Highly Pathogenic Avian Influenza A(H5N1) Virus Clade 2.3.4.4b Infections in Wild Terrestrial Mammals, United States, 2022.

We describe the pathology of natural infection with highly pathogenic avian influenza A(H5N1) virus of Eurasian lineage Goose/Guangdong clade 2.3.4.4b in 67 wild terrestrial mammals throughout the United States during April 1‒July 21, 2022. Affected mammals include 50 red foxes (Vulpes vulpes), 6 striped skunks (Mephitis mephitis), 4 raccoons (Procyon lotor), 2 bobcats (Lynx rufus), 2 Virginia opossums (Didelphis virginiana), 1 coyote (Canis latrans), 1 fisher (Pekania pennanti), and 1 gray fox (Urocyon cinereoargenteus). Infected mammals showed primarily neurologic signs. Necrotizing meningoencephalitis, interstitial pneumonia, and myocardial necrosis were the most common lesions; however, species variations in lesion distribution were observed. Genotype analysis of sequences from 48 animals indicates that these cases represent spillover infections from wild birds.

Domestic and wild animal samples and diagnostic testing for SARS-CoV-2.

From the first cases in 2019, COVID-19 infections caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have resulted in over 6 million human deaths in a worldwide pandemic. SARS-CoV-2 is commonly spread from human to human through close contact and is capable of infecting both humans and animals. Worldwide, there have been over 675 animal outbreaks reported that resulted in over 2000 animal infections including domestic and wild animals. As the role of animal infections in the transmission, pathogenesis, and evolution of SARS-CoV-2 is still unfolding, accurate and reliable animal diagnostic tests are critical to aid in managing both human and animal health. This review highlights key animal samples and the three main diagnostic approaches used for animal testing: PCR, serology, and Next Generation Sequencing. Diagnostic results help inform (often difficult) clinical decision-making, but also possible ways to mitigate spread among pets, food supplies, or wildlife. A One Health approach has been key to monitoring the SARS-CoV-2 pandemic, as consistent human-animal interactions can lead to novel variants. Having multiple animal diagnostic tests for SARS-CoV-2 available is critical to ensure human, animal, and environmental health.

H5N1 highly pathogenic avian influenza clade 2.3.4.4b in wild and domestic birds: Introductions into the United States and reassortments, December 2021-April 2022.

Highly pathogenic avian influenza viruses (HPAIVs) of the A/goose/Guangdong/1/1996 lineage H5 clade 2.3.4.4b continue to have a devastating effect on domestic and wild birds. Full genome sequence analyses using 1369 H5N1 HPAIVs detected in the United States (U.S.) in wild birds, commercial poultry, and backyard flocks from December 2021 to April 2022, showed three phylogenetically distinct H5N1 virus introductions in the U.S. by wild birds. Unreassorted Eurasian genotypes A1 and A2 entered the Northeast Atlantic states, whereas a genetically distinct A3 genotype was detected in Alaska. The A1 genotype spread westward via wild bird migration and reassorted with North American wild bird avian influenza viruses. Reassortments of up to five internal genes generated a total of 21 distinct clusters; of these, six genotypes represented 92% of the HPAIVs examined. By phylodynamic analyses, most detections in domestic birds were shown to be point-source transmissions from wild birds, with limited farm-to-farm spread.

Successful Detection of Delta and Omicron Variants of SARS-CoV-2 by Veterinary Diagnostic Laboratory Participants in an Interlaboratory Comparison Exercise.

BACKGROUND: Throughout the COVID-19 pandemic, veterinary diagnostic laboratories have tested diagnostic samples for SARS-CoV-2 both in animals and over 6 million human samples. An evaluation of the performance of those laboratories is needed using blinded test samples to ensure that laboratories report reliable data to the public. This interlaboratory comparison exercise (ILC3) builds on 2 prior exercises to assess whether veterinary diagnostic laboratories can detect Delta and Omicron variants spiked in canine nasal matrix or viral transport medium. METHODS: The ILC organizer was an independent laboratory that prepared inactivated Delta variant at levels of 25 to 1000 copies per 50 µL of nasal matrix for blinded analysis. Omicron variant at 1000 copies per 50 µL of transport medium was also included. Feline infectious peritonitis virus (FIPV) RNA was used as a confounder for specificity assessment. Fourteen test samples were prepared for each participant. Participants used their routine diagnostic procedures for RNA extraction and real-time reverse transcriptase-PCR. Results were analyzed according to International Organization for Standardization (ISO) 16140-2:2016. RESULTS: Overall, laboratories demonstrated 93% detection for Delta and 97% for Omicron at 1000 copies per 50 µL. Specificity was 97% for blank samples and 100% for blank samples with FIPV. No differences in Cycle Threshold (Ct) values were significant for samples with the same virus levels between N1 and N2 markers, nor between the 2 variants. CONCLUSIONS: The results indicated that all ILC3 participants were able to detect both Delta and Omicron variants. The canine nasal matrix did not significantly affect SARS-CoV-2 detection.

Comparison of magnetic bead and rapid swab RNA extraction methods for detecting rabbit hemorrhagic disease virus 2 in rabbit liver samples.

We compared a bead RNA extraction method with a one-tube method that required only a heat block and ice. RNA was first extracted from liver samples from nine rabbits dying from rabbit hemorrhagic disease virus 2 (RHDV2) using magnetic beads, and RT-PCR was used to detect RHDV2 sequence. Following freezing, RNA was extracted a second time using the SwiftX™ Swabs Viral RNA Extraction Reagent. RHDV2 was detected in all nine samples. Cycle threshold values were higher in the RT-PCR following SwiftX extraction (mean: 3.79), indicating that the second extraction method resulted in approximately a 1 log10 reduction in sensitivity. A second freeze-thaw for the samples and less tissue extracted using SwiftX may have contributed additionally to the loss in sensitivity.

An Evaluation of Avian Influenza Virus Whole-Genome Sequencing Approaches Using Nanopore Technology.

As exemplified by the global response to the SARS-CoV-2 pandemic, whole-genome sequencing played an important role in monitoring the evolution of novel viral variants and provided guidance on potential antiviral treatments. The recent rapid and extensive introduction and spread of highly pathogenic avian influenza virus in Europe, North America, and elsewhere raises the need for similarly rapid sequencing to aid in appropriate response and mitigation activities. To facilitate this objective, we investigate a next-generation sequencing platform that uses a portable nanopore sequencing device to generate and present data in real time. This platform offers the potential to extend in-house sequencing capacities to laboratories that may otherwise lack resources to adopt sequencing technologies requiring large benchtop instruments. We evaluate this platform for routine use in a diagnostic laboratory. In this study, we evaluate different primer sets for the whole genome amplification of influenza A virus and evaluate five different library preparation approaches for sequencing on the nanopore platform using the MinION flow cell. A limited amplification procedure and a rapid procedure are found to be best among the approaches taken.

Immunogenicity, Safety, and Anti-Viral Efficacy of a Subunit SARS-CoV-2 Vaccine Candidate in Captive Black-Footed Ferrets (Mustela nigripes) and Their Susceptibility to Viral Challenge.

A preliminary vaccination trial against the emergent pathogen, SARS-CoV-2, was completed in captive black-footed ferrets (Mustela nigripes; BFF) to assess safety, immunogenicity, and anti-viral efficacy. Vaccination and boosting of 15 BFF with purified SARS-CoV-2 S1 subunit protein produced a nearly 150-fold increase in mean antibody titers compared to pre-vaccination titers. Serum antibody responses were highest in young animals, but in all vaccinees, antibody response declined rapidly. Anti-viral activity from vaccinated and unvaccinated BFF was determined in vitro, as well as in vivo with a passive serum transfer study in mice. Transgenic mice that received BFF serum transfers and were subsequently challenged with SARS-CoV-2 had lung viral loads that negatively correlated (p < 0.05) with the BFF serum titer received. Lastly, an experimental challenge study in a small group of BFF was completed to test susceptibility to SARS-CoV-2. Despite viral replication and shedding in the upper respiratory tract for up to 7 days post-challenge, no clinical disease was observed in either vaccinated or naive animals. The lack of morbidity or mortality observed indicates SARS-CoV-2 is unlikely to affect wild BFF populations, but infected captive animals pose a potential risk, albeit low, for humans and other animals.

Second round of an interlaboratory comparison of SARS-CoV2 molecular detection assays used by 45 veterinary diagnostic laboratories in the United States.

The COVID-19 pandemic presents a continued public health challenge. Veterinary diagnostic laboratories in the United States use RT-rtPCR for animal testing, and many laboratories are certified for testing human samples; hence, ensuring that laboratories have sensitive and specific SARS-CoV2 testing methods is a critical component of the pandemic response. In 2020, the FDA Veterinary Laboratory Investigation and Response Network (Vet-LIRN) led an interlaboratory comparison (ILC1) to help laboratories evaluate their existing RT-rtPCR methods for detecting SARS-CoV2. All participating laboratories were able to detect the viral RNA spiked in buffer and PrimeStore molecular transport medium (MTM). With ILC2, Vet-LIRN extended ILC1 by evaluating analytical sensitivity and specificity of the methods used by participating laboratories to detect 3 SARS-CoV2 variants (B.1; B.1.1.7 [Alpha]; B.1.351 [Beta]) at various copy levels. We analyzed 57 sets of results from 45 laboratories qualitatively and quantitatively according to the principles of ISO 16140-2:2016. More than 95% of analysts detected the SARS-CoV2 RNA in MTM at ≥500 copies for all 3 variants. In addition, for nucleocapsid markers N1 and N2, 81% and 92% of the analysts detected ≤20 copies in the assays, respectively. The analytical specificity of the evaluated methods was >99%. Participating laboratories were able to assess their current method performance, identify possible limitations, and recognize method strengths as part of a continuous learning environment to support the critical need for the reliable diagnosis of COVID-19 in potentially infected animals and humans.

MORBIDITY AND MORTALITY OF HAWAIIAN GEESE (BRANTA SANDVICENSIS) AND LAYSAN ALBATROSS (PHOEBASTRIA IMMUTABILIS) ASSOCIATED WITH RETICULOENDOTHELIOSIS VIRUS.

Only one virus, Avipox, has been documented previously in wild birds in Hawaii. Using immunohistochemistry and PCR, we found that two native threatened Hawaiian Geese (Branta sandvicensis), one with multicentric histiocytoma and the other with toxoplasmosis, and one Laysan Albatross (Phoebastria immutabilis) with avian pox were infected with reticuloendotheliosis virus (REV). The virus was isolated from one of the geese by cell culture. Surveys of other Hawaiian geese with various pathologies, avian pox cases, and pox viral isolates using PCR failed to reveal REV, suggesting that the virus is uncommon, at least in samples examined. The full genome of the Gag, Pol, and Env genes were sequenced for all three infected birds and revealed geographic divergence of the Pol gene, suggesting it to be under strong selective pressure. Our finding of REV in Hawaii makes this only the second virus documented in native Hawaiian birds associated with pathology. Moreover, the presence of REV in a pelagic seabird is unusual. Future surveys should seek the reservoir of the virus in efforts to trace its origins.

Spatiotemporal changes in influenza A virus prevalence among wild waterfowl inhabiting the continental United States throughout the annual cycle.

Avian influenza viruses can pose serious risks to agricultural production, human health, and wildlife. An understanding of viruses in wild reservoir species across time and space is important to informing surveillance programs, risk models, and potential population impacts for vulnerable species. Although it is recognized that influenza A virus prevalence peaks in reservoir waterfowl in late summer through autumn, temporal and spatial variation across species has not been fully characterized. We combined two large influenza databases for North America and applied spatiotemporal models to explore patterns in prevalence throughout the annual cycle and across the continental United States for 30 waterfowl species. Peaks in prevalence in late summer through autumn were pronounced for dabbling ducks in the genera Anas and Spatula, but not Mareca. Spatially, areas of high prevalence appeared to be related to regional duck density, with highest predicted prevalence found across the upper Midwest during early fall, though further study is needed. We documented elevated prevalence in late winter and early spring, particularly in the Mississippi Alluvial Valley. Our results suggest that spatiotemporal variation in prevalence outside autumn staging areas may also represent a dynamic parameter to be considered in IAV ecology and associated risks.

An Opportunistic Survey Reveals an Unexpected Coronavirus Diversity Hotspot in North America.

In summer 2020, Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) was detected on mink farms in Utah. An interagency One Health response was initiated to assess the extent of the outbreak and included sampling animals from on or near affected mink farms and testing them for SARS-CoV-2 and non-SARS coronaviruses. Among the 365 animals sampled, including domestic cats, mink, rodents, raccoons, and skunks, 261 (72%) of the animals harbored at least one coronavirus. Among the samples that could be further characterized, 127 alphacoronaviruses and 88 betacoronaviruses (including 74 detections of SARS-CoV-2 in mink) were identified. Moreover, at least 10% (n = 27) of the coronavirus-positive animals were found to be co-infected with more than one coronavirus. Our findings indicate an unexpectedly high prevalence of coronavirus among the domestic and wild free-roaming animals tested on mink farms. These results raise the possibility that mink farms could be potential hot spots for future trans-species viral spillover and the emergence of new pandemic coronaviruses.

Interlaboratory comparison of SARS-CoV2 molecular detection assays in use by U.S. veterinary diagnostic laboratories.

The continued search for intermediate hosts and potential reservoirs for SARS-CoV2 makes it clear that animal surveillance is critical in outbreak response and prevention. Real-time RT-PCR assays for SARS-CoV2 detection can easily be adapted to different host species. U.S. veterinary diagnostic laboratories have used the CDC assays or other national reference laboratory methods to test animal samples. However, these methods have only been evaluated using internal validation protocols. To help the laboratories evaluate their SARS-CoV2 test methods, an interlaboratory comparison (ILC) was performed in collaboration with multiple organizations. Forty-four sets of 19 blind-coded RNA samples in Tris-EDTA (TE) buffer or PrimeStore transport medium were shipped to 42 laboratories. Results were analyzed according to the principles of the International Organization for Standardization (ISO) 16140-2:2016 standard. Qualitative assessment of PrimeStore samples revealed that, in approximately two-thirds of the laboratories, the limit of detection with a probability of 0.95 (LOD95) for detecting the RNA was ≤20 copies per PCR reaction, close to the theoretical LOD of 3 copies per reaction. This level of sensitivity is not expected in clinical samples because of additional factors, such as sample collection, transport, and extraction of RNA from the clinical matrix. Quantitative assessment of Ct values indicated that reproducibility standard deviations for testing the RNA with assays reported as N1 were slightly lower than those for N2, and they were higher for the RNA in PrimeStore medium than those in TE buffer. Analyst experience and the use of either a singleplex or multiplex PCR also affected the quantitative ILC test results.

Pathology of Lagovirus europaeus GI.2/RHDV2/b (Rabbit Hemorrhagic Disease Virus 2) in Native North American Lagomorphs.

Rabbit hemorrhagic disease, a notifiable foreign animal disease in the US, was reported for the first time in wild native North American lagomorphs in April 2020 in the southwestern US. Affected species included the desert cottontail (Sylvilagus audubonii), mountain cottontail (Sylvilagus nuttallii), black-tailed jackrabbit (Lepus californicus), and antelope jackrabbit (Lepus alleni). Desert cottontails (n=7) and black-tailed jackrabbits (n=7) collected in April and May 2020 were necropsied at the US Geological Survey National Wildlife Health Center and tested positive for Lagovirus europaeus GI.2, also known as rabbit hemorrhagic disease virus 2 (GI.2/RHDV2/b), by real-time PCR at the US Department of Agriculture's Foreign Animal Disease Diagnostic Laboratory. Gross and microscopic lesions were similar to those reported in European rabbits (Oryctolagus cuniculus) and other hare (Lepus) species with GI.2/RHDV2/b infection; they included epistaxis (12/13; 92%); massive hepatocellular dissociation (14/14; 100%) and necrosis or apoptosis (11/11; 100%); pulmonary congestion (12/12; 100%), edema (12/13; 92%), and hemorrhage (11/12; 92%); and acute renal tubular injury (3/8; 38%). As in previous reports, massive hepatocellular dissociation and necrosis or apoptosis was the most diagnostically distinct finding. As North American Sylvilagus and Lepus species appear to be susceptible to fatal GI.2/RHDV2/b infection, additional work is needed to understand the host range, pathogenicity, and potential population effects of GI.2/RHDV2/b in North America.

Emergence and molecular characterization of pigeon Paramyxovirus-1 in non-native Eurasian collared doves (Streptopelia decaocto) in California, USA.

Eurasian collared doves (Streptopelia decaocto) were introduced into Florida in the 1980s and have since established populations throughout the continental United States. Pigeon paramyxovirus-1 (PPMV-1), a species-adapted genotype VI Avian orthoavulavirus 1, has caused periodic outbreaks among collared doves in the U.S. since 2001 with outbreaks occasionally involving native doves. In California, PPMV-1 mortality events were first documented in Riverside County in 2014 with subsequent outbreaks in 23 additional counties from southern to northern California between 2015 and 2019. Affected collared doves exhibited torticollis and partial paralysis. Pale kidneys were frequently visible on gross necropsy (65.4%; 51/78) while lymphoplasmacytic interstitial nephritis often with acute tubular necrosis (96.0%; 24/25) and pancreatic necrosis (80.0%; 20/25) were common findings on histopathology. In total, PPMV-1 was confirmed by rRT-PCR and sequence analysis from oropharyngeal and/or cloacal swabs in 93.0% (40/43) of the collared doves tested from 16 California counties. In 2017, Avian orthoavulavirus 1 was confirmed in a native mourning dove (Zenaida macroura) found dead during a PPMV-1 outbreak in collared doves by rRT-PCR from formalin-fixed paraffin-embedded (FFPE) tissues, after the initial rRT-PCR from swabs failed to detect the virus. Molecular sequencing of the fusion protein of isolates collected from collared doves during outbreaks in 2014, 2016, and 2017 identified two distinct subgenotypes, VIa and VIn. Subgenotype VIn has been primarily isolated from collared doves in the southern U.S., while VIa has been isolated from mixed avian species in the northeastern U.S., indicating two independent introductions into California. While populations of collared doves are not expected to be substantially impacted by this disease, PPMV-1 may pose a threat to already declining populations of native columbids. This threat could be assessed by monitoring native and non-native columbids for PPMV-1. Based on our study, swab samples may not be sufficient to detect infection in native columbids and may require the use of non-traditional diagnostic approaches, such as FFPE tissues, to ensure virus detection.

SARS-CoV-2 Exposure in Escaped Mink, Utah, USA.

In August 2020, outbreaks of coronavirus disease were confirmed on mink farms in Utah, USA. We surveyed mammals captured on and around farms for evidence of infection or exposure. Free-ranging mink, presumed domestic escapees, exhibited high antibody titers, suggesting a potential severe acute respiratory syndrome coronavirus 2 transmission pathway to native wildlife.

Experimental challenge of a North American bat species, big brown bat (Eptesicus fuscus), with SARS-CoV-2.

The recently emerged novel coronavirus, SARS-CoV-2, is phylogenetically related to bat coronaviruses (CoVs), specifically SARS-related CoVs from the Eurasian bat family Rhinolophidae. As this human pandemic virus has spread across the world, the potential impacts of SARS-CoV-2 on native North American bat populations are unknown, as is the ability of North American bats to serve as reservoirs or intermediate hosts able to transmit the virus to humans or to other animal species. To help determine the impacts of the pandemic virus on North American bat populations, we experimentally challenged big brown bats (Eptesicus fuscus) with SARS-CoV-2 under BSL-3 conditions. We inoculated the bats both oropharyngeally and nasally, and over the ensuing three weeks, we measured infectivity, pathology, virus concentrations in tissues, oral and rectal virus excretion, virus transmission, and clinical signs of disease. We found no evidence of SARS-CoV-2 infection in any examined bat, including no viral excretion, no transmission, no detectable virus in tissues, and no signs of disease or pathology. Based on our findings, it appears that big brown bats are resistant to infection with the SARS-CoV-2. The potential susceptibility of other North American bat species to SARS-CoV-2 remains to be investigated.

Possibility for reverse zoonotic transmission of SARS-CoV-2 to free-ranging wildlife: A case study of bats.

The COVID-19 pandemic highlights the substantial public health, economic, and societal consequences of virus spillover from a wildlife reservoir. Widespread human transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) also presents a new set of challenges when considering viral spillover from people to naïve wildlife and other animal populations. The establishment of new wildlife reservoirs for SARS-CoV-2 would further complicate public health control measures and could lead to wildlife health and conservation impacts. Given the likely bat origin of SARS-CoV-2 and related beta-coronaviruses (ß-CoVs), free-ranging bats are a key group of concern for spillover from humans back to wildlife. Here, we review the diversity and natural host range of ß-CoVs in bats and examine the risk of humans inadvertently infecting free-ranging bats with SARS-CoV-2. Our review of the global distribution and host range of ß-CoV evolutionary lineages suggests that 40+ species of temperate-zone North American bats could be immunologically naïve and susceptible to infection by SARS-CoV-2. We highlight an urgent need to proactively connect the wellbeing of human and wildlife health during the current pandemic and to implement new tools to continue wildlife research while avoiding potentially severe health and conservation impacts of SARS-CoV-2 "spilling back" into free-ranging bat populations.

Detection of SARS-CoV-2 by RNAscope® in situ hybridization and immunohistochemistry techniques.

In situ hybridization (ISH) and immunohistochemistry (IHC) are essential tools to characterize SARS-CoV-2 infection and tropism in naturally and experimentally infected animals and also for diagnostic purposes. Here, we describe three RNAscope®-based ISH assays targeting the ORF1ab, spike, and nucleocapsid genes and IHC assays targeting the spike and nucleocapsid proteins of SARS-CoV-2.

Investigation of the 2018 thick-billed murre (Uria lomvia) die-off on St. Lawrence Island rules out food shortage as the cause.

Die-offs of seabirds in Alaska have occurred with increased frequency since 2015. In 2018, on St. Lawrence Island, seabirds were reported washing up dead on beaches starting in late May, peaking in June, and continuing until early August. The cause of death was documented to be starvation, leading to the conclusion that a severe food shortage was to blame. We use physiology and colony-based observations to examine whether food shortage is a sufficient explanation for the die-off, or if evidence indicates an alternative cause of starvation such as disease. Specifically, we address what species were most affected, the timing of possible food shortages, and food shortage severity in a historical context. We found that thick-billed murres (Uria lomvia) were most affected by the die-off, making up 61% of all bird carcasses encountered during beach surveys. Thick-billed murre carcasses were proportionately more numerous (26:1) than would be expected based on ratios of thick-billed murres to co-occurring common murres (U. aalge) observed on breeding study plots (7:1). Concentrations of the stress hormone corticosterone, a reliable physiological indicator of nutritional stress, in thick-billed murre feathers grown in the fall indicate that foraging conditions in the northern Bering Sea were poor in the fall of 2017 and comparable in severity to those experienced by murres during the 1976-1977 Bering Sea regime shift. Concentrations of corticosterone in feathers grown during the pre-breeding molt indicate that foraging conditions in late winter 2018 were similar to previous years. The 2018 murre egg harvest in the village of Savoonga (on St. Lawrence Is.) was one-fifth the 1993-2012 average, and residents observed that fewer birds laid eggs in 2018. Exposure of thick-billed murres to nutritional stress in August, however, was no different in 2018 compared to 2016, 2017, and 2019, and was comparable to levels observed on St. George Island in 2003-2017. Prey abundance, measured by the National Oceanic and Atmospheric Administration in bottom-trawl surveys, was also similar in 2018 to 2017 and 2019, supporting the evidence that food was not scarce in the summer of 2018 in the vicinity of St. Lawrence Island. Of two moribund thick-billed murres collected at the end of the mortality event, one tested positive for a novel re-assortment H10 strain of avian influenza with Eurasian components, likely contracted during the non-breeding season. It is not currently known how widely spread infection of murres with the novel virus was, thus insufficient evidence exists to attribute the die-off to an outbreak of avian influenza. We conclude that food shortage alone is not an adequate explanation for the mortality of thick-billed murres in 2018, and highlight the importance of rapid response to mortality events in order to document alternative or confounding causes of mortality.

Artificial intelligence and avian influenza: Using machine learning to enhance active surveillance for avian influenza viruses.

Influenza A viruses are one of the most significant viral groups globally with substantial impacts on human, domestic animal and wildlife health. Wild birds are the natural reservoirs for these viruses, and active surveillance within wild bird populations provides critical information about viral evolution forming the basis of risk assessments and countermeasure development. Unfortunately, active surveillance programs are often resource-intensive, and thus, enhancing programs for increased efficiency is paramount. Machine learning, a branch of artificial intelligence applications, provides statistical learning procedures that can be used to gain novel insights into disease surveillance systems. We use a form of machine learning, gradient boosted trees, to estimate the probability of isolating avian influenza viruses (AIV) from wild bird samples collected during surveillance for AIVs from 2006 to 2011 in the United States. We examined several predictive features including age, sex, bird type, geographic location and matrix gene rRT-PCR results. Our final model had high predictive power and only included geographic location and rRT-PCR results as important predictors. The highest predicted viral isolation probability was for samples collected from the north-central states and the south-eastern region of Alaska. Lower rRT-PCR Ct-values are associated with increased likelihood of AIV isolation, and the model estimated 16% probability of isolating AIV from samples declared negative (i.e., ≥35 Ct-value) using the rRT-PCR screening test and standard protocols. Our model can be used to prioritize previously collected samples for isolation and rapidly evaluate AIV surveillance designs to maximize the probability of viral isolation given limited resources and laboratory capacity.