APPARENT PREVALENCE, DIVERSITY, AND ASSOCIATED LESIONS OF PERIORBITAL NEMATODES IN A POPULATION OF BARRED OWLS (STRIX VARIA) FROM NORTHERN CALIFORNIA, USA

Barred Owls (Strix varia) are a common owl species found in forest habitats across North America (Livezey 2007). Although historically found predominantly in eastern North America, they have expanded their range over the last four decades through much of the western part of the continent, including California (Dark et al. 1998; Gutiérrez et al. 2007; Franklin et al. 2021). This expansion into coastal Washington, Oregon, and California has been proposed as a potential detriment to the maintenance of the federally threatened Northern Spotted Owl (Strix occidentalis caurina) because of potential hybridization between the species, predation, and competition for habitat and food (Dark et al. 1998; Wiens et al. 2014; Franklin et al. 2021; Wiens et al. 2021). As a result, removal of Barred Owls from Spotted Owl habitat has been used as an experimental management technique to maintain Spotted Owl populations (Kelly et al. 2003; Diller et al. 2016; Wiens et al. 2021).

Periocular nematodes, including those in the genera Ceratospira, Aprocta, Thelazia, and Oxyspirura, have been described in a variety of avian species including multiple species of owls and other raptors (Purwaningsih 1993; Suedmeyer et al. 1999; Willis and Wilkie 1999; Rodriguez-Tovar et al. 2008; Okulewicz and Sitko 2012). Most ocular nematode infections are considered incidental or a rare cause of morbidity in isolated cases (Beckmann et al. 2014). Since 2010, increasing reports have implicated infection by the nematode Oxyspirura petrowi in Northern Bobwhite Quail (Colinus virginianus) and Lesser Prairie Chickens (Tympanuchus pallidicinctus) affecting host fitness, although the effects of periorbital nematodes are not completely understood at the individual or population level (Dunham et al. 2014a, b; Olsen et al. 2016; Henry et al. 2017). Recently, a human was infected by a nematode closely related to O. petrowi, further emphasizing the increasing importance of studying periorbital nematodes (Dung et al. 2020).

A population of Barred Owls in Humboldt County, California, was observed to be infected by an unknown periorbital nematode during routine museum processing. Because of the potential effect of these parasites on the health of Barred Owls and Spotted Owls, an investigation into these parasites was initiated. The aims of our study were to determine the apparent prevalence and identity of the periorbital nematodes of these Barred Owls as well as potential associations of infection and burden with owl body weight. Additionally, histopathologic lesions were evaluated in a small subset of infected and noninfected owls to determine the severity of any lesions.

Barred owls from Humboldt County, California, were harvested by gunshot as part of a Spotted Owl restoration effort within the boundary of the Hoopa Valley Indian Reservation under United States Fish and Wildlife Service Scientific Collecting permit MB14305B-0. Data from each bird obtained in the field included the bird's body weight by spring scale and sex (by direct identification of sex organs); heart blood and liver samples were also collected. Owls were frozen at –20 C until future processing.

A total of 69 Barred Owls harvested from 2016 to 2020 were thawed and examined at the Southeastern Cooperative Wildlife Disease Study (SCWDS; University of Georgia, Athens, Georgia, USA) or the School of Veterinary Medicine, University of California-Davis (Davis, California, USA). In all cases, the periorbital tissues (conjunctivae, third eyelid, lacrimal duct, retrobulbar area, and Harderian gland) were dissected and examined for the presence of grossly visible nematodes within or adjacent to these structures. The number of nematodes surrounding each eye and in each owl was recorded. Gross examination was limited to the cornea; sclera; conjunctivae; lacrimal bone, gland, and duct; and retrobulbar space.

When nematodes were detected, they were extracted, placed into 70% ethanol, and stored for later evaluation or were kept in situ within the conjunctivae for formalin fixation. The extracted nematodes were cleared in either lactophenol or glycerol solution to enable evaluation of morphologic structures and were identified to genus with the use of multiple taxonomic keys (Yamaguti 1961; Anderson and Bain 1974; Chabaud 1974).

Nematodes also were identified phylogenetically by molecular methods with reference to published sequences by PCR and sequencing targeting the 18S rRNA with primers NEMFG1 and CRYPTOR as described by Vanstreels et al. (2018). Because our initial morphologic examinations indicated that some of the Barred Owl worms were an Aprocta sp., we obtained an ethanol-fixed specimen of Aprocta cylindrica from a European Robin (Erithacus rubecula) from a collaborator and obtained the 18S rRNA sequence as described above (Beckmann et al. 2014). Consensus sequences obtained from GenBank were produced and aligned with sequences from related nematodes in MEGA 7 (Kumar et al. 2016). A phylogenetic tree was constructed by maximum likelihood algorithms that used the Tamura-Nei parameter model and pairwise deletion in MEGA 7.

A subset of 45 eyes with surrounding conjunctiva and third eyelid were placed in 10% neutral buffered formalin and left to fix for at least 48 h after extraction and evaluation for grossly visible nematodes. Next, the orbits and surrounding conjunctiva were placed into 15% formic acid for 24 h to decalcify the ossicular ring. Singular sagittal sections of the orbit and surrounding conjunctivae were dehydrated in ethanol and embedded in paraffin wax. Sections 5 µm thick were cut, stained with H&E, and evaluated by light microscopy by a board-certified veterinary anatomic pathologist and trainee.

Apparent prevalence of eyeworms was calculated for males and females and compared by a chi-square test. Analysis of variance (ANOVA) tests were used to compare nematode abundance and infection burden, as well as weight between males and females and between infected and uninfected owls of each sex. Additionally, generalized linear regression models were used to predict nematode burden by individual, additive, and interactive effects of sex and weight. Models were evaluated by an information theoretic approach (Blankenship et al. 2002). Statistical analyses were performed in RStudio version 1.4.1717 (R Core Team 2020) with statistical significance assessed at α=0.05.

We examined 69 owls, including 29 collected in 2016, 13 in 2018, three in 2019, and 24 in 2020. Overall, 34 owls (49%) were infected with at least one periorbital nematode. Of the infected owls, 2/34 (5.9%) had nematodes only in the right orbit, 12/34 (35.3%) had nematodes only in the left orbit, and 20/34 (58.8%) had nematodes in both orbits. For the infected owls, the medians and ranges of nematode burdens in the right orbit, left orbit, and combined were 3 (1–13), 3 (1–16), and 5 (1–24), respectively. Nematodes were present within the conjunctival folds in 28/34 (82%) of the infected owls; two owls had worms extending visibly onto the surface of the cornea (Fig. 1A, B). Nematodes were within the lacrimal duct of 9/34 (27%). In 4/34 (12%) they were found behind the lacrimal bone and in the retrobulbar space (Fig. 1C). Gross lesions associated with eye worm infection were not detected.

A total of 38 eyes and associated periorbital tissues with grossly visible nematode infection and seven eyes and associated periorbital tissues without visible infection were examined histologically. Postmortem autolysis and freezing artifact were moderate to marked in all tissue sets. In 25/38 (66%) infected tissue sets, lymphocytes and rare plasma cells and heterophils were present in variably sized but typically small scattered clusters in the substantia propria of the conjunctiva and, less commonly, at the limbus (Fig. 1D, E). Segmental erosions or ulcerations were present on the conjunctival epithelium in 8/38 tissue sets (21%), and moderate to severe goblet cell hyperplasia was observed in 6/38 (16%) tissue sets of infected conjunctivae. Sections of nematodes were microscopically visible in 15/38 (40%) tissue sets and had two distinct morphologic profiles. In 12/15 cases (80%), the nematodes ranged from 250 to 400 µm in diameter and had, when visible, cuticular spikes, prominent intestines (Fig. 1F), a tri-radiated esophagus, eosinophilic coelomic fluid, and embryonated eggs consistent with Spiruroidea (Gardiner and Poynton 1999). In three cases (20%), the nematodes were 450–500 µm in diameter and had much lower coelomyarian musculature, prominent lateral chords, and a small, simple intestinal tract consistent with Filarioidea (Gardiner and Poynton 1999). In 2/7 (29%) eyes without grossly or histologically visibly nematodes, similar mild lymphoplasmacytic and heterophilic inflammation was present, but no other significant lesions. No obvious distinction was noted between parasite species with respect to associated histologic lesions.

A total of 115 nematodes from 17/34 infected owls were examined morphologically. The nematodes had two distinct morphologies and were identified to genus by three keys (Yamaguti 1961; Anderson and Bain 1974; Chabaud 1974). Briefly, one group had a smooth, transversely striated cuticle; a simple, unbound mouth without lips; a short, simple, nondividing esophagus; and rounded anterior (Fig. 2A) and posterior ends (Fig. 2G) in both sexes. Females had the vulva in the cranial, esophageal region (Fig. 2C), and eggs were small and thick shelled. Males had equal to slightly subequal spicules (Fig. 2E) and lacked caudal alae. These features were consistent with Aprocta spp. Detailed morphologic descriptions are depicted in Supplementary Material Table S1. The second group had cervical alae in both sexes (Fig. 2B), a well-developed buccal cavity, a divided esophagus, an absent preanal sucker, and a pointed, incurved tail (Fig. 2H). The males had unequal or dissimilar spicules (Fig. 2F). In females, the vulva was posteriorly placed (Fig. 2D), and the eggs lacked polar plugs. These features were consistent with Oxyspirura spp. Specimens from each genus were submitted to the National Parasite Collection at the Smithsonian under US National Museum (USNM) accession numbers (USNM 1522364) for Oxyspirura and (USNM 1522363) for Aprocta.

Fourteen of the 17 owls (82%) from which worms were collected had exclusively Oxyspirura sp. infections, totaling 98 of the 115 (85%) nematodes examined, whereas 1/17 owls (6%) had exclusively Aprocta sp. infection. The remaining two (12%) owls were co-infected with Aprocta and Oxyspirura spp.; in these cases, each nematode species was present in only one eye.

The 18S rRNA gene sequence (1,730 bp) of the Oxyspirura sp. had 97.4% sequence identity to three sequences of O. petrowi from Northern Bobwhite Quail from Texas (1,340–1,350 bp overlap, see Supplementary Material Table S2). The owl sequence was 97.5% similar to a shorter Oxyspirura sp. sequence (885 bp) from a human in Vietnam (Dung et al. 2020). Phylogenetically, the owl sequence grouped with the three O. petrowi sequences (Fig. 3). The Aprocta sp. sequence (1,860 bp) was approximately 95–96% similar to numerous Spiruromorpha species including the only Aproctoidea sequence available in GenBank (MG805661, Diomedenema diomedeae; see Supplementary Material Table S1). The sequence of the A. cylindrica specimen from a European Robin (1,736 bp long) was 97.5% similar to the Barred Owl Aprocta sp. Phylogenetically, the Barred Owl Aprocta sp. grouped with A. cylindrica within a clade including D. diomedeae and a member of the related Diplotriaenidae (Serratospiculum ten-do; Fig. 3). The Oxyspirura sp., Aprocta sp., and A. cylindrica sequences were deposited into GenBank under accession numbers OQ474909, OQ474908, and OQ474907, respectively.

The apparent prevalence of eyeworms was significantly higher in males (52%) compared with females (19%; χ²=6.438; df=1; P=0.0112), but nematode burden by sex was not different. Owl weights were significantly higher for females (mean 787.6 g) compared with males (mean 668.0 g; P<0.0001). Infected female owls weighed significantly more than uninfected females (mean 852 g vs. 767.5 g; F-ratio=4.70776, P=0.043), but there was no significant difference between infected males (mean 676.6 g) and uninfected males (mean 658.7 g; f-ratio=1.8202, P=0.184).

Sex was a significant predictor of nematode burden in the models (P<0.0001), but weight was only significant when included in models with sex. The top model predicting nematode burden was well supported (ω_i=0.634) and included the additive effects of sex and weight (Supplementary Material Table S3). According to this model, males were more likely to have higher nematode burdens, as were heavier animals.

This study describes high apparent prevalence of periorbital nematodes in a free-ranging population of Barred Owls. In contrast, most publications on periorbital nematodes in birds are related to morphologic and taxonomic descriptions in individual animals or isolated case reports of infection with or without clinical signs or associated mortality. Descriptions of avian orbital nematodes affecting populations of free-ranging birds are rare, with the exception of O. petrowi, a well-known and widespread parasite of wild birds (Dunham et al. 2014b; Henry et al. 2017).

Aside from the presence of nematodes, no grossly apparent evidence of conjunctivitis or other gross lesions were observed in infected birds, although significant autolysis, freeze artifacts, or lesions related to harvesting may have complicated interpretation of subtle changes. Similar to other studies (Bruno et al. 2015; Dunham et al. 2016b), most nematodes in these owls were present in conjunctival folds and less commonly in lacrimal glands and ducts. Oxyspirura petrowi has been reported in the nasal sinuses of quail (Dunham et al. 2014a) and Aprocta sp. in the peritoneum and mandible of owls in China (Zhang et al. 2008). The Oxyspirura sp. infection intensity in our owls was lower than that reported for O. petrowi in Bobwhite Quail in the Rolling Plains region of Texas (Dunham et al. 2016a) but was more similar to Bobwhite Quail in southern Texas (Shea et al. 2021). Whether owls tolerate eye worm infection better than do quail is unclear, perhaps because of the overall larger size of owl eyes and relatively lower parasite intensity, or whether periorbital nematodes in owls are inherently less pathogenic.

Although periorbital nematodes infect numerous avian species worldwide, there is a paucity of sequences from these two groups. Currently the only near–full-length 18S rRNA gene Oxyspirura sequences available in GenBank are from O. petrowi from Bobwhite Quail and intermediate hosts and from uncharacterized larvae from a human in Vietnam with an unknown natural host (Dung et al. 2020). Little is known about interspecific genetic variation among these nematode species, but on the basis of 18S rRNA gene sequence analysis, the Barred Owl parasite does not appear to be O. petrowi. For Aprocta spp., before this study no sequences had been submitted to GenBank, and the only other near–full-length 18S rRNA gene sequence from the Aproctoidea was from D. diomedeae from an albatross (a short sequence is available from Desmidocercella australis; Vanstreels et al. 2018), hence our decision to obtain an A. cylindrica specimen for comparison with our unknown species.

Surveys of eye worms in birds suggest a high diversity of hosts for both Oxyspirura and Aprocta spp. Future studies should include detailed morphologic characterization combined with sequence analysis of commonly used loci, including 18S and other ribosomal genes, to better understand the diversity of parasites within these hosts.

The only other reports of Oxyspirura in owl species in the Americas have been an unspecified Oxyspirura sp. detected in a captive fulvous owl (Strix fulvescens) from Mexico (Rodriguez-Tovar et al. 2008) and Oxyuris brevisubulata from a tropical screech owl (Megascops choliba) in Brazil (Ransom 1904). Both O. petrowi and Oxyspirura pusillae have been confirmed to occur in the US, and in addition to infections in quail and prairie chickens, other Oxyspirura spp. infections have been documented in mimids (Mimidae) from Texas (Dunham and Kendall 2014), a woodpecker (Melanerpes carolinus) in Florida (Foster et al. 2002), and multiple species in the orders Passeriformes, Pelecaniformes, and Piciformes in Louisiana (Pence 1972). Paucity of detailed morphologic comparisons between nematodes and lack of molecular data from past studies complicate our ability to determine the host-parasite relationships and diversity of Oxyspirura spp. in bird hosts.

Aprocta spp. have been reported in a variety of avian taxa from many continents including reports in psittacine, piciform, cuculiform, and galliform birds from South America (Freitas and Machado de Mendonça 1952; Machado de Mendonça 1961; de Oliveira Rodrigues and Sodre Rodrigues 1980; Pinto et al. 1997; Oniki et al. 2002; Tantaleán and Chavez 2004); in multiple passerine species in Africa (Quentin et al. 1976) and Asia (Makarenko 1959); and passerine birds including multiple corvid species in Europe (Skrjabin 1934; Manfredi et al. 1992; Beckmann et al. 2014). Aprocta nematodes also have been reported in several owl species in China, including Scops owls (Otus scops and Otus sunia), Little Owls (Athene noctua), and Eurasian Eagle Owls (Bubo bubo; Zhang et al. 2008, 2012).

Similar to previous descriptions of O. petrowi infections in quail (Bruno et al. 2015; Dunham et al. 2016b), we found that owls infected with Oxyspirura and Aprocta spp. had variably severe, predominantly mononuclear cell inflammation in the conjunctivae, third eyelid, Harderian gland, lacrimal gland, or a combination of these tissues. Additionally, these previous studies reported lymphoplasmacytic keratitis, corneal erosions and edema, and Harderian gland adenitis, fibrosis, and atrophy (Bruno et al. 2015; Dunham et al. 2016b). Mononuclear inflammation is a nondiagnostic change, and other infectious or immune-mediated disorders could have contributed to these lesions. Corneal lesions were not appreciated in the owls in our study, probably because their distribution was limited to the conjunctiva and rarely extended to the corneal surface; additionally, severe autolysis and freezing artifacts potentially hindered interpretation. Multiple sections of Harderian gland were evaluated in our study, and no lesions were detected, again possibly because of autolysis. Similarly, no gross or histologic lesions have been reported with Oxyspirura spp. infections in other species, including a Fulvous Owl and Lesser Prairie Chickens (Foster et al. 2002; Robel et al. 2003; Rodriguez-Tovar et al. 2008).

Definitive identification of the nematodes by morphology in tissue sections was challenging because of the potential overlap of key features in these groups, including the size, prominence of coelomyarian musculature, and larvated eggs. However, the presence of a tri-radiated esophagus, cuticular spines or crests, and a large gastrointestinal tract are more suggestive of Oxyspirura, whereas the small, simple intestinal tract and prominent lateral chords are more consistent with Aprocta (Gardiner and Poynton 1999; Beck-man et al. 2014). Overall, morphologic descriptions of these two genera in histologic sections are sparse in the literature, and narrowing the identification is more reliable if based on molecular methods or on morphologic features from isolated, intact nematodes.

When we compared the morphologic features of these two nematode groups with other avian Aprocta and Oxyspirura species (Supplementary Material Table S1), the Oxyspirura nematodes had morphologic features typically within the range of those previously described. Although not observed in cleared intact specimens, Ivanova et al. (2007) describe cuticular crests in ocular Oxyspirura spp. from primates with similar cuticular features as those seen in histologic sections in our study owls (Fig. 1F). However, these features are not unique to this genus and can be found in other spirurid nematodes. The limited number of intact male Aprocta specimens hindered the ability to draw conclusions on the differences in morphometry between our specimen and published references.

Which potential intermediate hosts might be involved in propagating periorbital nematode infection is unclear in this population of barred owls. Oxyspirura petrowi, for example, uses house crickets (Acheta domesticus), cockroaches (order Blattodea), and grasshoppers (Brachystola spp.) as intermediate hosts (Kistler et al. 2016; Almas et al. 2018; Kalyanasundaram et al. 2019; Henry et al. 2020). Aprocta cylindrica uses locusts as intermediate hosts (Quentin et al. 1976), but less is known about potential intermediate hosts in this genus compared with Oxyspirura. Additional studies to investigate the life cycle and intermediate host(s) of the nematodes infecting these Barred Owls are warranted, as is examining potential landscape changes in Northern California that may contribute to alterations in potential intermediate host populations.

Despite extensive autolysis and freezing artifacts that may have hindered interpretation of subtle changes, no gross or histologic evidence indicated that the lesions associated with nematode infection impaired the vision of these owls. It is important to emphasize that the owls in this study were seemingly normal at the time of harvest, without any overt clinical signs or behavioral abnormalities. Olsen et al. (2016) emphasized that experimental studies are the preferred method to study the potential health effects of orbital helminths in quail. Similar to others (Bruno et al. 2015; Dunham et al. 2016b), a relationship between histologic lesions and individual bird fitness was not clearly defined in our study because visual acuity was not explicitly evaluated, and inflammation appeared too mild to have affected vision. Future studies to score or quantify the severity of inflammation or other lesions in a larger number of well-preserved owls and to record how these lesions relate to burden or weight would be of value to further clarify any potential effects on these hosts. Targeted sampling and postmortem examinations of birds with overt illness would also help understand the potential of these eye worms to cause morbidity or mortality. Finally, nematode burden in these owls was higher in heavier birds, suggesting that fit birds may be more likely to catch prey harboring the nematodes rather than nematodes affecting an owl's ability to catch prey.

Our study highlights the importance of continued monitoring considering the conservation status of spotted owls. Further evaluation of Barred Owls throughout their range may provide additional insight on whether these parasites are endemic, were recently or historically introduced, and are host generalists. Similarly, examination and potential identification of these worms in clinically normal and dead, or clinically diseased, Spotted Owls may determine potential pathogenic consequences in this threatened host as well as shed light on the extent of host specificity.

The authors thank the Hoopa Tribe for funding as well as for access to research material. We also thank the Garden Wildlife Health project (www.gardenwildlifehealth.org), a wildlife disease surveillance program coordinated by the Institute of Zoology, UK, for donation of a sample of Aprocta cylindrica for phylogenetic comparison, and the histology laboratories at the Veterinary Medical Teaching Hospital, University of California–Davis and Athens Veterinary Diagnostic Laboratories. Funding was provided in part by the sponsorship of the Southeastern Cooperative Wildlife Disease Study (SCWDS) by the fish and wildlife agencies of Alabama, Arkansas, Florida, Georgia, Kentucky, Kansas, Louisiana, Maryland, Mississippi, Missouri, Nebraska, North Carolina, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Virginia, and West Virginia, USA. Support from the states to SCWDS was provided in part by the Federal Aid to Wildlife Restoration Act (50 Stat. 917). Additional funding was provided by the California Academy of Sciences. Becki Lawson receives financial support from Research England.

© Wildlife Disease Association 2023

Supplementary material for this article is online at http://dx.doi.org/10.7589/JWD-D-21-00186.