Leucocytozoon Prevalence Differs by Sex in Louisiana Wild Turkeys (Meleagris gallopavo) Open Access

Avian malaria parasites (Plasmodium) and related Haemosporidia (Haemoproteus and Leucocytozoon) are arthropod-borne pathogens that represent a major threat to avifauna worldwide, including zoo, agricultural, and free-living bird populations (1,2,3,4,5,6). Plasmodium is transmitted by Culicidae mosquitos (7,8), Haemoproteus by louse flies (Hippoboscidae), and biting midges (Culicoides) (9,10), and Leucocytozoon by black flies (Simuliidae), and biting midges (11,12,13). Sexual reproduction of these parasites occurs in their respective insect vectors, while birds serve as intermediate hosts for asexual amplification that bolsters their geographic spread (14). Much progress has been made in recent years in understanding the biology of Haemosporidian parasites and the diseases they cause in wild birds, but knowledge on Leucocytozoon parasites lags in comparison to Plasmodium and Haemoproteus species (12).

The disease caused by Leucocytozoon species (leucocytozoonosis) is similar to that caused by Plasmodium and Haemoproteus species, with symptoms including anemia, appetite loss, lethargy, debilitation, and loss of muscle coordination (12,14,15). These parasites can sequester in host organs, more severely in the brain, where they can cause convulsions and paralysis before death (12). However, adult birds usually survive acute Leucocytozoon infection, and exo-erythrocytic parasites can remain dormant in the host tissues, with seasonal relapses and disease emergence in and from reservoir hosts (16).

Wild turkeys (Meleagris gallopavo) in North America are susceptible to infection by some Leucocytozoon species (i.e., Leucocytozoon andrewsi, also known as Leucocytozoon schoutedeni; and Leucocytozoon smithi). This has frequently been observed through 100% infection prevalence in turkeys (17,18,19,20,21), often higher than in nearby domestic fowl (18,22). Likewise, wild turkeys were the only species among six land birds and waterfowl to have 100% infection success during natural and experimental inoculations (19,23,24). Leucocytozoon outbreaks in domestic turkey farms historically had high mortality rates as well as reduced egg production, egg fertility, and hatching success for turkey hens (25) before production was moved indoors to eliminate disease transmission (22,26,27). However, turkeys have historically not been infected by some Leucocytozoon species (notably, Leucocytozoon sabrazesi (28), also known as Leucocytozoon macleani; Leucocytozoon caulleryi and have resisted infection by others (e.g., Leucocytozoon anseris, also known as Leucocytozoon simondi (29,30)), suggesting avian host specificity of some Leucocytozoon species. The high prevalence of some Leucocytozoon infections in wild and domestic turkeys and the long-term persistence of gametocytes in the peripheral blood (e.g., L. smithi (31)) strongly suggests that wild turkey populations serve as reservoirs of Leucocytozoon species in North America (26,32,33,34,35).

From a pre-European settlement estimate of one million Eastern wild turkeys in Louisiana (36,37), populations declined in the late 1800s because of clear-cutting of virgin longleaf pine forests and turkey overharvesting, leaving only 14 isolated flocks comprising ∼1500 individuals in North America after World War II (38,39). Population restoration began in 1962 (40), and the Louisiana Department of Wildlife and Fisheries (LDWF) has released 3913 captive-bred turkeys to the wild since 1962 (38). Conservation efforts were initially successful, increasing Louisiana turkey populations to 5000 individuals in 1969, 6500 in 1973, and 90,000 turkeys in 1995 (41). A mere four years later, this estimate decreased to 60,000 and Louisiana’s 2023 population estimate was 40,000 wild turkeys. Overall, despite multifaceted conservation management techniques employed by the LDWF, including captive breeding and release, relocation, reduced hunting season lengths, and population monitoring, Louisiana wild turkey populations have generally decreased since a peak in the mid-1990s (38; Fig. 1). Indeed, when restoration effort and population productivity were compared, Louisiana was the only of eight southeastern states with reduced wild turkey productivity (poults-per-hen) after restoration was 95% complete (44); however, the cause for the recent population declines is unknown.

We hypothesize that Leucocytozoon infection might have played a role in Louisiana turkey population declines because of an April – May 2010 outbreak of the black flies (Simulium meridionale), also known as turkey gnats, which caused at least 225 backyard poultry deaths throughout the state (45). Further, collections in nearby Mississippi from 2009 to 2018 identified Leucocytozoon in turkey gnats (46) and reports of backyard poultry deaths in this state have also been linked to turkey gnat incidence. The historical sensitivity of turkeys to Leucocytozoon infection and potential role of wild turkeys as a reservoir of Leucocytozoon also support the hypothesis that Leucocytozoon infections may be partly responsible for the decline of Eastern wild turkeys (M. g. silvestris) in Louisiana. To test this hypothesis, we extracted DNA from 106 wild turkey blood samples collected in Louisiana over two years and used PCR to evaluate the prevalence and species identities of Plasmodium, Haemoproteus, and Leucocytozoon. Our objectives were to evaluate the prevalence of specific Leucocytozoon species in wild turkeys in Louisiana and identify host traits (i.e., sex and location) that predict Leucocytozoon prevalence in wild turkeys.

Although our investigation focused on Leucocytozoon specifically, we also tested for the prevalence of Plasmodium and Haemoproteus. This allowed us to evaluate whether one Haemosporidian genus occurred more frequently than others and quantify the frequency of co-infections. In addition, we report a case study of an emaciated wild turkey heavily infected with Haemosporidian parasites that was presented to us during our investigation. The results of this study will inform wildlife management practices to conserve the declining wild turkey populations of Louisiana and further our understanding of wild turkeys as disease reservoirs.

Wild turkey blood samples were collected in 2019 from harvested turkeys (n = 8, all male) and in 2020 from turkeys (n = 98, 44 males and 54 females) trapped by Louisiana Department of Wildlife and Fisheries personnel using cannon nets (note that Wildlife and Fisheries Department personnel are exempt from requiring Scientific Collection Permits in Louisiana). Nobuto filter paper strips were dipped into the blood of hunter-harvested turkeys in 2019 and stored at 4°C (47) in individually labeled envelopes to avoid contamination. Blood samples were collected from captured turkeys by venipuncture of the medial metatarsal vein using a 22G hypodermic needle and 3 cc syringe. Approximately 0.1 ml of blood was deposited onto Nobuto filter paper strips for absorption. Netted turkeys were released within 10 min of capture. All filter-strip blood samples were stored at −20°C and protected from light until DNA extraction.

On November 17, 2020, a private landowner in Tangipahoa Parish, Louisiana, reported a lethargic wild turkey that was unable to fly. Personnel from the LDWF detected lice and severe emaciation. The animal was euthanized and the carcass shipped to the Southeastern Cooperative Wildlife Disease Study (SCWDS), at the College of Veterinary Medicine at the University of Georgia for postmortem analyses. Postmortem findings included serous atrophy of fat, esophageal/crop mucosal retention, and early urate stasis in kidneys, all indicative of decreased food and water intake that resulted in emaciation. SCWDS determined that debilitation of this turkey was from Leucocytozoon infection.

A blood smear was obtained from this turkey, stained with Wright-Giemsa (Harleco, Sigma Aldrich. Germany), and parasitemia as the number of infected red blood cells out of the 10,000 examined was quantified with an oil-immersive 100× objective lens using an Olympus TH4-100 light microscope. In addition, the first 100 leucocytes were classified as either lymphocytes, monocytes, heterophils, or eosinophils.

DNA was extracted from turkey blood blots using an ammonium acetate precipitation protocol (48). Nested PCR was conducted with the extracted DNA at a concentration of 20 ng/μl, according to previously established protocols (48,49,50,51). In the first round of PCR, primers Haem NF1 and Haem NR3 amplify parasite mitochondrial DNA found in all three genera (52); the second round of PCR targeted regions of cytochrome B specific to Haemoproteus and Plasmodium (primers HaemF and HaemR2) and Leucocytozoon (primers HaemFL and HaemR2L) (53). The first round of PCR reactions used 10-μl volumes with 5 μl DreamTaq (2x; ThermoFisher), 0.12 μl of 50 μM HaemNF1, 0.12 μl of 50 μM HaemNR3, 2.5 μl of 20 ng/μl (50 ng) template DNA, and 2.26 μl ddH₂O. The second round of PCR reactions included 0.06 μl of 50 μM forward and 0.06 μl 50 μM reverse primers for the respective parasites, 1 μl of round one amplicons as template DNA, and 3.88 μl ddH₂O. The PCR amplicons of round two (5 μl) were loaded onto a 1% agarose gel stained with RedSafe (iNtRON Biotechnology, Korea), run at 100 V for 90 min, and visualized under UV light. Each round of PCR included a positive control: DNA from a house sparrow (Passer domesticus) co-infected with Haemoproteus and Plasmodium for HaemF and HaemR2 primers, and DNA from a Leucocytozoon-infected turkey for HaemFL and HaemRL2 primers. We submitted the remaining 5 μl of PCR product for samples that tested positive for Haemosporidian infection (revealed by gel bands) to Louisiana State University’s (LSU) Genomics Core Facility for PCR cleanup and Sanger sequencing on a 3130xl Genetic Analyzer (Applied Biosystems). Co-infections with Haemoproteus and Plasmodium resulted in double bands that were incompatible with Sanger sequencing, so we did not obtain species-level resolution for co-infected samples. Sequences were queried against the National Center for Biotechnology Information (NCBI)’s Basic Local Alignment Search Tool (BLAST) for species-level resolution of the Haemosporidian parasites found in wild turkeys (54). None of the top matches for Leucocytozoon sequences were resolved to species, so we also consulted the MalAvi database for this genus (55).

χ-square tests of independence were used to compare the frequency of infection by each genus (Plasmodium, Haemoproteus, and Leucocytozoon). We first tested whether the frequency of infection by genus differed overall, and if so, which genera significantly differed (n = 4, including the latter post hoc χ-square tests). We next assessed whether the frequency of being uninfected, singly infected, co-infected, and tri-infected differed overall, and if so, which infection categories significantly differed from each other (n = 7, including post hoc χ-square tests). Finally, we examined whether the frequency of co-infection types (genus combinations) differed overall, and if so, which combinations significantly differed from one another (n = 4, including post hoc χ-square tests) for a total of 15 tests.

χ-square tests of independence were also used to compare the frequency of infection by species of each genus to test whether the prevalence of each species was more or less than expected by chance. We tested whether the prevalence of each species within respective genera differed overall (n = 3) and, if significant differences were detected, performed post hoc χ-square contrasts to determine which species significantly differed (n = 10), for a total of 13 tests. To control Type I error (false positives), we divided the conventional α by the total number of tests (i.e., 28) and considered P < 0.0018 as statistically significant (56).

We used three binary logistic regressions that differed by the presence (infected or not) of infection by Haemoproteus, Plasmodium, and Leucocytozoon as the dependent variable to test whether sex and location (independent variables) affected the likelihood of a wild turkey being infected by each genus. We did not test for a year effect because of limited sampling efforts in 2019 (n = 8; in 2020, n = 98). Next, we categorized the number of infections per individual as a number (i.e., 0, 1, 2, or 3) and included it as the dependent variable of a generalized linear model with a Poisson distribution for right-skewed data to test whether sex and location (independent variables) affected the likelihood of co-infection.

The frequency of Haemoproteus, Plasmodium, and Leucocytozoon significantly differed from one another (Fig. 2a–c; Table 1a). Post hoc χ-square tests revealed that Haemoproteus infections occur more frequently than Plasmodium and Leucocytozoon (Table 1a). The frequency of turkeys infected with either zero, one, two, and three infections significantly differed from one another (Table 1b, Fig. 2d), and post hoc χ-square tests indicated that only the frequency of uninfected and triply infected turkeys did not significantly differ from each other (Table 1b). The frequency of genera co-infection types significantly differed from one another overall (Table 1c). There were no Plasmodium-Leucocytozoon co-infections present in our turkey samples (Fig. 2e). Post hoc tests revealed that there were more Haemoproteus-Leucocytozoon co-infections than Plasmodium-Leucocytozoon and Plasmodium-Haemoproteus co-infections (Table 1c).

Haemoproteus infections comprised of five species: Haemoproteus sacharovi (62%), Haemoproteus plataleae (22%), Haemoproteus noctuae (13%), Haemoproteus enucleator (3%), and Haemoproteus syrnii (1%; Fig. 3a). Within Haemoproteus-infected samples, the different species’ infection frequencies significantly differed from one another (Table 2a). Post hoc comparisons revealed that the frequency of H. sacharovi infections was significantly higher than all other Haemoproteus species, and H. plataleae infections were significantly more frequent than H. enucleator and H. syrnii (Table 2a).

Plasmodium infections were comprised of two species, Plasmodium cuculus and Plasmodium elongatum that were equally prevalent (n = 3 each; Table 2b; Fig. 3b). Leucocytozoon infections were comprised of L. schoutedeni (74%), L. sabrezesi (22%), and one sample that could not be identified to species (4%; Fig. 3c). Within the subset of Leucocytozoon-infected samples, L. schoutedeni infections occurred significantly more frequently than L. sabrezesi infections (Table 2c). Queries for both databases (NCBI and MalAvi) were consistent: neither database’s top match for a sequence provided species-level resolution; however, the top match to a species for each database returned the same species (see Supplementary Material).

Haemoproteus infections were not predicted by sex or location (all P > 0.99), nor were Plasmodium infections (all P > 0.35; Table 3). Notably, all male turkey samples in this study tested positive for Haemoproteus infection, as did all but two females. Year was not a significant predictor of Leucocytozoon infections (P = 0.99), but sex was (z = 3.1, df = 105, P = 0.001); 50% of male turkeys were infected with Leucocytozoon compared to only 18% of females (Table 3). Neither sex nor location were significant predictors of co-infection (all P > 0.14; Fig. 4).

Sanger sequencing of the PCR product from this turkey’s blood revealed the specific species of Leucocytozoon to be L. schoutedeni, with 97% identity to GenBank KT290937.1, Gallus gallus, Malaysia (unpublished) and 96% identity to GenBank MW043728.1, G. gallus, Thailand (58). In addition to Leucocytozoon infection, SCWDS also detected lymphoproliferative disease virus, known to cause tumors in the skin and organs, and Parahaemoproteus/Haemoproteus infection, which we identified as H. sacharovi. We did not detect Plasmodium from this turkey. Blood smear quantification revealed 82 Haemoproteus-infected cells and 88 Leucocytozoon-infected cells, both indicative of acute infections (Fig. 5). This turkey’s blood also had a high proportion of eosinophils (77%), followed by 20% lymphocytes and 3% monocytes; 0%–2% is the typical range for eosinophils for a healthy bird (59).

Leucocytozoon infections were detected in 34% of blood samples from eastern wild turkeys of Louisiana; only two species were found: L. sabrazesi and L. schoutedeni. Leucocytozoon sabrazesi occurred less frequently than L. schoutedeni, and it was found exclusively in the Southeast Loblolly management region. We consider this a conservative estimate of Leucocytozoon prevalence, because our sampling may have overlooked individuals that died from infection before capture, underestimating the actual prevalence of infection. Our understanding of Leucocytozoon mortality rates in wild turkeys relies upon studies nearly a century old (22,26,27), which is enough time for host-parasite coevolution to have altered the high mortality rates that were previously observed in turkeys. Indeed, wild turkeys have historically not been infected by L. sabrazesi (28), although we detected this species in seven of our samples. The 2020 turkey case study may suggest high pathogenicity of Leucocytozoon infection in turkeys but the co-occurrence of H. sacharovi infection and lymphoproliferative disease virus limits our ability to conclude that symptoms were specific to Leucocytozoon infection. Further experimental studies will be needed to clarify the pathology of Leucocytozoon infections in different populations of wild turkeys.

Inoculation experiments would also help clarify why Leucocytozoon infections were found more frequently in male turkeys than in females; for example, whether males are more susceptible to Leucocytozoon infection because of sex differences in habitat use, or whether females have higher Leucocytozoon mortality rates that result in the observed lower frequency of infection. Eastern wild turkeys have mating systems of harem defense or male dominance polygyny, where males form dominance hierarchies that grant access to moving female groups, and females are the sole providers of parental care (60). This mating structure may differentially affect habitat preference (61) and movement (62) in males and females. For example, males are hypothesized to have larger home ranges in the spring to increase access to hens (63,64), and females decrease their home range size to areas of high food abundance during brood rearing (65). It is possible, then, that these sex differences in habitat use may translate to differential vector exposure and subsequent disease incidence. Alternatively, male turkeys are typically larger (66) and may emit more carbon dioxide than females, thereby increasing the “attractiveness” of males to black fly vectors (67). Although the higher incidence of Leucocytozoon infection in males than females may suggest that the former is a more effective disease reservoir, our investigation only presents one year of data with a robust sample size; observing this pattern over multiple years should test this hypothesis.

Historical records of L. schoutedeni (also called L. andrewsi), the most common Leucocytozoon species we found in Louisiana wild turkeys, are few in North America. The first recorded incidence dates to 1950 in chickens in South Carolina (18) and the species was not reported again until the new century (30), with examples in the literature from 2009 in poultry in Mississippi (6) and from 2018 to 2020 in Northern bobwhite quail (Colinus virginianus) in western Oklahoma (68). Thus, L. schoutedeni appears to have been uncommon in North America until recently and was perhaps introduced from its common range in Africa and southeast Asia (59,69,70,71,72,73,74). Leucocytozoon schoutedeni typically causes mild disease, including long-term chronic parasitaemia and reduced laying in chickens and turkeys (12,14,30).

Like L. schoutedeni, L. sabrazesi (also called L. macleani), the other Leucocytozoon species we found in Louisiana wild turkeys, typically results in mild pathology and is most frequently found in Africa and southeast Asia (12,14,74,75,76). Historically, turkeys were not infected by L. sabrazesi (28), and we were unable to find any previous record of it occurring in North America. However, it is important to note that although 45 Leucocytozoon morphospecies can be visually distinguished, only 13 can be characterized using molecular methods (12) and lineage detection may be biased by the primers used (77). For example, the fact that L. schoutedeni had the closest sequence similarity (96%) to the case study turkey does not unequivocally confirm this strain; indeed, L. schoutedeni is morphologically characterized by round host cells, but the blood smear showed an elongated parasite. Therefore, it is possible this was another species that has not been described, and which was not present in the NCBI database but had the highest genetic similarity to L. schoutedeni. Alternatively, a co-infection of two Leucocytozoon species may have interfered with sequencing. Thus, whether our investigation represents the first incidence of L. sabrazesi in North America and the first detection in wild turkeys as a host, or an entirely new species, will require additional surveys and genetic analyses.

Haemoproteus parasites had the highest frequency of detection (92%) out of all genera and the highest species diversity. Only one of the five Haemoproteus species, H. sacharovi, occurred more frequently than expected by chance. Our observation of Haemoproteus infections exceeding the prevalence of Plasmodium and Leucocytozoon is consistent with some historical (21,78) and recent (79,80) observations in wild turkeys of North America, but others have noted a higher prevalence of Leucocytozoon (17,34). Variable prevalence of different Haemosporidian species from year to year is expected, because abiotic factors, including precipitation, humidity, and temperature, influence the life cycle and feeding activity of insect vectors as well as parasite development times (81,82,83,84,85,86,87). Forrester et al. (78) and Annetti et al. (79) quantified infection prevalence over successive years, but only the former detailed annual variation, and found that Haemoproteus exceeded Leucocytozoon prevalence in three of the four years investigated. Haemoproteus has minimal pathogenic effects (88) and low observed circulating parasitaemia in turkeys (21), though turkeys have high susceptibility to infection from vectors (89). Combined with the high prevalence we detected in Louisiana wild turkeys, this suggests that turkeys are tolerant of Haemoproteus infection.

In addition to single infections, we also quantified Leucocytozoon, Plasmodium, and Haemoproteus co-infection patterns to consider whether co-infection may pose a threat to Louisiana wild turkeys. Unexpectedly, we detected over four times as many samples infected with all three genera than samples free from any Haemosporidian infection. This “tri-infection rate” was nearly double that of turkeys tested in Illinois (9.2% versus 5%; 79). Co-infections of Leucocytozoon-Haemoproteus were five times more frequent than Haemoproteus-Plasmodium co-infections, and we did not detect any co-infections of Leucocytozoon-Plasmodium. This may be because of differing ecological requirements of black flies and mosquitoes, the vectors of these two parasite species (90,91). Haemosporidian co-infections appear frequently in wild turkeys (18,78,79), and in birds generally (92,93). Co-infections with Haemoproteus and Plasmodium had additive negative impacts on body condition, but not reproductive parameters, in free-living purple martins (Progne subis) (94). Likewise, experimental inoculations with two different species of Plasmodium did not increase infection virulence in multiple host species (95,96,97). The high prevalence of co-infection we detected in Louisiana wild turkeys suggests that turkeys can survive co-infection, but whether co-infection affects metrics such as reproductive output or lifespan requires additional study, especially considering that Leucocytozoon infection is known to reduce fecundity in domestic turkey hens (25).

In our investigation of Louisiana wild turkey Haemosporidian parasites, we found that Haemoproteus had the highest infection prevalence and species diversity; male turkeys were more likely to be infected by Leucocytozoon parasites than females; and the number of triply infected turkeys exceeded that of uninfected turkeys. Studies of Haemosporidian parasites in wild turkeys were abundant when turkeys were declining in North America, but publications tapered off after turkey populations rebounded. This, combined with the fact that Leucocytozoon studies in wild birds are generally limited (12), means there are still many open questions about the role of Leucocytozoon species in the decline of wild bird populations. Experimental inoculations with Leucocytozoon parasites in wild turkeys that describe pathology would help elucidate whether Leucocytozoon infection contributes to population decline in Louisiana through mortality or negative impacts upon reproductive output, and whether one sex is more tolerant to Leucocytozoon infection than the other.

Supplemental data associated with this article can be found at https://doi.org/10.1637/aviandiseases-D-25-00022.s1.

We thank Louisiana Department of Wildlife and Fisheries Turkey Program Manager Cody Cedotal, the wildlife biologists and technicians, as well as SCWDS for the diagnostics report of the emaciated turkey. We thank Danylo Zaitsev and Aidan Matthews for providing laboratory assistance and Scott Herke of Louisiana State University’s Genomics Facility for logistical support. LDWF personnel conducted the trapping and blood collection from wild turkeys. They are exempt from requiring a Scientific Collecting Permit. The Pennington Regents Chair for Wildlife Research Professorship funded L. F., T. R. K. received financial support as a Merck Awardee of The Life Sciences Research Foundation, C. R. L. received start-up funds from LSU, and K. I. L. from the LSU President’s Future Leaders in Research program.

LDWF =

Louisiana Department of Wildlife and Fisheries;

LSU =

Louisiana State University;

SCWDS =

Southeastern Cooperative Wildlife Disease Study