Identification of Blood Parasites in Individuals from Six Families of Freshwater Turtles
Abstract
Freshwater turtles from 28 species representing 6 families, categorized as “wild,” “captive,” or “wild-caught, captive-raised,” were analyzed for the presence of intraerythrocytic, parasitic protozoa. Microscopic analyses of blood smears from 327 individual turtles revealed parasites in the blood of 29% of these individuals (n = 96), with levels of parasitemia ranging from 0.003% to 2.1%. SYBR ® Green–based quantitative polymerase chain reaction confirmed these results with 29% overall prevalence (94 of 324), with 66% prevalence in the Geoemydidae, 26% prevalence in the Emydidae, and 23% prevalence in the Kinosternidae. No infections were detected in members of the Chelidae, Pelomedusidae, and Trionychidae. Prevalence was highest in wild turtles with 67%–100% from 7 locations, followed by wild-caught, captive-raised turtles with 4%–26% from 2 locations, while detections in captive turtles were zero in 2 locations and 3% in the remaining third location. Comparative sequence analyses of 583-bp amplicons of 18S rRNA gene fragments allowed us to identify 80 Haemogregarina infections and one Hemolivia infection. Parasites representing the genus Hepatozoon were not detected. Because parasites were generally detected in wild turtles and only rarely in captive turtles, captive breeding with subsequent release of turtles would seem to pose little risk to free-ranging turtles with regard to the spread of these parasites. On the contrary, wild turtles would be more likely sources of parasite transmission to captive populations if used in breeding programs.
A large variety of environmental stressors, including habitat destruction and fragmentation by urban and agricultural growth or related fertilizer and chemical run off, have been identified as underlying reasons for steep declines of turtle populations, with effects on both turtle abundance and diversity (Gibbons et al. 2000; Dresser et al. 2017). Harvest and trade of turtles for food or herbal remedies are other causes adding to population declines (Gibbons et al. 2000; Mali et al. 2014; Harju et al. 2018). Conservation efforts for endangered turtle populations often include removal of individuals from at-risk areas or confiscation at borders or trade markets, captive propagation, and the subsequent release of offspring into the environment (Miorando et al. 2013; Hartley and Sainsbury 2017). Capture, transport, and high-density propagation are stressful conditions for turtles, generally increasing susceptibility to diseases and parasites but also promoting their transmission (Bower et al. 2019). Turtles are extremely sensitive to diseases and are affected by more diseases than is any other reptilian group (Dodd 2016). Transmission of diseases from introduced hosts to indigenous animals could result in more severe infections of naïve animals potentially lacking an acquired immunity for protection (Hartley and Sainsbury 2017).
Diseases are often transmitted through vectors such as leeches, the most common ectoparasites in reptiles (McCoy et al. 2007; Readel et al. 2008). Leeches have been identified as potent vectors, especially for the transmission of blood parasites that require both an invertebrate and a vertebrate host (Davies and Johnston 2000; Siddall and Desser 2001; Dvořáková et al. 2014). In freshwater turtles, the most common of these parasites belong to the Haemogregarinidae (Readel et al. 2008) and more specifically to 3 genera, Haemogregarina, Hemolivia, and Hepatozoon (Cook et al. 2015; Arizza et al. 2016). Infections with these intraerythrocytic, parasitic protozoa are thought to have little or no effect on their natural host; however, they may result in anemia, weakness, and generally poor health conditions of naïve turtle hosts, along with limited motility and decreased growth rate (Maia et al. 2014; Özvegy et al. 2015). Thus, the potential release of infected offspring back into at-risk areas could be harmful to conservation efforts (Verneau et al. 2011).
The goal of the present study was to assess the potential threat that captivity or captive breeding of turtles might pose for the transmission of blood parasites from captive to free-ranging turtles. For this purpose, we analyzed more than 300 freshwater turtles from 28 species representing 6 families for the presence of blood parasites of the Haemogregarinidae. These turtles were categorized as “wild,” “captive,” or “wild-caught, captive-raised.” Prevalence and parasitemia were initially analyzed by microscopy, with prevalence confirmed by quantitative polymerase chain reaction (qPCR). Comparative sequence analyses of amplified 18S rRNA gene fragments were then used for genetic identification of the parasites.
METHODS
Blood Sample Collection and Preservation
Adult and juvenile turtles were collected using basking traps and hoop nets or collected by hand from ponds (Institutional Animal Care and Use Committee [IACUC] permit 7E1EC3_02, TPWD SPR-0102-191). Captive turtles, i.e., turtles born and kept in captivity (n = 160), were obtained from both indoor and outdoor facilities from Concordia Turtle Farm (CTF; Wildsville, Louisiana; 31°37′10.9200″N, –091°47′04.0920″W), Guthrie Turtle Farm (GTF; Birmingham, Alabama; 32°49′07.8240″N, –086°12′30.7080″W), and Waterlife (W; Austin, Texas; 29°47′30.8760″N, –097°56′02.5440″W). These locations also provided wild turtles, i.e., turtles free-ranging or in captivity for less than a year. Wild turtles (n = 123; Table 1) were also caught at Cypress Point (CP; San Marcos, Texas; 29°53′29.4720″N, –097°55′59.3400″W), a private ranch (D; Deanville, Texas; 30°25′41.1240″N, –096°41′41.7480″W), Oasis Ranch (OR; Sheffield, Texas; 30°28′07.3920″N, –101°47′57.9480″W), Capital Aggregate (CA; Marble Falls, Texas; 30°41′14.1720″N, –098°14′44.5560″W), and Griffith League Ranch (GLR; Bastrop, Texas; 30°11′41.0280″N, –097°14′36.1320″W). Turtles held in captivity for more than a year were obtained from GTF and W and classified as wild-caught, captive-raised turtles (n = 49). Turtle families sampled in this study included the Geoemydidae, Chelidae, Emydidae, Kinosternidae, Pelomedusidae, and Trionychidae, with a total of 28 species (Table 1).
All turtles were marked, weighed, and sexed, with body measurements and the presence of leeches recorded. Blood samples were obtained from the femoral artery or vein of 332 individuals using a 20-ga needle. For histological analyses, blood samples were directly dropped onto slides while those for molecular analyses (approximately 700 µl) were mixed with 700 µl of blood storage buffer (100 mM Tris-HCl, 100 mM Na2EDTA, 30 mM SDS, pH 8.0) in 2-ml Nunc® cryotubes. These blood vouchers were stored at –80°C in the Forstner tissue collection (MF5454 to MF7949, not consecutively enumerated) held at Texas State University.
Microscopic Analyses of Blood Samples
Thin blood smears were made in triplicate for 327 of 332 individuals following the Centers for Disease Control and Prevention (CDC) protocol (Diagnostic Procedures for Blood Specimens; http://www.dpd.cdc.gov). Blood smears on slides were fixed in absolute methanol for 1 min, stained with Wright Stain (Carolina Biological Supply Company, Burlington, NC) for 3 min, and placed in Wright Buffer (Carolina Biological Supply Company) for 6 min (Clark 1981). Blood smears were subsequently examined for approximately 10,000 cells for all 3 slides per individual. The level of parasitemia was determined by dividing the number of infected cells per number of total cells.
Molecular Analyses of Blood Samples
SYBR® Green–based qPCR using 18S rRNA gene targeting primers JM4F (5′-ACT CAC CAG GTC CAG ACA TAG A-3′) and JM5R (5′-CTC AAA CTT CCT TGC GTT AGA C-3′) (Maia et al. 2014) was used to test for presence–absence of hemogregarine infections in DNA extracts from 10 µl of blood voucher samples (Qiagen Blood and Tissue Kit; Qiagen, Hilden, Germany) of 326 individuals. qPCR reactions were performed in 10-µl volumes containing 5 µl of SsoADV SYBR Green mix (BioRad, Hercules, CA), 0.2 µl each of 10 µM JM4F and JM5R primers, 3.6 µl of nuclease-free H2O, and 1 µl of DNA template in an Eco Real-time PCR system (Illumina, San Diego, CA). Initial incubations at 50°C for 2 min and 95°C for 10 min were followed by 40 cycles of 95°C for 10 sec, 64°C for 30 sec, and 72°C for 20 sec. The amplification of 215-bp fragments was followed by a melting curve analysis.
DNA from positive samples was then used in endpoint PCRs performed with primers HepF300 (5′-GTT TCT GAC CTA TCA GCT TTC GAC G-3′) and HepR900 (5′-CAA ATC TAA GAA TTT CAC CTC TGA C-3′; Ujvari et al. 2004) to amplify a longer fragment of the 18S rRNA gene (633 bp). PCR was performed in a final volume of 25 µl containing 3 µl of sample DNA, 0.5 µl of 10 mM dNTPs, 2.5 µl of buffer, 18.875 µl of nuclease-free H2O, 0.5 µl of each primer (10 µM each), and 0.125 µl of Taq polymerase. PCR conditions consisted of an initial incubation at 95°C for 10 min, followed by 35 cycles at 95°C for 10 sec, 50°C for 30 sec, and 72°C for 45 sec, and a final incubation at 72°C for 7 min. PCR products were analyzed by gel electrophoresis (2% agarose in TAE buffer) and visualized using a BioRad GelDoc EZ Imager.
Amplicons were cleaned using Shrimp Alkaline Phosphatase and Exonuclease I (Affymetrix, Santa Clara, CA) following the manufacturer's protocols and then sequenced bidirectionally using BigDye™ Terminator v3.1 (Applied Biosystems, Foster City, CA) with the same primers used for PCR. Products were purified by Sephadex G-50 gel filtration before sample plates were loaded onto an ABI-3500 Sequencer (Applied Biosystems). Amplification products that could not be sequenced successfully by direct sequencing attempts were cloned using the TOPO TA Cloning Kit for Sequencing (Invitrogen, Carlsbad, CA) following the manufacture's protocols. Resulting sequences were assembled in Geneious 9.1.4 (https://www.geneious.com) and deposited at GenBank under accession numbers MN160404 to MN160484.
Resulting bidirectionally verified sequence contigs were then assembled in Geneious 9.1.4 with primer sequences trimmed and 583-bp sequences compared against sequences in the GenBank/European Molecular Biology Laboratory (EMBL) databases using the Basic Local Alignment Search Tool (BLAST) algorithm (Pearson and Lipman 1988). Representative sequences from confirmed hemogregarines were added from GenBank/EMBL databases and aligned to our sequences using the Geneious alignment tool. The identity and relationship among the amplified sequences were evaluated using neighbor joining (NJ; Saitou and Nei 1987) and maximum likelihood (ML; Stamatakis 2006) analyses using Geneious 9.1.4. The NJ analyses utilized the HKY85 model to correct for substitution bias (Hasegawa et al. 1985). Model parameters for ML, which were estimated by the general time reversible (GTR) model with gamma (Tavaré 1986), were used as input in an ML heuristic search using PhyML (Guindon et al. 2010). Bootstrap values (Felsenstein 1985) were estimated from a heuristic search with a random stepwise addition sequence for 10,000 iterations for both NJ and ML analyses.
RESULTS
Microscopic analyses of blood smears from 327 individual turtles, representing 6 families and 28 species, detected parasites in the blood of 96 individuals (Fig. 1). Detections included 6 of 9 species of the Geoemydidae (prevalence 66%, n = 44 individuals), 5 of 7 species of the Emydidae (prevalence 27%, n = 235 individuals), and 2 of 3 species of the Kinosternidae (prevalence 25%, n = 12 individuals), but only in 0 of 7 species of the Chelidae (32 individuals), 1 species of the Pelomedusidae (3 individuals), and 1 species of the Trionychidae (1 individual; Table 1). Detection of parasites within families and species was highly variable, with entire species without parasite detections (e.g., Callagur borneoensis, all 7 individuals from W), while other species were found to be generally carrying parasites (e.g., Heosemys grandis, 20 of 22 individuals from W; Table 1). Levels of parasitemia in infected individuals were highly variable as well, ranging from 0.003% to 2.1% in the Geoemydidae, 0.01% to 2% in Emydidae, and 0.01% to 0.6% in the Kinosternidae (Fig. 2).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 19, 1; 10.2744/CCB-1411.1
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 19, 1; 10.2744/CCB-1411.1
Comparative qPCR analyses largely corroborated the histological results, except for 12 cases (Pseudemys texana and Pseudemys gorzugi) where qPCR failed to detect parasites identified by microscopy (Table 1) and another 8 cases (Siebenrockiella crassicollis and Trachemys scripta elegans) where qPCR detected parasites while microscopy did not (Table 1). Overall, qPCR detected parasites in 94 of 324 individuals, with parasites detected in 66% (31 of 47 individuals) of the Geoemydidae, 26% (58 of 230 individuals) of the Emydidae, and 23% (3 of 13 individuals) of the Kinosternidae (Table 1), with an average prevalence of 29% at all sites. Prevalence was lowest at CTF with 3% (4 of 135 individuals) followed by OR with 12% (2 of 17 individuals), W with 27% (19 of 69 individuals), and GTF with 29% (8 of 28 individuals; Fig. 3). All other locations had prevalence values greater than 67%, with 100% prevalence at GLR (1 individual) and CA (5 individuals), 67% (32 of 48 individuals) at CP, and 74% (17 of 23 individuals) at D (Fig. 3). From these locations as well as from OR, only wild turtles were available and analyzed while turtles from CTF were all captive animals. The only locations from which captive, wild, and wild-caught, captive-raised turtles were analyzed were GTF and W. Analyses of prevalence showed steep gradients, with the highest prevalence of parasites in wild turtles (70% and 72%, respectively) followed by wild-caught, captive-raised turtles (26% and 4%, respectively) and no infections of captive turtles from either location (Fig. 4).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 19, 1; 10.2744/CCB-1411.1
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 19, 1; 10.2744/CCB-1411.1
Of the 94 samples with parasites, a total of 81 amplicons of 18S rRNA gene fragments were obtained and sequenced. Comparative sequence analyses of these 583-bp fragments with those of reference sequences resolved clades for Hepatozoon, Hemolivia, and Haemogregarina with high bootstrap support for both NJ and ML analyses (Fig. 5). None of our sequences clustered to sequences of Hepatozoon reference organisms, and only one was identified as Hemolivia sp. (Fig. 5). This Hemolivia sp. was detected in a wild-caught, captive-raised Rhinoclemmys punctularia from GTF, with an identical sequence to Hemolivia sp. ex Rhinoclemmys pulcherrima (KF992714; Fig. 5). All remaining 80 sequences were identified as Haemogregarina sp. (Fig. 5). Of these, 79 sequences were very similar (99%–100%), with some identical to Haemogregarina sp. STcl6 ex Macrochelys temminckii (KX507248) or similar to Haemogregarina sp. STcl8 ex Macrochelys temminckii (KX507250) (Fig. 5). The remaining sequence from a wild Graptemys versa from CA had high similarity (99.7%) to sequences in a well-supported clade with Haemogregarina balli (HQ224959) and Haemogregarina stepanowi (KF257928) while similarity values to remaining Haemogregarina sequences were below 97% (data not shown). Sequences were obtained mainly from parasites in wild turtles (n = 73; 90% of the sequences) and in wild-caught, captive-raised animals (n = 6; 8%) while only 2% (n = 2) of the parasite detections occurred in captive turtles.
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 19, 1; 10.2744/CCB-1411.1
DISCUSSION
Hemogregarine infections have been detected in members of the Geoemydidae (Rossow et al. 2013; Dvořáková et al. 2015), the Emydidae (Davis and Sterrett 2011; Dvořáková et al. 2014; Özvegy et al. 2015; Arizza et al. 2016), the Kinosternidae (Davis and Sterrett 2011; Rossow et al. 2013), the Chelidae (Scheelings and Rafferty 2012; Goes et al. 2018), and the Pelomedusidae (Soares et al. 2014; Picelli et al. 2015; de Oliveira et al. 2018). Detections of hemogregarine parasites in turtle species from these families are generally documented for wild turtle populations while many of our analyses included captive and wild-caught, captive-raised populations. We likely failed to detect haemogregarine infections in the Chelidae and Pelomedusidae because wild turtles were not available from these families. Only captive (n = 5) and wild-caught, captive-raised (n = 27) individuals were sampled for the Chelidae and only wild-caught, captive-raised (n = 3) individuals were sampled for the Pelomedusidae. This speculation is supported by our overall results showing hemogregarine infections were more commonly found in wild turtles, with very few detections in captive and wild-caught, captive-raised individuals.
In our study, prevalence values of more than 67% and parasitemia levels of up to 2% for wild populations reflect those published for wild populations worldwide, with prevalence values often higher than 50% and up to 100% and parasitemia levels between 0.01% and 3% (Mihalca et al. 2008; Davis and Sterrett 2011; Rossow et al. 2013; Picelli et al. 2015; de Oliveira et al. 2018). Prevalence values in wild turtles by far exceeded those in captive (3%) and wild-caught, captive-raised (14%) individuals. These individuals were all obtained from turtle farms (CTF, GTF) or private facilities (W) seeking to assist with conservation efforts. At these locations, animals are kept and raised in indoor and outdoor facilities under controlled environmental conditions that include water treatments aimed at reducing the presence of leeches that have been shown to be potent vectors for the transmission of hemogregarines between turtle species (Siddall and Desser 2001). Placobdella parasitica and other leech species have been detected in wild populations of different turtle species including Chrysemys picta, Chelydra serpentina, Trachemys scripta scripta, Sternotherus odoratus, Graptemys geographica, and Kinosternum subrubum, among others, with high prevalence (i.e., on 20%–99% of all individuals) and with high abundance (i.e., with 1 to more than 100 leeches per individual; Ryan and Lambert 2005; McCoy et al. 2007; Readel et al. 2008; Davy et al. 2009). A reduction or elimination of the vector for hemogregarine parasites could therefore reduce the potential for parasite transmission and thus maintain turtle populations with no or low parasite prevalence.
Comparative sequence analyses of 18S rRNA gene fragments are commonly used to identify hemogregarine parasites in blood samples from freshwater turtles, with studies generally retrieving sequences that represent Haemogregarina sp. (Dvořáková et al. 2014, 2015; Özvegy et al. 2015; Alhaboubi et al. 2017; de Oliveira et al. 2018; Ungari et al. 2018). Sequences representing Hemolivia spp. have been obtained only from terrestrial animals such as the tortoise Kinixys zombensis (Cook et al. 2015) or the wood turtle Rhinoclemmys pulcherrima manni (Kvičerová et al. 2014) while Hepatozoon spp. have been detected in Sternotherus odoratus (Davis and Sterrett 2011), Kinosternon scorpioides (Soares et al. 2017), and Mauremys leprosa (Marzal et al. 2017), which are aquatic species. Primers HepF300 and HepR900 have been developed and used frequently for the detection of Hepatozoon spp. (Ujvari et al. 2004; Vilcins et al. 2009; Cook et al. 2015) but also perfectly match sequences from Hemolivia spp. The forward primer HepF300 shows 2 bp mismatches to sequences of Haemogregarina spp. However, they are located at the 5′-end and thus should not affect PCR amplification. Our data confirm this, as 80 of 81 sequences represented Haemogregarina spp. and the remaining sequence represented a Hemolivia sp.
We detected Haemogregarina spp. in predominantly aquatic turtle species while the Hemolivia sp. was found in one terrestrial wood turtle (Rhinoclemmys pulcherrima). These data are congruent with published data (Telford et al. 2009; Cook et al. 2015) and support the assumption that leeches are transmission vectors for Haemogregarina spp. in aquatic turtle species while ticks are transmission vectors for Hemolivia spp. in terrestrial turtle species (Široký et al. 2009; Telford et al. 2009; Rossow et al. 2013; Alhaboubi et al. 2017). Sequences representing Haemogregarina spp. resolved into 3 clusters; however, only the one in which 2 sequences from Graptemys versa and Siebenrockiella crassicollis clustered with those from species H. balli, H. stepanowi, H. sacaliae, and H. pellegrini was supported by strong bootstrap values (Fig. 5). The remaining 2 clusters had weak bootstrap support, which separated sequences from American and Asian turtle species, except for one wild individual of S. crassicollis. Haemogregarina spp. from American turtle species were closely related to those obtained from an alligator snapping turtle (Macrochelys temminckii) from northern Texas (USA; Alhaboubi et al. 2017) while those from Asian turtles did not cluster with any published sequences in GenBank. All Asian turtles were obtained from 2 locations, W (Austin, Texas), with 20 sequences of wild individuals in this cluster, and GTF (Birmingham, Alabama), with 4 sequences of wild (n = 1) and wild-caught, captive-raised (n = 3) individuals in this cluster (Fig. 5). The American turtles were from 6 different locations, with 3 turtles, including 2 wild-caught, captive-raised Kinosternon sonoriense and the S. crassicollis, from GTF. No American turtles were analyzed from W. It is tempting to speculate about coevolution of Haemogregarina spp. and Asian and American turtles; however, our data set is too biased and limited to provide more-definitive statements. In addition to sampling bias, e.g., sampling or keeping Asian turtles from different locations as the American turtles, or the potential for parasite transmission in captivity or during transport, our phylogenetic analyses were also based on comparative analyses of relatively short sequences (583 bp) from generally conserved molecules (18S rRNA genes). However, this fact is mainly a consequence of limited reference data for parasites in general and highlights the critical need for targeted studies that provide more genetic tools for the study of parasitic protozoa. Small sequence differences in multiple gene copies, as indicated for Haemogregarina spp. from Macrochelys temminckii (Alhaboubi et al. 2017), and also amplification bias or sequencing errors might therefore be one basis in our cluster assignments. Confirmation of these Haemogregarina clusters requires additional studies with sequences from gene fragments with higher phylogenetic information content, which will first require the development of suitable primers for this taxonomic group.
Our study demonstrates significant infection of freshwater turtles with Haemogregarina spp., but not with Hemolivia spp. or Hepatozoon spp., and that infections are generally found in wild animals, much less in wild-caught, captive-raised animals, and only rarely in captive animals. Given the critically endangered status of many turtle taxa across the planet, captive propagation and reintroduction is nearly certain to be required to enable recovery of wild populations or reestablish locally extirpated regions. Our results would indicate that the relative risk of transmissions of haemoprotozoans from captive-produced and captive-raised turtles introduced to wild turtle populations is quite low (0.1%) where leech vector management in captivity is comprehensive. On the contrary, wild turtles are more likely sources (33%) of parasite transmission to captive populations if individuals are removed from at-risk areas, or they are confiscated at borders or trade markets, and used in breeding programs. However, standard quarantine procedures for new animals into assurance colonies would likely mitigate many of those risks through leech and tick control (vector control) in the period of isolation. Our results suggest the potential transfer of parasites seemingly dominant in Asia to captive American taxa where the captive colonies contain wild-caught Asian and captive-raised American turtles. However, additional development of more taxonomically targeted molecular markers will help to resolve infections of specific parasites. The overall results strongly support that current management practices (e.g., vector control) for captive colonies significantly reduce or eliminate these parasites in captive animals.
Microscopic detection of parasites in erythrocytes of freshwater turtles from Cypress Point (CP), San Marcos, Texas. (A) Immature gamonts in erythrocytes (Pseudemys texana, collection #MF5810); (B) multiple immature gamonts or merozoites (Pseudemys texana, collection #MF5798); (C) mature gamonts (Trachemys scripta elegans, collection #MF5807).
Mean parasitemia levels in individual turtles from the Geoemydidae, Emydidae, and Kinosternidae with standard errors.
Prevalence of parasites in turtles from locations including Concordia Turtle Farm (CTF), Guthrie Turtle Farm (GTF), Waterlife (W), Capital Aggregate (CA), Cypress Point (CP), Deanville (D), Griffith League Ranch (GLR), and Oasis Ranch (OR). Dark bars represent percentages of positive detections while light bars represent percentages with negative detections, based on qPCR analyses. Numbers in bars represent numbers of individuals.
Prevalence of parasites in turtles from locations including the Guthrie Turtle Farm and Waterlife. Dark bars represent percentages of positive detections while light bars represent percentages with negative detections, based on qPCR analyses. Samples were obtained from captive turtles, i.e., turtles born in captivity; wild turtles, i.e., free-ranging turtles or those in captivity for less than a year; and wild-caught, captive-raised turtles, i.e., turtles held in captivity for more than a year. Numbers in bars represent numbers of individuals.
Maximum likelihood (ML) tree (general time reversible [GTR] model in PhyML, Geneious 9.1.4) showing the phylogenetic position of 81 parasites detected in blood samples of 15 turtle species belonging to the Geoemydidae, Emydidae, and Kinosternidae, with MF collection numbers based on comparative sequence analysis of 18S rRNA gene fragments with those of reference organisms. Bootstrap support from neighbor-joining (NJ) analysis is provided in front of the ML bootstrap value for that node. Sampling locations for turtles are color-coded and include CTF (green), GTF (red), W (dark brown), CA (blue), CP (yellow), D (brown), and GLR (gray); see Figure 3 caption for definitions. Color codes highlight the separation of turtles into categories: captive turtles (C), i.e., turtles born in captivity; wild turtles (W), i.e., free-ranging turtles or those in captivity for less than a year; and wild-caught, captive-raised turtles (WCCR), i.e., turtles held in captivity for more than a year.
Contributor Notes
Handling Editor: Peter V. Lindeman
