Baseline haematological parameters in three common Australian frog species

Field collection

Frogs were collected through hand-capture during nocturnal surveys (May 9^th–13^th, 2022) in the Strathbogie Ranges, Victoria, Australia (36°48′15″S 145°46′03E) (n = 126, see Table 1). Frogs were identified to be present at the site level by call, and once captured were identified to the species level by visual examination. Determination of sex was made using physical features such as nuptial pads and throat colour to identify adult males. Juveniles were identified by their small size and lack of male secondary sexual characteristics. Animals were weighed using spring scales (±0.1 g), and snout-to-vent length was measured using dial callipers (±0.1mm). Any ectoparasitic fly larvae that were visible were removed and preserved in 100% methanol and sent to our collaborators (J Keem at University of Melbourne, D De Angelis at La Trobe University, N Johnston at University of Technology Sydney, and K Newman at University of Melbourne) who are compiling instances of ectoparasitic fly larvae found to be causing myiasis in frogs in Australia.

All frogs were swabbed for B. dendrobatidis following standard methods (Boyle et al., 2004, as stated below). Blood (approximately 5 μL per frog) was then collected from the facial (maxillary) vein of each frog according to Forzán et al. (2012), and a blood smear was made. Frogs were then released at their original collection site on the same night.

Batrachochytrium dendrobatidis swabbing and qPCR

A dry sterile rayon swab (MW-113, Medical Wire and Equipment, Wiltshire, UK) was rolled five times along each of: the ventrolateral surface of the abdomen, the ventral surface of all limbs and digits, and the medial surface of the thighs. The tip of the swab was then broken off into a 1.5 mL Eppendorf tube and stored at −20 °C within 2 h of sample collection. The samples remained in −20 °C until processing (approximately 3 months). Skin swab samples were extracted using PrepMan Ultra (Thermo Fisher Scientific, Waltham, MA, USA). The extraction method followed the manufacturer’s directions which included: adding 50 μL of PrepMan Ultra and 30–40 mg of 0.5 mm silica beads (BioSpec, Bartlesville, OK, USA) to each sample, then homogenizing the samples (using a cell homogenizer) for 2 min at 1,400 oscillations per sec, incubating samples at 95 °C to lyse the cells for 10 min, and collecting and diluting the supernatant 6:100 in ultra pure water before directly analysing for pathogen presence and quantity using qPCR (Brannelly et al., 2020). The remaining extracted DNA was stored at −20 °C. With every extraction performed one B. dendrobatidis positive control sample (zoospores from culture) and one negative control (swab only) was extracted.

We used qPCR (Rotor-Gene Q, Qiagen, Hilden, Germany) to amplify and quantify the B. dendrobatidis DNA in each sample following (Boyle et al., 2004). We used a 15 μL reaction volume, with 5 μL of template DNA, 7.5 μL of Lo-ROX 2x master mix (SensiFAST, Bioline, London, UK), and a final reaction concentration of 900 nM ITS1–3 (Purification: RP-Cartridge Gold, Sequence: CCTTGATATAATACAGTGTGCCATATGTC, Eurogentec, Integrated Science, Liège, Belgium), 900 nM 5.8S Chytr (Purification: RP-Cartridge Gold, Sequence: AGCCAAGAGATCCGTTGTCAAA, Eurogenetec, Integrated Science, Liège, Belgium), 250 nM Chytr MGB2 5′ 6-FAM-labelled probe (5′ 6-FAM - CGA-GTC-GAA-CAA-AAT - MGB-Eclipse^® 3′; Integrated Science, Liège, Belgium) and 400 ng/μL bovine serum albumin (Thermo Fisher Scientific, Waltham, MA, USA). The amplification conditions were initial denaturation conditions of 2 min at 50 °C, 10 min at 95 °C, followed by 40 cycles of denaturation at 95 °C for 15 s and annealing at 60 °C for 1 min. On each qPCR reaction plate we included a series of seven plasmid-based B. dendrobatidis standards (purchased from Pisces Molecular, Bolder, Colorado, US containing 4.2, 42, 420, 4,200, 42,000, 420,000, and 4,200,000 DNA copies per reaction). On each qPCR reaction plate we also included a no template control, where ultrapure water replaced the template DNA. The extraction positive and negative control samples were analysed via qPCR like all sample-extracted DNA.

DNA copies of the B. dendrobatidis ITS gene extracted from each swab were estimated by extrapolating the gene copies detected by qPCR (Rotor-Gene Q 2.3.5 software), considering the elution volume and dilution of template DNA. Each sample was determined to be B. dendrobatidis positive if its reaction well was amplified to at least two ITS DNA copies detected. If the sample was considered negative for B. dendrobatidis, the B. dendrobatidis DNA copies was coded as 0 for that sample (Brannelly et al., 2020). Samples were analysed via qPCR in singlicate, which is a common and accepted practice for chytrid infection studies (Brannelly et al., 2015, 2017). All positive extraction controls were positive, all negative extraction controls were negative, and all no template qPCR controls were negative for B. dendrobatidis DNA copies.

Blood sampling and analysis

Blood samples were taken in accordance with Forzán’s protocol for collecting blood from the facial (maxillary) vein (Forzán et al., 2012). A 20G needle was used to pierce the skin and score the vein between the upper jawline and the rostral side of the tympanum (Forzán et al., 2012). The score released blood onto the skin surface that was then collected with a heparinised capillary tube. Blood from the capillary tube was transferred onto microscope slides, smeared, and left to air dry. Immediately after samples dried (within 2 h after collection), they were fixed with 100% methanol and stained using DiffQuik, air-dried, and then mounted on a coverslip using Permamount^™ mounting medium (Thermo Fisher Scientific, Waltham, MA, USA).

Blood samples were examined using light microscopy to a) visually identify the different cell types within the blood smears, b) perform a differential cell count of leukocytes and erythrocytes within the smear (i.e., the percentage of leukocytes and erythrocytes), c) perform a differential cell count of leukocytes (i.e., the percentage of the different types of leukocyte cells), and d) detect the presence of any haemoparasites using visual microscopic assessment. We performed these analyses on blood smears that were high quality: e.g., few broken cells, few blood clots, effective staining, high quality smears, large enough blood volume within the sample. Blood smears were examined in the monolayer at a total magnification of 1,000× (100× oil immersion objective lens).

While standard cytological analysis includes assessing erythrocyte concentrations using a hemocytometer (Forzán et al., 2016), we were unable to perform these analyses due to the limited volume of blood collected from each frog. Therefore, we used an adapted analysis and performed a differential count of leukocytes to erythrocytes within the smear, where all cell types were counted until at least 500 total cells were reached (including erythrocytes, thrombocytes, and leukocytes) to calculate their relative proportion within the sample. To perform a differential leukocyte count, we counted until a minimum of 100 leukocytes was reached (Chabot-Richards & George, 2015). The leukocyte differential count included the number of lymphocytes, monocytes, neutrophils, basophils, and eosinophils, which were used to calculate their relative proportion within the smear sample.

An animal was considered positive for parasites if there were macroparasites or the animals had infection with B. dendrobatidis based on the qPCR results.

Ethics and permits

The work was carried out under The University of Melbourne Animal Ethics application number 20062, which is conducted in compliance with the Australian code for the care and use of animals for scientific purposes and Wildlife Act 1975 Research Authorisation for the Victorian Department of Environment Land Water and Planning permit number 1000984. In accordance with our animal ethics application, we took all measures possible to follow the 3R tenets as described within the methods above. All fieldwork protocols conducted were compliant with The Australian Government Department of the Environment and Energy Threat Abatement Plan for the infection of amphibians with chytrid fungus resulting in chytridiomycosis (2016).

Statistical analyses

All analyses were conducted in R/RStudio using the packages ‘glmmTMB’ and ‘car’ (RStudio Team, 2020; R Core Team, 2022; Brooks et al., 2023; Fox et al., 2023). We used an alpha value of p = 0.05. Effect size was estimated using Cohen d’s statistic, and Tukey’s post hoc tests were conducted where appropriate using the package ‘emmeans’ (Lenth et al., 2022). Model assumptions were assessed to ensure no assumption was violated. For each analysis we conduced, we assessed the best fit model using the Akaike Information Criterion (AIC), a criterion used to compare statistical models to determine the best fit for data. We chose the model for each analysis conducted with the lowest AIC. If two models were <2 AIC units in difference, we chose the least complex model. For the statistical analyses that we performed, the possible predictor variables were species (Lit. paraewingi, C. signifera, and Lim. dumerilii), sex (male, female, and juvenile), size (mass, g), parasite infection status (1 = ectoparasites or B. dendrobatidis infection; 0 = no parasites observed) where appropriate, and two-way interactions among the predictor variables.

To compare differentials in the animals that had no signs of parasites (n = 79) we used generalised linear models (GLM) with a beta distribution. To assess the differential of erythrocyte to leukocyte cells, the ratio of leukocytes to erythrocytes was the response variable, and the predictor variables were species and sex. To compare the leukocyte differential counts, we conducted a series of GLMs where the proportions of the five different leukocyte cell types (lymphocyte, neutrophil, monocyte, basophil, and eosinophil) were the response variables, and species was the predictor variable.

To understand if there was an effect of parasites on the cell differentials, we investigated only the results from Lit. paraewingi and C. signifera because we did not find parasites in Lim. dumerilii (n = 70). We used generalised linear models (GLM) with a beta distribution to compare differentials across both species and sex in the animals that had no signs of parasites. To assess the differential of leukocyte to erythrocyte counts, the ratio of leukocytes to erythrocytes was the response variable, and the predictor variables were species, parasite status, and the interactive effect of species and parasite status. To compare the leukocyte differential counts, we conducted a series of GLMs where the proportions of three different leukocyte cell types (lymphocyte, neutrophil, monocyte) were the response variables, and species and parasite status were the predictor variables. We did not include basophils and eosinophils in this analysis because the counts of either type were too low for both Lit. paraewingi and C. signifera.

Parasites

Eight animals tested positive for B. dendrobatidis: five Lit. paraewingi (9.8% of the total collected) and two C. signifera (6.5%). The median B. dendrobatidis load for the infected Lit. paraewingi was 51,831 DNA copies and for C. signifera was 1,500 DNA copies. Six of the eight infected animals had corresponding blood smear slides examined.

Two Lit. paraewingi had visible fly larvae ectoparasites (Fig. 1A), one of which was severely anaemic with very few mature erythrocytes (Fig. 2). This animal was removed from the study because it was not considered clinically healthy. One C. signifera had a macroparasite found in the blood smear (Fig. 1B), which has not been identified. Using microscopic techniques, we found no confirmed haemoparasites in these blood samples.

In total, eight individual frogs were considered parasite positive (either had macroparasites (ectoparasitic fly larvae or unidentified macroparasite) or B. dendrobatidis fungal infection: five Lit. paraewingi (12% of the samples included in analysis; one female, one juvenile, three males) and three C. signifera (10%; three males). No individuals within this study had both B. dendrobatidis fungal infection and a coinfection with another parasite observed here. No parasites were detected in Lim. dumerilii.

Cell differentials of healthy animals

In this study we found that the median proportional ratio of cells within the blood smear was 0.946 erythrocyte cells: 0.034 leukocyte cells: 0.020 thrombocyte cells (Figs. 3 and 4). All three species had equivalent proportions of leukocyte to erythrocyte cells within their blood smears, and the proportions were equivalent in males, females, and juveniles (GLM: species, χ²₂ = 0.784, p = 0.676; sex, χ²₂ = 4.167, p = 0.125).

We found an effect of species on leukocyte proportions within the blood, indicating that the proportions of different leukocyte cells varied widely across species. Species was a significant predictor for four of the five leukocyte cell types (GLM for lymphocyte proportion: χ²₂ = 35.343, p < 0.001; GLM for monocyte proportion: χ²₂ = 11.912, p = 0.002; GLM for neutrophil proportion: χ²₂ = 45.914, p < 0.001; GLM for eosinophil proportion: χ²₂ = 131.580, p < 0.001; GLM for basophil proportion: χ²₂ = 5.183, p = 0.075; Figs. 3B and 5).

In all species we found that lymphocytes were the most prevalent leukocyte cell type and occurred in the greatest proportion in Lit. paraewingi (Figs. 4 and 5). Neutrophils and monocytes were equally common in Lit. paraewingi and Lim. dumerilii, but in C. signifera, neutrophils were much more common than in the other two frog species (Figs. 4 and 5). Eosinophils were rarely found in Lit. paraewingi—only seen once in the clinically sick animal with ectoparasites that was removed from the study (Figs. 2 and 4), while they were relatively common in Lim. dumerilii (Figs. 4 and 5).

The impacts of parasites on cell differentials

When we investigated the effect of parasite status (macroparasite or fungal infection presence) on blood smear differentials in the two species that had parasites, we found no effect of parasite status on the proportion of leukocytes (GLM: species, χ²₁ = 0.7349, p = 0.391; parasite status, χ²₁ = 0.084, p = 0.7719; species*parasite status, χ²₁ = 1.569, p = 0.210).

We found no effect of parasite status on leukocyte differentials when we analysed the two species that had parasites. Species was a significant predictor of the proportion of lymphocytes, monocytes and neutrophils, which coincides with the analyses conducted above (Fig. 5; GLM for lymphocyte proportion: Species, χ²₂ = 23.900, p < 0.001; parasite status, χ²₂ = 0.673, p = 0.412: GLM for monocyte proportion: Species, χ²₂ = 9.30, p = 0.003; parasite status, χ²₂ = 1.508, p = 0.219: GLM for neutrophil proportion: Species, χ²₂ = 38.388, p < 0.001; parasite status, χ²₂ = 0.541, p = 0.462).

Leukocyte differentials

In the healthy frogs without evidence of parasites there was no effect of species or sex on the relative proportions of leukocytes to erythrocytes within the blood sample. However, when we investigated these effects on leukocyte differentials, we found that there was a significant effect of species. Across all three species, the leukocyte present in the highest proportion was lymphocytes. Lymphocytes and neutrophils make up the majority of leukocytes in amphibians, similar to mammals and birds (Davis, Maney & Maerz, 2008). While the proportion of lymphocytes varies greatly across species, in amphibians it is frequently seen as the most common leukocyte (Davis & Durso, 2009; Young et al., 2012a; Forzán et al., 2016; Chen et al., 2022).

The relatively high proportion of eosinophils in Lim. dumerilii is of particular interest because none of the Lim. dumerilii that we sampled had detectable parasites. Eosinophils in most vertebrate animals have typically been considered as effector immune cells against extracellular, multicellular parasites like helminths (Lee et al., 2010). They have also been implicated in hypersensitivity diseases and there are several known disorders of eosinophils in mammals (Lee et al., 2010; Seimon et al., 2017; Wechsler et al., 2021). However, more recent research revealed that eosinophils may be involved in a range of homeostatic processes, suggesting a role in maintaining normal health (Wechsler et al., 2021). The high proportion of eosinophils in Lim. dumerilii might indicate highly effective innate immune protection against parasite infection or be related to a different mechanism for maintaining health.

We noted that in Lit. paraewingi, eosinophils were only found in one animal that was heavily infected with an ectoparasite (four parasitic fly larvae). Though investigating sick animals was outside the scope of this study and this was a singular finding, with further exploration it may demonstrate the links between eosinophils, maintaining normal health, and immunity against parasites. There is ongoing work exploring the health impacts of ectoparasite fly larvae on Australian frog species (J. Keem, 2022, personal communication).

The blood cell differentials of neither Lit. paraewingi nor C. signifera were affected by parasite status. Certain changes in leukocyte differentials, including monocytosis, lymphopaenia, and neutrophilia, are highly conserved across vertebrate species and associated with inflammatory disease processes (Davis, Maney & Maerz, 2008; Grzelak et al., 2017). The lack of significant changes in leukocyte differentials in this study could be because we combined macroparasites and fungal infection, and these parasites might elicit different haematological responses; infection with B. dendrobatidis can affect blood differentials, and the effect is often species dependent (Young et al., 2012b; Greenspan et al., 2017). However, parasites might cause a different response to a B. dendrobatidis skin infection, therefore leaving this study design unable to ascertain small differences in the leukocyte differentials. Future studies should collect a larger sample size in order to consider the pathogens/parasites separately. More likely, we might not have an observed change in leukocyte differentials because our sample of parasite positive animals was small (only 8% of all samples analysed). Future studies exploring the health of free-ranging amphibians are needed to understand the subclinical impacts of disease.

We collected approximately 5 μL of blood from each animal; a sufficient volume to create a blood smear but insufficient for haemocytometry. Typically, absolute erythrocyte counts are performed using a haemocytometer to calculate concentration (Forzán et al., 2016). Using a blood smear to analyse a ratio of erythrocytes within the blood, we were unable to determine absolute blood counts or estimate blood cell concentrations, as these values cannot be accurately extrapolated from blood smears (Forzán et al., 2017). Future studies would benefit from using haemocytometry. Furthermore, because the entire sample was used in the smear, we were unable to perform molecular analysis on the samples to identify any haemoparasites that might have been present but missed in our cytological analysis.