Long-term molecular surveillance of Cryptosporidium and Giardia in wildlife in protected drinking water catchments

Anson V Koehler; Tao Wang; Melita A Stevens; Shane R Haydon; Robin B Gasser

doi:10.1186/s13071-025-07048-8

. 2025 Oct 14;18:413. doi: 10.1186/s13071-025-07048-8

Long-term molecular surveillance of Cryptosporidium and Giardia in wildlife in protected drinking water catchments

Anson V Koehler ^1,^✉, Tao Wang ¹, Melita A Stevens ², Shane R Haydon ², Robin B Gasser ^1,^✉

PMCID: PMC12523218 PMID: 41088250

Abstract

Background

This study presents findings from a 15-year longitudinal surveillance program (2009–2024) monitoring Cryptosporidium and Giardia in protected drinking water catchments in Melbourne and environs in the State of Victoria, Australia. As one of the few major cities worldwide sourcing largely unfiltered water from forested catchments, Melbourne presents a unique opportunity to assess the occurrence and prevalence of protozoan parasites in a minimally disturbed ecosystem.

Methods

A total of 14,960 animal faecal samples were analysed using polymerase chain reaction (PCR)-based sequencing, including 8695 samples collected over the past 9 years.

Results

Cryptosporidium was detected in 3.15% of samples and Giardia in 0.16%. A total of 12 recognised Cryptosporidium species and genotypes were identified, nine of which have known zoonotic potential, as well as two sub-assemblages (AI and AIII) of Giardia duodenalis, including four novel assemblage AI variants. Parasite diversity was the highest in eastern grey kangaroos, which hosted at least 18 Cryptosporidium variants. Temporal analyses revealed significant inter-annual variation, with peak prevalence during the 2023 La Niña year and seasonal differences by host group. Notably, C. ubiquitum, C. muris and C. occultus were recorded for the first time in these catchments. In spite of the low prevalence of high-risk species such as C. parvum and the absence of C. hominis, the detection of emerging and previously uncharacterised genotypes emphasises the importance of sustained surveillance.

Conclusions

These findings have broad implications for managing zoonotic risk in unfiltered water systems worldwide. Advances in metagenomics and high-throughput sequencing platforms will be critical for enhancing future pathogen monitoring and catchment management strategies in the context of increasing climate and environmental pressures.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s13071-025-07048-8.

Keywords: Cryptosporidium, Giardia, Wildlife, Protected drinking water catchments, Long-term monitoring, Molecular tools

Background

The supply of fresh drinking water to people and the prevention of waterborne diseases are a major challenge, particularly in disadvantaged communities [1]. Worldwide, diarrhoeal diseases are the third leading cause of death and malnutrition in children under 5 years of age, equating to ~1.7 billion cases per annum [2–4]. Globally, Cryptosporidium alone accounts for a disease burden of 12.9 million disability-adjusted life-years (DALYs) [5], ranking second only to rotavirus as cause of severe diarrhoea in young children [6]. Together, Cryptosporidium and Giardia are responsible for most protozoal waterborne disease outbreaks worldwide [7].

Currently, there are 46 recognised species of Cryptosporidium and more than 120 genotypes [8, 9]. Of these, 19 species and four genotypes have been reported in humans, with C. parvum and C. hominis accounting for 95% of all Cryptosporidium species described to date [8]. There are more than eight phylogenetically distinct species of Giardia [10]. Cryptosporidium and Giardia are unique in that very small numbers of infective stages (oocysts and cysts, respectively) can lead to disease in humans [11, 12], and these stages are resistant to chlorination and other common water treatments [13]. Although outbreaks of cryptosporidiosis can have major adverse human health and economic impacts [14, 15], sub-clinically infected (asymptomatic) people and animals represent key reservoirs for transmission to susceptible or immune-suppressed or deficient people. This context emphasises the major public health importance of waterborne cryptosporidiosis and giardiasis and the need for sustained and effective prevention of these diseases.

Melbourne in the State of Victoria, Australia, is one of the few cities in high-income countries receiving largely unfiltered, chlorinated drinking water from protected wilderness catchment areas [16]. All water from these protected catchments is disinfected with chlorine before it is provided to the city. Each day, approximately 1250 million litres of drinking water are supplied to the more than 5 million residents of Greater Melbourne [16]. The management of these water catchment areas includes restricted access for humans, long water retention times and an intense program of testing and monitoring for pathogens in source water [16]. These catchments represent habitat for native and feral animals, such that the monitoring of zoonotic pathogens is central to management and the rigorous protection of the safety of drinking water [16].

Within the context of an integrated catchment management (ICM) framework, we have been monitoring Cryptosporidium and Giardia in faecal samples from various mammals and birds in Melbourne’s catchments [17, 18]. Since 2009, we consolidated a joint longitudinal program to consistently monitor (every 1–2 months) the presence of Cryptosporidium genotypes and subtypes as well as Giardia assemblages in faecal samples from animals in key catchments using molecular tools. All findings are reported to Melbourne Water, which reports to the Department of Health (DH) of the State of Victoria, Australia. Some of the findings produced have been published previously [17–21]. Here, we report the results from December 2015 to July 2024 and also review trends seen over the 15-year period since the beginning of the sample collection program in June of 2009.

Methods

Melbourne’s water catchments

Melbourne’s closed drinking water catchments, designed to minimise the risk of waterborne contamination by restricting domestic animal and human access, are primarily located in the Yarra Ranges, 40–90 km east of Melbourne, spanning more than 150,000 hectares of eucalypt forests. Additional off-stream reservoirs are located closer to the city of Melbourne (Fig. 1). All water is treated in accordance with national and international guidelines [22]. A total of ten major reservoirs and numerous weirs serve the city. The faecal sampling reported here has occurred from most of the catchments for these reservoirs and weirs. The Armstrong weir (AR), Maroondah reservoir (MR), O’Shannassy reservoir (OS), Thomson reservoir (TH) and Upper Yarra reservoir (UY) are situated in the dense eucalypt forests of the Yarra Ranges catchment. The remaining reservoirs – Yan Yean (YY), Cardinia (CA), Greenvale (GV) and Silvan (SV) – primarily function as off-stream storages for the larger catchments and are surrounded by eucalypt and/or pine forests. It should be noted that Winneke (WI) is a large conventional water treatment plant supplied from the off-stream Sugarloaf reservoir, which is in turn filled by pumping from a large multi-use catchment. Samples for this site are from around the treatment plant, not the Sugarloaf catchment. In total, 14,960 samples were collected and tested (from June 2009 to July 2024). Data sets from previous studies [17, 18] were used for selective, comparative analyses.

Fig. 1 — Map of Melbourne Water’s protected drinking water catchments (in blue)

Sampling and isolation of genomic DNA

In the context of our Cryptosporidium and Giardia monitoring program, faecal samples have been – and continue to be – collected from forested or grassy areas in close proximity to water reservoirs – where wildlife typically graze. From December 2015 to July 2024, 8695 faecal deposits from Wallabia bicolor (swamp wallaby), Macropus giganteus (eastern grey kangaroo), Vombatus ursinus (common wombat), Phascolarctos cinereus (koala), Trichosurus vulpecula (common brushtail possum), Rusa unicolor (sambar deer), Cervus elaphus (red deer), Dama dama (fallow deer), Oryctolagus cuniculus (rabbit), Mastacomys fuscus (broad-toothed rat), Rattus fuscipes (bush rat), Rattus lutreolus (swamp rat), Canis familiaris (dog), Canis familiaris dingo (dingo), Vulpes vulpes (red fox), Dromaius novaehollandiae (emu) and Chenonetta jubata (wood duck) and samples of unknown host origin were collected from nine locations. Specifically, samples were collected from AR (n = 15), CA (n = 1902), GV (n = 1188), MR (n = 1482), OS (n = 302), SV (n = 1489), UY (n = 620), WI (n = 366) and YY (n = 1331) (Fig. 1). It is important to note that some animal species are not evenly distributed across catchments. For instance, fallow deer and emus are seen only in Cardinia, while Greenvale is home to kangaroos, rabbits and foxes. Scats were initially identified using a field guide [23].

From aliquots (~50 mg) of individual faecal samples, genomic DNA was extracted using the PowerSoil kit (MoBio Powersoil Kit; 2015 to April 2021), later rebranded as the DNeasy PowerSoil Pro kit (Qiagen; May 2021 to July 2024) following the manufacturers’ instructions. The specific identity of animals from which individual samples originated was verified by polymerase chain reaction (PCR)-based sequencing of a region of the mitochondrial cytochrome b (cytb) gene from faecal DNA using an established method [24]. For some analyses, data for samples from three species of deer (i.e. sambar, red and fallow) were pooled (designated as ‘deer’), as for samples from eastern grey kangaroos and swamp wallabies (designated ‘macropods’).

PCR-coupled sequencing, sequence alignments and associated analyses

During the monitoring program (from 2015 to 2024), distinct genetic loci have been used to identify Cryptosporidium and Giardia species via nested PCR-based sequencing (reagents and cycling conditions are presented in Table 1). For the identification of Cryptosporidium species, small subunit ribosomal ribonucleic acid (RNA) gene (SSU) sequences (800–830 bp) were obtained using well-established methods [18, 21, 25–28]. On occasions, large subunit ribosomal RNA gene (LSU) sequences (500 bp) were obtained using another technique [29]. For the differentiation of C. parvum, C. hominis and C. cuniculus, a 60-kDa glycoprotein (gp60) gene region (1000 bp) was sequenced using an established approach [18, 30, 31]. For the identification and differentiation of Giardia species and assemblages, a triose-phosphate isomerase (tpi) gene region (530 bp region) was amplified and sequenced using a published methodology [17, 18, 32]. Known test-positive, test-negative and no-template (including ‘carry-over’) controls were included in each step of each set of PCRs. Bidirectional sequencing of amplicons was conducted by Macrogen (South Korea) or the Australian Genome Research Facility (AGRF; Melbourne).

Table 1.

The primers and conditions used to obtain Cryptosporidium and Giardia

Region	PCR	Primers	Size (bp)	Denaturation/annealing/extension	References
LSU	1°	LSU2040F—5′ CGAATAGCGTTATCTTTGCTATTT 3′	500	94 °C/30 s 58 °C/30 s 72 °C/50 s 35 cycles	[29]
	1°	LSU3020R—5′ GTCTTCCGCGAAGATCAG 3′		94 °C/30 s 58 °C/30 s 72 °C/50 s 35 cycles
	2°	LSU2065F—5′ TTACCATGGAAT(C/T)AGTTCAGC 3′		94 °C/30 s 58 °C/30 s 72 °C/50 s 35 cycles
	2°	LSU2557R—5′ AACACCATTTTCTGGCCATC 3′		94 °C/30 s 58 °C/30 s 72 °C/50 s 35 cycles
SSU	1°	18SiCF2—5′ GACATATCATTCAAGTTTCTGACC 3′	800	94 °C/30 s 58 °C/30 s 72 °C/30 s 45 cycles	[26]
	1°	18SiCR2—5′ CTGAAGGAGTAAGGAACAACC 3′		94 °C/30 s 58 °C/30 s 72 °C/30 s 45 cycles
	2°	18SiCF1—5′ CCTATCAGCTTTAGACGGTAGG 3′		94 °C/30 s 58 °C/30 s 72 °C/30 s 45 cycles
	2°	18SiCR1—5′ TCTAAGAATTTCACCTCTGACTG 3′		94 °C/30 s 58 °C/30 s 72 °C/30 s 45 cycles
SSU	1°	SSU-F2—5′ TTCTAGAGCTAATACATGCG 3′	830	94 °C/45 s 55 °C/45 s 72 °C/60 s 35 cycles	[25, 71, 86]
	1°	SSU-R2—5′ CCCATTTCCTTCGAAACAGGA 3′		94 °C/45 s 55 °C/45 s 72 °C/60 s 35 cycles
	2°	SSU-F3—5′ GGAAGGGTTGTATTTATTAGATAAAG 3′		94 °C/45 s 55 °C/45 s 72 °C/60 s 35 cycles
	2°	SSU-R4—5′ CTCATAAGGTGCTGAAGGAGTA 3′		94 °C/45 s 55 °C/45 s 72 °C/60 s 35 cycles
gp60	1°	gp15-ATG—5′ ATGAGATTGTCGCCTCATTATC 3′	1000	94 °C/30 s 55 °C/45 s 72 °C/60 s 35 cycles	[31]
	1°	gp15-STOP—5′ TTACAACACGAATAAGGCTGC 3′		94 °C/30 s 55 °C/45 s 72 °C/60 s 35 cycles
	2°	gp15-15A—5′ GCCGTTCCACTCAGAGGAAC 3′		94 °C/30 s 55 °C/30 s 72 °C/30 s 35 cycles
	2°	gp15-15E—5′ CCACATTACAAATGAAGTGCCGC 3′		94 °C/30 s 55 °C/30 s 72 °C/30 s 35 cycles
tpi	1°	AL3543—5′ AAATTATGCCTGCTCGTCG 3′	530	94 °C/45 s 50 °C/45 s 72 °C/60 s 35 cycles	[32]
	1°	AL3546—5′ CAAACCTTTTCCGCAAACC 3′		94 °C/45 s 50 °C/45 s 72 °C/60 s 35 cycles
	2°	AL3544—5′ CCCTTCATCGGTGGTAACTT 3′		94 °C/45 s 50 °C/30 s 72 °C/60 s 35 cycles
	2°	AL3545—5′ GTGGCCACCACTCCCGTGCC 3′		94 °C/45 s 50 °C/30 s 72 °C/60 s 35 cycles

Open in a new tab

DNA sequence data obtained were matched to sequences in the GenBank database using the Basic Local Alignment Search Tool (BLAST; www.ncbi.nlm.nih.gov) and subsequently aligned to curated sequences representing all known distinct Cryptosporidium species/genotypes and Giardia species/assemblage as well as outgroup taxa in our in-house sequence repository. Pairwise sequence comparisons were conducted, and sequence identities were calculated using Geneious Prime, version 2024.0.7, software (www.geneious.com). Subsequently, sequences were aligned using the MAFFT program [33] and alignments manually adjusted in Mesquite, version 3.81 [34].

Phylogenetic analysis of sequence data was performed by Bayesian inference (BI) using Monte Carlo Markov Chain (MCMC) analysis in MrBayes, version 3.2.7 [35]. Likelihood parameters for the BI analysis of SSU data were selected on the basis of the Akaike information criterion (AIC) test in MEGA11 [36], with the general time-reversible (GTR) model having the lowest AIC score. The number of substitutions model (Nst) was set at six, with a gamma distribution and a proportion of invariable sites. Posterior probability (pp) values were calculated after 10,000,000 generations with four simultaneous tree-building chains, saving trees at every 100th generation. At the end of each run, convergence was confirmed by a standard deviation of split frequencies < 0.01, and the potential scale reduction factor approached 1. A 50% majority-rule consensus tree was constructed from the final 75% of trees generated. Each analysis was run three times to ensure convergence and insensitivity to priors. The outgroup included in the analysis was C. abrahamseni. Additional statistical analyses were performed using R Statistical Software (version 4.5.1), with graphs generated using ggplot2, dplyr and tidyr packages.

Results

Identification and classification of Cryptosporidium on the basis of SSU

From 8695 faecal DNA samples (prevalence: 3.15%), 274 SSU amplicons were obtained, and 13 species and genotypes of Cryptosporidium were identified (Additional File 1: Supplementary Table S1). In total, 35 unique sequences were assigned GenBank accession numbers (PV658852 to PV658885 and PV668769) (Table 2) and used for phylogenetic analysis (Fig. 2). In total, we obtained 23 novel SSU sequences (i.e. with < 100% identity to sequences in GenBank) – not previously published or deposited in public databases. Each sequence was linked to a host species and water catchment area, and its prevalence was recorded (Additional File 1: Supplementary Table S1). The final designation of species and genotype was based on the alignment and position of a sequence within the phylogenetic tree (Fig. 2). In this tree, most sequences were assigned to the three predominant clades – i.e. C. ryanae, C. fayeri and C. macropodum; the remaining sequences were external to these clades (Fig. 2). In the following sections, we explore Cryptosporidium species and genotypes according to host group.

Table 2.

Summary of Cryptosporidium taxa (species and genotypes) identified in faecal samples collected from animals in Melbourne’s protected drinking water catchment areas (December 2015 to July 2024)

Species/genotype	Total	Designation	Subtotal		Human reports^*	GenBank no.
Cryptosporidium baileyi	4	Cryptosporidium baileyi	4		None	PV658852
Cryptosporidium canis	2	Cryptosporidium canis	2		Many	PV658853
Cryptosporidium cuniculus	6	Cryptosporidium cuniculus	6		Many	PV658854
Cryptosporidium fayeri	14	Cryptosporidium fayeri	63		Few	PV658855
Cryptosporidium fayeri EGK1	1	Cryptosporidium fayeri				PV658856
Cryptosporidium fayeri EGK1 v2	7	Cryptosporidium fayeri				PV658857
Cryptosporidium fayeri EGK1 v3	1	Cryptosporidium fayeri				PV658858
Cryptosporidium fayeri EGK1 v4	18	Cryptosporidium fayeri				PV658859
Cryptosporidium fayeri EGK1 v5	1	Cryptosporidium fayeri				PV658860
Cryptosporidium fayeri KG1-like	8	Cryptosporidium fayeri				PV658861
Cryptosporidium fayeri KG1-like v1	1	Cryptosporidium fayeri				PV658862
Cryptosporidium fayeri KG1-like v2	2	Cryptosporidium fayeri				PV658863
Cryptosporidium fayeri KG1-like v3	2	Cryptosporidium fayeri				PV658864
Cryptosporidium fayeri v1	1	Cryptosporidium fayeri				PV658865
Cryptosporidium fayeri v2	1	Cryptosporidium fayeri				PV658866
Cryptosporidium fayeri v3	1	Cryptosporidium fayeri				PV658867
Cryptosporidium fayeri v4	1	Cryptosporidium fayeri				PV658868
Cryptosporidium fayeri v5	4	Cryptosporidium fayeri				PV658869
Cryptosporidium macropodum	112	Cryptosporidium macropodum	117		None	PV658870
Cryptosporidium macropodum v1	2	Cryptosporidium macropodum				PV658871
Cryptosporidium macropodum v2	2	Cryptosporidium macropodum				PV658872
Cryptosporidium macropodum v3	1	Cryptosporidium macropodum				PV658873
Cryptosporidium muris	4	Cryptosporidium muris	4		Many	PV658874
Cryptosporidium occultus	1	Cryptosporidium occultus	2		Several	PV658875
Cryptosporidium occultus	1	Cryptosporidium occultus	2		Several	PV668769
Cryptosporidium parvum-like	1	Cryptosporidium parvum-like	1		Many	PV658876
Cryptosporidium ryanae-like v1	21	Cryptosporidium ryanae	68		None	PV658877
Cryptosporidium ryanae-like v2	3	Cryptosporidium ryanae				PV658878
Cryptosporidium ryanae-like v3	1	Cryptosporidium ryanae				PV658879
Cryptosporidium sp. deer genotype v1	4	Cryptosporidium ryanae				PV658880
Cryptosporidium sp. deer genotype v2	4	Cryptosporidium ryanae				PV658881
Cryptosporidium sp. deer genotype v3	35	Cryptosporidium ryanae				PV658882
Cryptosporidium sp. possum genotype	1	Cryptosporidium sp. possum genotype	1			PV658883
Cryptosporidium ubiquitum cf C3604	1	Cryptosporidium ubiquitum	3			PV658884
Cryptosporidium ubiquitum cf W13491	2	Cryptosporidium ubiquitum	3			PV658885
Cryptosporidium viatorum	3	Cryptosporidium viatorum	3		Many	MG021320
Total	274			^*According to ref. [8]

Open in a new tab

Cryptosporidium in deer

Within the C. ryanae clade (Fig. 2), four SSU sequence variants representing C. ryanae and three representing ‘Cryptosporidium sp. deer genotype’ [37] were recorded for 66 of a total of 1774 samples from deer (Fig. 2; Tables 2 and 3). Cryptosporidium sp. deer genotype v3 was the most prevalent, and recorded in 35 samples, followed by C. ryanae-like v1 in 16 samples; three were classified as C. ryanae-like using LSU [29] but were not characterised by SSU and, therefore, were not included in the phylogeny (Fig. 2). The majority of these 66 records were from sambar deer (n = 49), followed by red deer (n = 7) and fallow deer (n = 3) (Additional File 1: Supplementary Table S1); seven were from deer of which species identity was not determined by molecular means [24]. In addition, C. muris was detected once in sambar deer in CA and YY catchments in 2021 and again at YY in 2023; C. ubiquitum (accession no. PV658885) and C. occultus (accession no. PV668769) were each detected once in sambar deer in the UY catchment.

Table 3.

Summary of Cryptosporidium taxa detected in 8695 faecal samples from animals in Melbourne’s protected drinking water catchments (2015–2024)

Catchment	Macropod	Deer	Rabbit	Canid	Rat	Waterbird	Possum	Wombat	Cryptosporidium total	Faecal total
Armstrong									0	15
Cardinia	19 macropodum, 6 fayeri	13 ryanae, 1 muris		1 canis					40	1902
Greenvale	36 macropodum, 12 fayeri			1 canis					49	1188
Maroondah	11 macropodum, 13 fayeri	4 ryanae							28	1482
O’Shannassy	1 parvum-like	12 ryanae	1 cuniculus						14	302
Silvan	5 fayeri, 2 macropodum		4 cuniculus			4 baileyi			15	1489
Upper Yarra		17 ryanae, 1 occultus 1 ubiquitum			3 viatorum, 1 occultus			1 ubiquitum	24	620
Winneke	22 macropodum, 1 ubiquitum								23	366
Yan Yean	27 fayeri, 27 macropodum	22 ryanae, 3 muris	1 cuniculus				1 possum genotype		81	1331
Total	182	74	6	2	4	4	1	1	274	8695

Open in a new tab

Cryptosporidium in marsupials

Within the C. fayeri clade, 13 variants were recorded among a total of 57 SSU sequences. Within the C. macropodum clade, five variants were detected among a total of 116 SSU sequences obtained from a total of 6014 samples from macropods (94.1% from eastern grey kangaroos; Fig. 2; Tables 2 and 3). Among the positive Cryptosporidium detections, nearly all were from kangaroos, with the exception of two wallabies that carried C. macropodum v2. The first record of C. ubiquitum in a macropod was from a kangaroo at Winneke, Victoria (accession no. PV658885), while a single variant of C. ubiquitum (represented by accession no. PV658884) was detected among 217 wombats. In addition, one Cryptosporidium sp. possum genotype was detected from seven brush-tailed possum samples. Notably, a C. parvum-like sequence (accession no. PV658876) was detected in a wallaby, with two nucleotide sites displaying multiple peaks that distinguish it from typical C. parvum sequences. Unfortunately, the gp60 region did not amplify from this sample. Cryptosporidium was not detected in the sole koala sample tested.

Cryptosporidium in other animals (canids, lagomorphs, rodents and birds)

Of the 39 canid samples tested, C. canis was detected in a fox from Greenvale and a dingo from Cardinia (Table 3). From the 82 rodents examined, three tested positive for C. viatorum [21] and one for a variant of C. occultus – confirmed by the alignment criteria set forwards from Stensvold et al. [38] (Table 3). Of the 198 samples from rabbits tested, five were positive for C. cuniculus (Table 3). Of these, two samples were sub-typed using the gp60 gene sequence: the first sequence (accession no. PV665486) defined subtype VbA31R4 originating from OS (December 2018) but with one point mutation over a 300 bp region when compared with the GenBank sequence accession no. KU852732 (822 bp) representing C. cuniculus from a human from Greece; the second sequence (accession no. PV665487) defined subtype VbA15, from YY. Of the 223 samples from waterbirds (predominantly wood ducks) tested, four contained C. baileyi (Table 3), but none of the 106 samples from emu tested positive for Cryptosporidium.

Giardia

Giardia was detected in 14 of the 8695 faecal DNA samples tested (Table 4). G. duodenalis was identified in seven deer – six belonging to sub-assemblage AIII and one to sub-assemblage AI. In eastern grey kangaroos, sub-assemblage AI of G. duodenalis was detected in five individuals, including four novel genetic variants (GenBank accession nos. PV665481–PV665484). Giardia sub-assemblage AI was also detected in one dog. No G. duodenalis was detected in rabbits, rodents or birds.

Table 4.

Total count of Giardia by catchment and host

Catchment	Canid	Deer	Macropod	Unknown	Total
Armstrong
Cardinia		4 (1 AI, 3 AIII)			4
Greenvale			2 (2 AI)		2
Maroondah			1 (AI)		1
O’Shannassy	1 (AI)				1
Silvan			1 (AI)		1
Upper Yarra
Winneke
Yan Yean		3 (2 AI, 1 AIII)	1 (AI)	1 (AI)	5
Total	1	7	5	1	14

Open in a new tab

Coinfections

Both Cryptosporidium and Giardia were detected in a small number of samples (n = 3). Cryptosporidium sp. deer genotype v3 and G. duodenalis subtype AIII were detected in a red deer at YY in the summer of 2018, and Cryptosporidium sp. deer genotype v1 and G. duodenalis subtype AI were detected in a fallow deer at CA in autumn 2021. In addition, C. fayeri v2 and G. duodenalis subtype AI were identified in one kangaroo in the SV catchment in the summer of 2020.

Temporal effects

Cryptosporidium prevalence ranged from a low of 0.99% in 2016 to a high of 6.15% in 2023, with an overall mean prevalence of 3.15% (274 records among 8695 samples). When data from 2009 onwards was considered, the total prevalence across all 16 years was 2.67% (399 cases out of 14,960 samples) (Fig. 3). A statistically significant increase in prevalence was seen across years (adjusted R-squared 0.294; P = 0.0175).

Fig. 3 — Prevalence of *Cryptosporidium* in animals in Melbourne’s protected drinking water catchments by year since 2009 – with 95% confidence interval bars

Over the entire period from 2009 to 2024, Cryptosporidium prevalence was the lowest in spring at 2.24% (78 records in 3482 samples) and the highest in summer at 3.57% (110 records from 3078 samples). Prevalences in autumn and winter were 2.48% (96 records from 3877 samples) and 2.47% (110 records from 4460 samples), respectively (Fig. 4). According to host groups and season, macropods had the highest Cryptosporidium prevalence in summer (4.10%; in 81 of 1976 samples) and spring (2.83%; in 53 of 1873 samples), with lower prevalences in autumn (1.80%; in 38 of 2106 samples) and winter (1.95%; in 47 of 2405 samples). In contrast, deer exhibited the highest Cryptosporidium prevalences in autumn (5.47%; in 32 of 1145 samples) and winter (2.47%; in 35 of 1418 samples), with lower prevalences in spring (1.51%; 17 of 1123 samples) and summer (0.67%; 4 of 600 samples) (Fig. 5).

Fig. 4 — Prevalence of *Cryptosporidium* in animals in Melbourne’s protected drinking water catchments by season since 2009 – with 95% confidence interval bars

Fig. 5 — Prevalence of *Cryptosporidium* in macropods and deer – by season – in Melbourne’s protected drinking water catchments since 2009 – with 95% confidence interval bars

An analysis of the monthly prevalence across all years for the three dominant Cryptosporidium species (C. fayeri, C. macropodum and C. ryanae) revealed distinct seasonal trends. C. macropodum exhibited a peak prevalence in the summer month of January, with the lowest prevalence recorded in June. In contrast, C. fayeri and C. ryanae showed peak prevalences in the winter months (Fig. 6).

Fig. 6 — Prevalences of *Cryptosporidium fayeri*, *C. macropodum* and *C. ryanae* in animals – by month – within Melbourne’s protected drinking water catchment system since 2009 – with 95% confidence interval bars

Discussion

Ensuring safe drinking water from natural catchments requires effective risk management and an understanding of protozoan parasite populations in the wildlife found in these areas. Melbourne Water’s chlorination treatment targets bacterial and viral pathogens but is not designed to remove the key protozoan parasites – Cryptosporidium and Giardia – which can pose health risks that are not always treatable with current chemotherapeutic options. Thus, ongoing surveillance of parasitic pathogens in catchments has been essential. This longitudinal study provides a strong and consistent foundation for understanding Cryptosporidium and Giardia populations in Melbourne’s catchment environments. Given that these parasites cannot be accurately identified or readily cultured with conventional laboratory techniques, this study employed cost-effective molecular genetic methods for their genetic identification and characterisation.

Here, we report the most recent results from one of the largest, continuous and most comprehensive longitudinal investigations of Cryptosporidium and Giardia in animals inhabiting protected drinking water catchment areas, serving the major metropolitan city of Melbourne in Australia. Nonetheless, smaller studies have been conducted in Victoria, New South Wales, South Australia, Queensland and Western Australia [39–46]. Overall, the prevalence of these two genera of protozoan parasites was relatively low (3.15% for Cryptosporidium and 0.16% for Giardia), with some fluctuations in prevalence from 2009 to 2024 and some spikes of Cryptosporidium in summer 2023. The prevalence of Giardia was extremely low, with sub-assemblages AI and AIII detected (assemblage A). Notably, four novel AI variants were identified. Sub-assemblage AI was present in deer, macropods and a canid, whereas AIII was exclusively found in deer – in accord with previous studies overseas [10, 47]. However, the low number of Giardia detections limits our ability to draw major conclusions from these data.

The low prevalences of Cryptosporidium and Giardia recorded compared with surveys in other states of Australia [39–46] might relate to our much larger sample sizes, the host species assessed, variation in habitats and/or catchment management practices, including the control or removal of animals from catchments. Although Giardia duodenalis assemblage AI (accession no. PV665479) has recognised zoonotic potential and was detected at a low prevalence (Table 4), Cryptosporidium remains the predominant concern for the water industry, as waterborne cryptosporidiosis in humans is typically far more challenging to manage and treat – particularly in immunosuppressed or immunodeficient individuals [3, 48].

Using PCR-based sequencing-phylogenetic analysis, we recorded 12 recognised species and genotypes of Cryptosporidium (Table 2). Nine species, known to occur in humans (C. canis, C. cuniculus, C. fayeri, C. muris, C. occultus, C. parvum, C. ubiquitum and C. viatorum) were recorded here in canids, deer, rats, rabbit, kangaroo, wallaby or wombat. Of the 46 recognised species of Cryptosporidium [8, 9], C. hominis and C. parvum are the most likely to cause human cryptosporidiosis [49] and are usually the key agents linked to waterborne outbreaks [15]. Despite their significance, the lack of C. hominis and the very low prevalence of C. parvum (0.01%) in the vast natural catchments (Table 3) suggests a negligible risk of waterborne transmission to humans in Greater Melbourne. Nonetheless, it is possible that some of the other Cryptosporidium species and genotypes discovered might have zoonotic potential.