Genetic characterisation of Cryptosporidium parvum in dairy cattle and calves during the early stages of a calving season

Cryptosporidium parvum is a causative agent of cryptosporidiosis, an infectious gastroenteritis in neonatal ruminants, which can be fatal in severe cases. The aim of this study was to determine the prevalence of infections in dairy cattle/calves during the early stages of a calving season and the species/genotypes of the Cryptosporidium present. Faecal samples collected from pre- and post-partum dams (n = 224) as well as calves from age ∼1 day onwards (n = 312) were examined. Oocysts were concentrated, DNA extracted and tested by Cryptosporidium 18S rRNA gene PCR and sequencing, while genotypes of C. parvum were determined by gp60 and VNTR analysis. Results showed that 31.3% and 30.4% of pre- and post-partum dams tested positive for Cryptosporidium, respectively. In the adults, C. parvum (n = 52), C. bovis (n = 4) and C. andersoni (n = 19) were identified, while in the calves 248 out of 312 (79.5%) were PCR-positive for C. parvum. The proportion of positive calf samples was significantly higher (P < 0.0001) than the proportion of positive adult cattle during the first seven weeks of the calving season. In adult cattle, three distinct gp60 genotypes were identified, a predominant genotype IIaA15G2R1 (n = 36) and genotypes IIaA15R1 (n = 2) and IIaA14G2R1 (n = 1). In the calves, only genotype IIaA15G2R1 was detected (n = 125). Although C. parvum was observed in adult cattle two weeks after the start of the calving season, the predominant genotypes were not detected until Week 4 in both adults and calves, meaning it is still unclear whether adult cattle are the initial source of C. parvum infections on the farm. Historically calves on this dairy farm demonstrated the IIaA19G2R1 genotype, which, has now clearly been replaced with the IIaA15G2R1 genotype that is now found in both adults and calves. During the study season, significantly higher levels of neonatal calf mortality were observed compared to the seasons before (P = 0.046) and after (P = 0.0002). This study has shown comparable levels of C. parvum infection in both pre- and post-partum dams but higher levels of infection in neonatal calves.


Introduction
Cryptosporidiosis is one of the most commonly diagnosed causes of infectious enteritis in neonatal dairy calves (Shaw et al., 2021).From a One Health perspective, the most important species is Cryptosporidium parvum, not only because it can cause potentially fatal disease in neonatal ruminants, but as it is also zoonotic and a major cause of human cryptosporidiosis worldwide, with young children being especially vulnerable (Helmy et al., 2015).There are three other main species of Cryptosporidium that are found in cattle, C. bovis, C. ryanae which are generally found in older calves (4 weeks to ≤ 6 months of age) (Thomson et al., 2019b), while C. andersoni is usually found in post-weaned calves and adult animals (Wang et al., 2011).Currently, there are at least 47 recognised species of Cryptosporidium, as well as many more genotypes infecting a range of host species, from mammals and birds to amphibians and reptiles (Prediger et al., 2021;Ryan et al., 2021;Huang et al., 2023;Tůmová et al., 2023).
Cryptosporidium parvum is the only species infecting cattle that is associated with clinical disease in naturally infected neonatal calves.The clinical signs of cryptosporidiosis in calves include profuse watery diarrhoea, dehydration, loss of appetite and depression, which in severe cases can result in calf mortality (Thomson et al., 2017).This susceptibility to C. parvum means that following infection young calves can shed huge numbers (3.89 × 10 10 ) of oocysts in a very short period (6-12 days) (Nydam et al., 2001).Although older calves (≥ 6 months) and adult cattle are known to be naturally infected with C. parvum (Shaw et al., 2021) these infections are generally sub-clinical/asymptomatic, with infected adult animals shedding much lower numbers of oocysts, compared to neonates.
There is very little information available for cattle as to whether parturition influences levels of C. parvum oocyst shedding.In one study no oocysts were found in faecal samples collected from dams 12 h after calving (Atwill and Pereira, 2003), while another study found that 6/32 (18.8%) periparturient cows (within ± 7 days of giving birth) were excreting C. parvum oocysts (Castro-Hermida et al., 2002).In another study, Thomson et al. (2019a) found that four weeks before calving 27.3% (n = 57) of adult cattle samples tested positive for Cryptosporidium.These data suggest that levels of oocyst shedding in/around pregnancy in cattle are relatively low.
We were interested in investigating what role (if any) the dams play in the initiation and spread of C. parvum infections during the early stages of a calving season on a commercial working dairy farm.We were also interested in the species and genotypes of Cryptosporidium present in both adults and calves, as well as the timings of infections.To do this we collected faecal samples from adult cattle in the last two weeks of pregnancy and from post-partum cattle.We also collected samples from the calves once they had been separated from the dams (~1 day of age).

Study design and farm setup
The study was carried out on a working dairy farm located in Midlothian (Central Scotland, UK).The farm contains a closed herd, with all replacement stock being bred on-site using artificial insemination of sexed semen from known genetic lines, while the cattle not being used to produce replacement calves were bred using beef cattle semen.The farm has a history of C. parvum infection in their cattle (Thomson et al., 2019b) and also in farm-based infections of people (Gait et al., 2008) The calving season takes place from mid-August to late March.No calves are born between April and mid-August.At the end of the calving season, all the pens are cleared, cleaned, disinfected, and then left clean and empty until the next calving season.
The cattle and calves were fully managed by the farm staff (stockmen), with their movement between pens being coordinated as per normal farm practices.During the study period, all the animals were housed in concrete sided pens that contained straw bedding, which was changed routinely during the calving season.The straw in all of the pens (adults and calves) was refreshed 3-4 times a week.Each batch calf pen was mucked out between batches of calves and pressure washed and disinfected with Downland Blitz (Downland Partners in Animal Health, Carlisle, UK).The adult cattle pens were mucked out on a monthly basis.
The adult cattle sampled during the present study were divided into two separate pens.All dams are vaccinated with Rotovec (MSD Animal Health, Milton Keynes, UK), 4-8 weeks prior to calving.Pen A contained pre-partum dams in the last 1-2 weeks of pregnancy.Once a dam showed signs of labour, they were moved into Pen B, where they were left until they gave birth.Calves were usually removed from Pen B within approximately 12 h of birth.Post-partum dams remained in Pen B for 2-5 days, to ensure that there were no post-partum complications, after which time the dams were returned to the herd.
Following removal from their mothers (Pen B), the calves were initially housed in small groups (1-7 calves) in Pens CA -CE for 1-7 days (aim of 2-3 days only) where the calves were bottle-fed by the stockmen, before being moved to large group pens of up to 25 calves per pen (Pens D, E and G), where the calves were housed for the next 10-11 weeks.
Pens CA, CB and CC are located directly adjacent to Pen B, while Pens CD and CE are in a separate building and had not been used to house calves for several years prior to this calving season.Pens D and E are opposite to each other in a separate building away from the newborn and adult cattle.Pen G was also situated in another separate building away from all the other animals.
Once in the larger group pens (Pens D, E and G) each calf was fitted with a calf-feeder transponder collar and fed using automated milk machines, which monitored and limited the amount of milk each animal received.This system also allowed the farm staff to identify calves that were not feeding sufficiently and intervene when needed.A schematic diagram of the farm is provided in Supplementary Fig. S1.
Data were also collected on the total number of live/dead calves born as well as examining levels of calf mortality (before the age of 42 days), during the entire calving season.
Clinical observations were made by the scientist between weeks 5-19 whilst collecting samples from the pens; these observations focused particularly on the neonatal calves and included information on the general demeanour of the animals in specific pens and when clear signs of diarrhoea were seen.

Adult samples
Adult cattle faecal samples (n = 224) were collected directly from the floor of each pen (Pen A and Pen B).For each sample, approximately 100 g of faeces was collected directly into a plastic bag, which was sealed until processing, usually within 1-5 days.Samples were collected randomly from different locations around the entirety of each pen.Each sample was appropriately labelled and then stored at 4 • C until processing.

Calf samples
Calf faecal samples were also collected directly from the floor of each pen.For each individual sample, as large a quantity of faeces as possible was collected and placed into a labelled plastic bag.Each calf pen was thoroughly searched (usually by two or more scientists) to try to collect as many faecal samples as possible at each sampling time point.In the laboratory, the faeces were weighed, mixed with an equal volume of water and then transferred into a plastic tube.Once diluted, the samples were thoroughly mixed and then stored at 4 • C for 24-48 h prior to processing, to allow the faecal material to soften, aiding oocyst recovery.

Adult samples
The oocysts in adult faecal samples (~50 g) were concentrated using the acid flocculation and saturated salt floatation method previously described by Wells et al. (2016) with the following alterations.The pellets from the acid flocculation were resuspended in a 50 ml tube containing 30 ml saturated salt and 7.5-10 ml of water was trickled on top of the salt solution prior to centrifugation.The layer containing the oocysts was removed with a disposable plastic pipette and made up to a total volume of 25 ml in water prior to centrifugation.The final dilution volume of T1 buffer (Macherey-Nagel tissue kit) was 200-400 μl depending on the size of the pellet.Samples were stored at this stage at − 20 • C prior to DNA extraction.

Calf samples
As the calf samples were generally much smaller in volume than the adult samples, acid flocculation was not used.On arrival in the laboratory, all the bedding material (i.e.straw) was removed from the calf sample, which was then weighed and diluted with an equal volume of water (as described above in Section 2.2.2).For samples where the final diluted faecal volume was less than 15 ml, the entire sample was used for salt floatation.For samples where ≥ 15 ml of diluted faeces was available, then 15 ml was taken forward to saturated salt floatation (as described in Section 2.3.1).

DNA extraction
Prior to DNA extraction, each sample in T1 buffer underwent 10 cycles of freezing/thawing in liquid nitrogen/56 • C water bath, to disrupt the oocysts walls.The samples were then processed using a modified Macherey-Nagel NucleoSpin tissue kit (Macherey-Nagel, Düren, Germany) protocol, according to the manufacturerʼs instructions with the following amendments: the samples were incubated with 25 μl (10 mg/ml) proteinase K overnight (~16 h).The samples were then vigorously vortexed and centrifuged at 14,000× g for 1 min and 200 μl of the resultant supernatant was removed into a fresh microfuge tube prior to further processing.For the final elution, each DNA sample was eluted in 50 μl of DNase/RNase-free water (dH 2 O).All DNA samples were stored at − 20 • C prior to PCR analysis.An extraction control (dH 2 0) was processed with every batch of 23 samples to ensure that no crosscontamination occurred.
In the secondary amplification, a multiplex PCR was run with the following primers (AL1598: 5′-GGA AGG GTT GTA TTT ATT AGA TAA AG-3′, AL3032: 5′-AAG GAG TAA GGA ACA ACC TCC A-3′ and CphF: 5′-AGA GTG CTT AAA GCA GGC ATA-3′) included in the reaction mixture and only 12.85 μl of dH 2 0 was added.Furthermore, the entire primary nPCR product was diluted with 100 μl of dH 2 0 and 2 μl of the diluted amplicon was used in the secondary amplification.All PCR reactions were performed in triplicate; a sample was considered positive if at least 1/3 replicates tested positive.A negative control (dH 2 0), a DNA extraction control, along with a positive control for C. parvum DNA were included with each batch of samples.All samples that tested positive with the Cryptosporidium spp.primers but negative with the C. parvum primers were sent for sequencing to confirm the species of the Cryptosporidium present.A proportion of the samples that gave positive results using the C. parvum primers were also sent for sequencing to confirm the specificity of the primers.Results from each plate of PCR were only accepted if all of the positive, negative and extraction controls produced the expected result.

gp60 PCR
Samples that produced a positive result using the C. parvum-specific 18S primers were further analysed using a nested gp60 PCR, which has been previously described by Brook et al. (2009) and amplifies a final fragment of 375 bp.The PCR amplifications were completed in 20 μl reactions using the same basic master mix as previously described (Section 2.5), the forward and reverse primers (GP60_1F: 5′-ATA GTC TCC GCT GTA TTC-3 and GP60_1R: 5′-GAG ATA TAT CTT GGT GCG-3′) for the primary amplification and the primers for the secondary amplification (GP60_2F: 5′-TCC GCT GTA TTC TCA GCC-3′ and GP60_2R: 5′-CGA ACC ACA TTA CAA ATG AAG-3′).The entire primary PCR amplicon was diluted with 100 μl of dH 2 0, before 2 μl was used for the secondary amplification.Each sample was tested in triplicate, along with negative, extraction and positive controls for C. parvum.
The cycling conditions for both the 18S and the gp60 PCR amplifications were 3 min at 94 • C, followed by 35 cycles of 45 s at 94 • C, 45 s at 55 • C and 1 min at 72 • C and a final extension step of 7 min at 72 • C. The completed PCR amplifications were held at 10 • C.

Variable Number Tandem Repeat (VNTR) multiplex PCR
Samples that produced positive results from the gp60 PCR, where a clear gp60 genotype was identifiable (using sequence analysis, see Section 2.10), were further analysed using the Variable Number Tandem Repeat (VNTR) PCR.The details for the VNTR PCR have been previously published as part of a Multi Locus VNTR Analysis (MLVA) typing scheme for Cryptosporidium parvum (Robinson et al., 2022).In brief, the PCRs comprise seven sets of primers, that are divided into two separate multiplex PCRs (one containing four primer sets and the other with three primer sets; Table 1).Each primer set was labelled with a unique fluorescent dye (PET, VIC, NED or FAM) and generated fragments of different sizes (Table 1).Each PCR reaction had a total volume of 25 μl, containing 12.5 μl of the 2× Type-it Microsatellite PCR master mix (Qiagen, Hilden, Germany), 2.5 μl of a 10× primer mix, 8 μl of dH 2 O and 2 μl of DNA.The 10× primer mix was composed of each appropriate set of primers used in the 4-plex and 3-plex PCRs, all at a concentration of μM.Each PCR reaction was performed in a single well, a negative (dH 2 0) and a positive control of C. parvum DNA were used with each batch of samples.The PCR cycling conditions were: 5 min at 95 • C, 30 cycles of 30 s at 95 • C, 90 s at 63 • C and 30 s at 72 • C, with a final extension of min at 60 • C. The PCR was then held at 4 • C.

Agarose gel electrophoresis
All secondary 18S and gp60 PCR products were analysed by agarose gel electrophoresis using 1.5% agarose (in 1× TAE buffer) containing Gel Red at 1:10,000 dilution (Biotium, Fremont, CA, USA) with a run time of 90 min at 120 V.The VNTR PCR products were analysed using a 2.5% gel and were run at 70 V for 3 h.For each sample, 8 μl of the amplicon was loaded onto the gel, along with a 100 bp-1.5 Kb DNA molecular weight marker XIV (Roche, Welwyn Garden City, UK) and was then photographed under UV light (GBox).

PCR purification and sequence analysis
Prior to sequencing, PCR amplicons were purified using the Promega Wizard SV gel and PCR clean-up system (Promega, Southampton, UK) according to the manufacturerʼs instructions.The concentration of DNA for each sample was determined using spectrophotometry (NanoDrop One, Thermo Fisher, Waltham, MA, USA).The purified amplicons were stored at 4 • C for the short term and at − 20 • C for long term.
For sequencing, ~100 ng of DNA and 2 μl of appropriate primer (10 mM) in a final volume of 17 μl were sent for Sanger sequencing (MWG Eurofins).A consensus was created for each sample using overlapping  Bartley et al. forward and reverse sequences (DNASTAR program SeqMan Pro version 17; DNASTAR, USA).Cryptosporidium species were confirmed using NCBI (National Centre for Biotechnology Information) BLAST, while the gp60 genotypes were determined by manually counting the number of trinucleotide repeats present in specific conserved regions of sequence (Chalmers et al., 2019).

Fragment size analysis
The (VNTR) multiplex PCR products (~10 μl) were sent to the University of Dundee (DNA Sequencing and Services, UK), for fragment size analysis.The fragment size data were analysed using Geneious Prime software, version 2022 (Biomatters, Auckland, New Zealand).Each allele was identified based on fluorescence, amplicon size and the number of repeat units, which were used to generate the multilocus genotype (MLG) of the sample.The MLG was found for each of the seven fragments, with the multi-locus VNTR analysis (MLVA) profile being completed in chromosomal order, which was cgd1, cgd4, MSF, cgd5, cgd6, cgd8 and MM19.Care was taken when examining the peaks to determine the size of the fragment.Wide and low peaks, stutter peaks and peaks bleeding across fluorescent channels were disregarded as true peaks.If no peak appeared, the corresponding fragment was defined as a null allele (labelled as Ø) and mixed alleles were only identified if the peak height ratio between the two peaks was greater than 0.25.
For each allele, representative samples from both adults and calves were sent for sequence analysis.The numbers of sequence repeats were manually counted, and the information was compared against the size fragment analysis data to ensure the results being reported were accurate (see Supplementary Table S1).

Species of Cryptosporidium found in adult cattle
A total of 224 faecal samples were collected from the adult cattle over the first seven weeks of the calving session, from both pre-partum dams (Pen A, n = 112; mean n = 16; weekly range 12-21) and postpartum dams (Pen B, n = 112; mean n = 19; weekly range 14-24).One cow gave birth while still housed in Pen A. Following the birth, the dam was moved to Pen B and the calf was moved to Pen CA.

Pre-partum cattle
In Pen A, 31.3% (35/112) of the faecal samples tested positive for the presence of Cryptosporidium spp.DNA.Positive samples were seen from Week 2 after the beginning of the calving season (see Table 2), where samples positive for C. parvum (n = 1) and C. bovis (n = 1) were identified.In Week 3, C. andersoni was identified in Pen A in two positive samples.The highest number of positive samples was observed in Pen A during Week 4, where 77.8% (14/18) of the samples were found to contain Cryptosporidium DNA (all C. parvum-positive), while in Weeks 5, 6 and 7, 7/16, 5/16 and 5/21 positive samples were identified, respectively (5 C. parvum-positive samples in each week).
In Pen A, adult cattle were 49 times (odds ratio, OR: 49.00, 95% confidence interval, CI: 4.85-495.23,P = 0.001) more likely to test C. parvum-positive during Week 4 of calving (14/18 C. parvum-positive samples), compared to Week 2 when the first C. parvum-positive sample (1/14) was identified (Table 2).The statistical significance of this finding was confirmed by using Fisherʼs exact test (P < 0.0001).The adults in Pen A continued to have increased odds of shedding C. parvum at Weeks 5, 6 and 7 (OR 6.36, 3.23 and 4.37, respectively) compared to Week 2; however, this risk was not considered statistically significant (P = 0.135, P = 0.346 and P = 0.222, respectively).

Post-partum cattle
In Pen B, 30.4% (34/112) of the faecal samples tested positive for Cryptosporidium spp.DNA, with C. parvum DNA-positive samples being identified on at least one occasion each week where samples were collected (i.e.calving season Weeks 2-7).The highest number of positive samples in Pen B was observed during Week 4 of calving, where 71.4% (10/14) of the samples tested positive (Table 2).
At Week 4 of the calving season, the animals in Pen B had a statistically significantly higher risk of shedding C. parvum, compared to when the first C. parvum-positive samples were identified in this pen at Week 2 (OR: 50.00, 95% CI: 4.92-508.33;Fisherʼs exact test P < 0.0001).Compared to Week 2 (first C. parvum-positive sample), the likelihood of testing C. parvum-positive continued to remain elevated at Weeks 5, 6 and 7 (OR: 9.10, 6.67 and 18.18, respectively).The risk had a borderline significance on Week 5 (P = 0.056) and was significant again on Week 7 (P = 0.009).
There was no significant difference in the likelihood of being C. parvum-positive when comparing pre-and post-natal cattle (Pen A and Pen B) (OR: 1.03, 95% CI: 0.55-1.93,P = 0.922).
Representative sequences of the 18S rRNA gene fragments for C. andersoni and C. bovis were submitted to the GenBank database under the accession numbers OR593321-OR593323.

Clinical observations
Between weeks 5 and 19 of the calving season, evidence of diarrhoea was repeatedly observed in the neonatal animals across most pens.Animals, which showed clear physical evidence of diarrhoea, were generally the youngest animals in each pen and as the calves became older, the incidence of diarrhoea tended to decrease.Along with diarrhoea, the calves also demonstrated other common clinical signs of cryptosporidiosis; these included a general depression in their demeanour along with signs of listlessness and a reluctance to stand.In some severe cases, calves also demonstrated a reluctance to feed, resulting in the animals needing to be bottle-fed by the stockmen.In very severe cases veterinary intervention was required and the calves received fluids intravenously.There was no evidence of clinical cryptosporidiosis amongst the adult animals during the study period.

Rates of calf mortality
The numbers of calves that were born alive or dead and that died before 42 days of age during the 2020-2021, 2021-2022 and 2022-2023 calving seasons are provided in Table 3.The number of calves being born alive each year is relatively consistent (252, 242 and 241 respectively).Although the number of calves being born dead each year were much higher in the 2020-2021 (n = 20) season compared to 2021-2022 and 2022-2023 where only (n = 9 and n = 14, respectively) died at birth, these differences were not significant (Fisherʼs exact test, P = 0.084 and P = 0.479, respectively).When levels of neonatal calf mortality were compared, the highest numbers of calves that died during the first 42 days after birth was recorded during our main study season (2021-2022), with 7.9% (19/242) compared to 3.2% (8/252) during 2020-2021 and 0.8% (2/241) during 2022-2023.These differences in neonatal calf mortality were statistically significant in both comparisons (P = 0.046 and P = 0.0002, respectively).It is also interesting to note that the majority (9/19) of the calf deaths were observed between Weeks 11-15 of the calving season.Although the mean age (in days) when neonatal deaths occurred was comparable across the three seasons being 15.4,16.0 and 12.5 days respectively, the range for the 2021-2022 season was wider and included calves from 2 to 28 days of age, while the range for the other seasons ranged was 7-19 days and 9-16 days, respectively.

Cryptosporidium parvum in calf samples
A total of 312 calf samples were collected and tested across the first 19 weeks of the calving season (see Fig. 1).Positive results with the C. parvum 18S PCR primers were found in 79.5% (248/312) of the samples tested.
In the calves, the first C. parvum-positive samples were detected in Pen D at Week 3 of calving (Fig. 1), where 46.2% (6/13) of the samples tested positive.However, by Week 4 in Pen D, 96.6% (28/29) of the samples were C. parvum-positive, which meant that the calves were 32 times (OR: 32.67, 95% CI: 3.36-317.24)more likely to test positive for Cryptosporidium in the space of a week (Fisherʼs exact test, P = 0.00043).
Samples from Pen CA tested PCR-positive for C. parvum from Week 4 onwards (at least once per week) in every week of the calving season, except for Week 6; however, during this week only one sample was collected from this pen.Every sample collected from Pen CA tested PCRpositive during 6 out of 17 weeks of the study (Fig. 1).Overall, 67.2% (82/122) of the samples collected from Pen CA tested PCR-positive for C. parvum.
Pen CB was used for the first-time during Week 10 of calving, where it immediately tested positive and on all subsequent times thereafter, every sample (100%; 46/46) collected from Pen CB tested PCR-positive for C. parvum.
Pen CC was only used for three weeks between the calving weeks 12-15.Samples from Pen CC tested positive continuously from the first week of use; during these sampling points 91.7% (11/12) of the samples were PCR-positive for C. parvum and this is another example where animals are testing positive directly on being introduced into a pen.
Pens CD and CE were unusual, in that they had not been used to house calves in several calving seasons and were used during this calving season to allow the main neonatal Pens CA and CB to be cleaned and disinfected.Even though no cattle had been housed in Pens CD and CE for several years, 82.6% (14/17) and 73.9% (17/23) of the samples tested PCR-positive for C. parvum, respectively.Both Pens CD and CE had positive samples from the very first sampling point and demonstrated at least one week each where every sample tested positive.
Samples from calves in Pens D, E and G were only analysed for a short period of time, as almost all the samples from each pen tested positive for the presence of C. parvum within one to three weeks of calves being placed in the pens.For Pen D, 96.6% (28/29) of the samples were positive in the third week of use.In Pen E, all (17/17) samples tested PCR-positive on the second week of use and in Pen G every sample collected (23/23) was PCR-positive from its first week of use.
A comparison of the overall percent of positive calf samples between Weeks 1-19 is illustrated in Fig. 2 and shows that the overall percentage of positive calf samples increases rapidly between Week 3 (28.6%)and Week 4 (91.9%) and remains elevated for the remainder of the calving season, meaning a significantly higher proportion of positive calves in Week 4 compared to Week 3 (P = 0.0001) (Fig. 2).From Week 4 onwards an average of 84.1% of calf samples were positive (range 41.2-100%).

Comparison of infection rates with C. parvum in adult cattle and calves
A comparison of the likelihood of being C. parvum infected between the adult cattle and calves, over the first seven weeks of calving (where comparable samples were available) revealed that calves were 10.84 times more likely to be C. parvum-positive (OR: 10.84, 95% CI: 7.23-16.25,P ≤ 0.0001) than adult cattle.
A "N-1" χ 2 test was applied to compare the proportion of positive adult cattle and calves throughout the first seven weeks of the calving period, where comparable data were available.Overall, 23.7% (53/224) of the adult samples tested C. parvum-positive; this is significantly lower (P ≤ 0.0001) compared to 65.4% (72/110) of the calf samples collected during the first seven weeks.

Pre-partum cattle
In Pen A, gp60 genotypes could be determined for 20 of the C. parvum-positive adult samples.One sample from Week 2 had a IIaA15R1 genotype, and the remaining 19 samples, were all identified as IIaA15G2R1.Complete MLVA profiles were generated for 13 out of samples with the IIaA15G2R1 gp60 genotype, all of which had the MLVA profile of 4-14-5-8-18-37-16 (see Supplementary Table S2A).Unfortunately, the IIaA15R1 sample from Pen A failed to produce a complete MLVA profile (Ø-Ø-Ø-8-Ø-Ø-Ø), even after repeated testing (see Supplementary Table S2A).

Post-partum cattle
In Pen B, the first two C. parvum-positive samples identified (Weeks and 3) were IIaA15R1 and IIaA14G2R1, respectively (see Fig. 1).The remaining 16 samples with identifiable gp60 genotypes from Week 4 onwards were all IIaA15G2R1.Examination of the MLVA profiles from Pen B revealed that the IIaA15R1-positive sample had an incomplete profile (4-14-5-8-18-Ø-14) compared to the 4-14-5-8-18-37-16 profile produced by the IIaA15G2R1-positive samples from Weeks 4 and 6 of the calving season.Even with a null peak for the cgd8 allele, the results demonstrated a clear difference at MM19, where 14 repeat sequences were seen, compared to the 16 repeats observed in the IIaA15G2R1.

Genotypic variation of the C. parvum-positive calf samples
For the calves, 125 samples returned readable gp60 genotypes.All the calf samples (125/125) collected from Pens CA-CE, D, E and G were identified as IIaA15G2R1.Additionally, 31 samples (Pens CA-CE, D, E and G) also returned complete MLVA profiles of 4-14-5-8-18-37-16, which was identical to that seen in the adults in Pens A and B. Identical gp60 and MLVA profiles were observed in the calf samples, from all the pens in samples from four weeks of calving onwards (see Supplementary Tables S2C-2J).The remaining calf samples produced incomplete MLVA profiles.Although the profiles of many of the calves were incomplete, no unique genotypes were identified in any of the calf samples.
Representative sequences of the gp60 gene fragments from the different C. parvum genotypes were submitted to the GenBank database under the accession numbers (OR590769-OR590782).

Comparison of efficiency of VNTR PCRs
The efficiency of each of the VNTR PCR primer pairs was compared using a combination of C. parvum-positive adult and calf samples (n = 123) that had a readable result for at least one allele.The most efficient VNTR PCR primers amplified the locus MSF, which had positive results for 65.9% (81/123) of the samples tested.The least efficient PCR primers amplified cdg4 with only 64 out of 123 (52.0%) positive results (Fig. 3).The remaining PCRs for cdg1, cdg5, cdg6 cdg8 and MM19 were comparable in their efficiency (ranging between 53.7 and 58.5%).When the efficiencies of the MSF and cdg4 PCR were compared using a Fisherʼs exact test, MSF was found to be significantly more efficient than cdg4 (P = 0.038).No other statistical differences were observed when comparing the other PCR. .parvum gp60 genotype IIaA15G2R1 first identified (this is the only gp60 genotype identified from that point onwards).Abbreviations: AP, animals in pen, but samples were not collected; E, pen empty (no animals); NSF, animals in pen, but no samples were found; SC, samples collected but not analysed.

Discussion
To our knowledge, this is the first longitudinal study of dairy cattle to examine faecal samples from adult cattle before (pre-partum) and at the beginning of a calving season (post-partum), as well as sampling neonatal calves from very early stages of the calving season for the presence of Cryptosporidium.The results from this study clearly demonstrate that infections with Cryptosporidium spp.are common among both pre-and post-partum cattle and that infections with C. parvum are highly prevalent amongst neonatal dairy calves, with infections occurring at a very early age.
In this study, 30.8% (69/224) of the samples collected from the adult cattle tested positive for Cryptosporidium spp.Cryptosporidium parvum was identified in 76.8% (53/69) of the positive samples, either as a single infection, or in co-infections with both C. andersoni and C. bovis.These findings are consistent with previous longitudinal studies on this farm where 27.3% of the adult samples tested Cryptosporidium-positive (Thomson et al., 2019b).Also, in another study C. parvum was the most commonly recognised species of Cryptosporidium in adult dairy cattle in Belgium (70%) and the Netherlands (75%) (Pinto et al., 2021).Cryptosporidiosis is known to be more of an issue with dairy cattle than beef cattle.Previous studies have demonstrated that dairy cattle (particularly calves) posed a greater risk to human health than beef cattle/calves (OʼHandley, 2007;Dixon et al., 2011).This is likely a farm management issue, with dairy cattle and their calves being housed indoors for extended periods of time (i.e. over winter).
Possible C. parvum transmission routes can be inferred from when calves became infected.Calf-to-calf transmission would seem to be one of the major routes of infection, due to the synchronicity of infection between calf pens.The most probable scenario would be that an infected calf in pen CA was transferred to pen D, thus spreading infection.The lack of variation in all the calf pens suggest that the IIaA15G2R1 population of C. parvum was fully established in pen CA very early in the calving season.Once the infection became established, it spread throughout the farm with relative ease.Other transmission methods, which can be proposed, are as follows.First, the diarrhoeic nature of the faeces from infected calves in pen CA could spread oocysts to pens CB-CC due to their proximity to each other.Secondly, C. parvum oocysts from previous calving seasons could have persisted in the pens causing infection.Thirdly, the stockmen themselves could be a route of transmission through contaminated PPE, as they work with both infected and uninfected animals and will most likely get faeces containing C. parvum oocysts onto their overalls.Finally, the adult cattle could be the source of infection.
Where the initial infection in calves comes from at the beginning of the calving season is still unclear, as though both the pre-and postpartum adult cattle tested C. parvum-positive during Week 2 of the calving season, the genotype that was identified (IIaA15R1) was never detected in any of the neonatal calf samples of any age.The only genotype identified in the calves was IIaA15G2R1.The first calf samples to test C. parvum-positive were isolated from Pen D during the third week of the season, where 46.1% (6/13) of the samples were C. parvum-positive.Unfortunately, we were unable to identify the gp60 genotypes of any of these samples even after repeated testing; this may have been due to low numbers of oocysts in the samples at the early stages of calving/ infection.However, it is highly likely that these samples were the same IIaA15G2R1 genotype that was identified in all the other calf samples (n = 7) at the following sampling point three days later and was also highly prevalent amongst the adult cattle during Week 4 of calving.
There are many factors that influence the spread of C. parvum amongst neonatal farm animals.One of importance is the high rates of oocyst shedding in neonatal dairy calves.Several studies have calculated the numbers of oocysts shed by young calves.A study by Nydam et al. (2001) demonstrated that a single naturally infected 6-day-old dairy calf could produce 3.89 × 10 10 C. parvum oocysts by the age of 12 days.Additionally, dairy calves can shed oocysts for 14-21 days (Zambriski et al., 2013), highlighting the potential scale of the problem faced by dairy farmers in trying to control the spread of cryptosporidiosis amongst young animals.The situation is further complicated by the low infectious dose of C. parvum, meaning that infections can become quickly established in calf pens.In experimentally infected neonatal cattle, an estimated 16.6 oocysts were associated with parasite shedding and diarrhoea in 50% of infected calves (Zambriski et al., 2013), while in healthy human volunteers, 30-1000 oocysts (median 132 oocysts) caused infection and oocyst shedding (DuPont et al., 1995) again highlighting the danger that C. parvum poses not only to animals but also to human health.
At the beginning of the calving season, the infecting doses were clearly relatively low, as there were only a few positive samples amongst the young animals over the first few weeks of calving and the animals appeared healthy (non-diarrhoeic).Even those samples that were C. parvum-positive did not return positive gp60 results, which is suggestive of low numbers of oocysts, as gp60 is a single copy gene (Strong et al., 2000), meaning many more oocysts are required to produce a positive PCR result, than are required for amplification of the multi-copy 18S rRNA gene (where each oocyst contains 20 copies) (Le Blancq et al., 1997).However, as the calving season progressed, the number of oocysts in the environment would have increased quickly.An increasing number of oocysts in the environment means not only an increased chance of exposure, but as the infective dose of oocysts increase then it is also likely that the pre-patent period before a calf became diarrhoeic would decrease.The duration each calf shed oocysts for could also increase, as has previously been shown in experimentally infected calves (Zambriski et al., 2013).
The rates of neonatal calf mortality were significantly higher in the 2021-2022 season compared to other seasons.Although we know from the numbers of positive samples that levels of infection were high, we do not know how many of these deaths were as a consequence of cryptosporidiosis or other pathogens and whether the change in dominant genotype present in the calf population also played a role in these deaths.Recent experimental data from lambs clearly demonstrated differences in the pathogenicity of isolates of C. parvum, where the IIaA15G2R1 genotype was more virulent than the IIaA19G2R1 (originally isolated from this study farm), in terms of severity of clinical symptoms and numbers of oocysts shed by infected animals (Bartley et al., 2023).Clinical observations from the present study also show that the animals that were diarrhoeic also generally demonstrated other clinical symptoms associated with an infection with C. parvum.
Historically the study farm demonstrated greater genetic diversity of gp60 amongst the adult cattle as well as a completely different genotype predominating in the calf population.In a previous study, the IIaA19G2R1 genotype was identified in human cases associated with veterinary students handling calves on the farm as well as within the calf population (Gait et al., 2008).In a previous study Thomson et al. (2019a) identified only the IIaA19G2R1 genotype in all calves ≤ 6 weeks of age, during calving seasons over two years.At the same time four different gp60 genotypes, IIaA15R1, IIaA15G2R1, IIaA18G2R1 and IIaA19G2R1, were routinely identified in adult cattle (Thomson et al., 2019b).
During the present study, calf (≤ 6 weeks-old) samples were tested and typed that were collected during the 2019-2020 calving season (two years before the main study).Positives amongst these were identified as the IIaA19G2R1 (4-14-5-7-27-32-19) genotype.Calf (< 6 weeks-old) samples, which were collected during the 2022-2023 calving season (the year after the main study), were identified as IIaA15G2R1, but failed to produce readable VNTR profiles.The historical and the new genotype data collected on this farm indicates that the IIaA19G2R1 genotype that had been stably maintained within the calf population on the farm for over a decade was replaced sometime between 2020 and 2021 with the IIaA15G2R1 genotype, which is still present in the calves in 2023.It is unclear if the IIaA15G2R1 is more pathogenic than the IIaA19G2R1, or if it is somehow just better adapted to calves and therefore more able to establish infections.Recently published data from lambs experimentally challenged with two distinct isolates of C. parvum clearly demonstrated that there are differences in pathogenicity between isolates of C. parvum, which also affected oocyst shedding profiles (Bartley et al., 2023).Another reason for the increased fecundity of the IIaA15G2R1 genotype may be due to increased levels of the viral symbiont Cryptosporidium parvum virus (CPV), which has been hypothesised to be linked to parasite virulence, though this would need to be further investigated (Jenkins et al., 2008).
As we only collected environmental samples and did not sample per rectum, we cannot definitively state how many C. parvum-infected neonatal calves were actively shedding oocysts on the farm.However, this number is likely to be very high, as on 16 separate occasions, every sample collected from an individual calf pen tested positive for C. parvum.Except for Pen D, where a maximum of 96.5% (27/28) of the samples were positive, every other calf pen had at least one occasion where all samples were positive.For some of the pens every sample tested positive for several weeks in a row and for Pen CB every sample collected (46/46) tested positive for C. parvum (see Fig. 1).

Conclusions
This study demonstrated that there were comparable levels of C. parvum infections in pre-and post-partum dairy cattle at the beginning of a calving season.Comparatively, there were significantly higher levels of C. parvum infection in neonatal calves than in the adult cattle.This study also demonstrated that the time taken for calves to become infected decreases as the calving season progresses, with almost all samples testing positive from about week six of calving onwards.After Week 4 of calving, only a single genotype of C. parvum (IIaA15G2R1) was found in the adults, while IIaA15G2R1 was the only genotype of C. parvum identified in the calves.

Fig. 1 .
Fig. 1.Numbers of C. parvum-positive samples in adult cattle and calves throughout the beginning of a calving season.Notes: a C. parvum gp60 genotype IIaA15R1 identified; b C. parvum gp60 genotype IIaA14G2R1 identified; *C.parvum gp60 genotype IIaA15G2R1 first identified (this is the only gp60 genotype identified from that point onwards).Abbreviations: AP, animals in pen, but samples were not collected; E, pen empty (no animals); NSF, animals in pen, but no samples were found; SC, samples collected but not analysed.

Fig. 2 .
Fig. 2. Percentage of C. parvum-positive calf samples over the first 19 weeks of the calving season.

Fig. 3 .
Fig. 3. Efficiency of individual Variable Number Tandem Repeat (VNTR) PCRs.Results from the calf samples, illustrating the percentage of reactions that generated clearly identifiable peaks for each of the VNTR loci.

Table 2
Species identification of Cryptosporidium-positive samples detected in pre-and post-partum adult cattle.
a Samples contained more than one Cryptosporidium spp.P.M.Bartley et al.

Table 3
Number of calves born alive or dead and number of calves that died before 42 days of age.