Molecular Detection of Cyclospora cayetanensis in Two Main Types of Farm Soil Using Real-Time PCR Assays and Method Modification for Commercial Potting Mix

Cyclospora cayetanensis is a foodborne protozoan parasite that causes outbreaks of diarrheal illness (cyclosporiasis) with clear seasonality worldwide. In the environment, C. cayetanensis oocysts are very robust, and contact with contaminated soil may serve as an important vehicle in the transmission of this organism, and it is considered a risk factor for this infection. The present study evaluated a flotation concentration method, previously shown to provide the best detection results when compared with DNA isolation directly from soil samples, in two main types of farm soil, silt loam soil and sandy clay loam, as well as in commercial potting mix samples inoculated with different numbers of C. cayetanensis oocysts. The flotation method was able to detect as few as 10 oocysts in 10 g of either type of farm soil without modifications, but needed an extra wash and samples of reduced size for the processing of the commercial potting mix to be able to detect 20 oocysts/5 g. A recently modified real-time PCR method for the detection of C. cayetanensis based on a mitochondrial gene target was also evaluated using selected samples of each type of soil. This comparative study confirmed that the concentration of oocysts in soil samples by flotation in high-density sucrose solutions is a sensitive method that can detect low numbers of oocysts in different types of soil.


Introduction
Cyclospora cayetanensis is an important foodborne protozoan parasite responsible for sporadic cases and outbreaks of intestinal disease with clear seasonality worldwide [1][2][3]. In the United States, the parasite has been reported to have caused clinical cases and outbreaks since the mid-1990s. In 2022, there was a total of 1129 laboratory-confirmed domestically acquired cases from 33 states, with 74 of the people affected needing hospitalization [4].
Fresh produce, often consumed raw and with little or no washing, has been the main type of commodity implicated in C. cayetanensis outbreaks worldwide [3,4]. Fresh produce may become contaminated with oocysts due to poor hygiene practices by food handlers, contaminated soil, and/or contaminated agricultural water [5].
Unsporulated C. cayetanensis oocysts are excreted by infected humans in feces. Sporulation occurs in the environment. Although there are some studies on the experimental conditions in which C. cayetanensis oocysts survive in the laboratory [6,7], unfortunately, there is no information on what triggers oocyst sporulation in field conditions. As is the case with other protozoa, C. cayetanensis oocysts are thought to be environmentally resistant; low numbers of oocysts with a heterogenous distribution are expected to be present in soil

Cyclospora cayetanensis Oocysts used and Types of Soil Analyzed
The oocysts used in the experiments were purified from individual human stool samples from Guatemala and were stored in 2.5% potassium dichromate, as described elsewhere [8]. The study was approved by the institutional review board of the FDA (protocol number 15-039F). The oocysts in the sample used in the present study were found to be unsporulated via microscopy. To enable comparability, the same preparation of oocysts was used for each seeding experiment in the different types of soil samples in the present study.
Silt loam soil was collected from a farm in Ohio (OH). The composition of this soil was analyzed and found to have a high percentage of silt (65%), a medium percentage of clay (21%), and a low percentage of sand (14%). The sandy clay loam soil was collected from a farm in Georgia (GA), and after analysis it showed a composition of 54% sand, 24% clay, and 23% silt. Similar levels of clay were observed in the two types of soil. Produce grown in the soil collected from OH included, among other crops, berries and leafy greens. The produce grown in the soil collected from GA included mainly leafy greens.

Protocol for Seeding, Flotation Method, and DNA Extraction in Sandy Clay and Silt Loam Soils
The sandy clay and silt loam soil samples were autoclaved prior to the experiments to avoid bacterial and fungal growth [8]. Samples of both types of soil (sandy clay loam and silt loam) weighing 10 g were weighted and aliquoted into 50 mL centrifuge tubes and seeded with different levels of oocysts: 5 samples were seeded with 100 C. cayetanensis oocysts, 10 samples were seeded with 20 oocysts, 10 samples were seeded with 10 oocysts, and 5 soil samples were processed unseeded to serve as a negative control, as previously reported [8]. Concentration of the C. cayetanensis oocysts was then performed via flotation in high-density sucrose solutions [8]. A total of 30 samples were processed for each type of soil. Briefly, C. cayetanensis-seeded and -unseeded control soil samples (10 g) were dispersed by adding 40 mL of deionized water to the 50 mL tube containing the 10 g of soil sample and well mixed manually (1 min). The tubes were then centrifuged at 2000× g for 20 min, after which the supernatants were removed. The pellet was mixed with cold sucrose solution (specific gravity (SG): 1.12) until completely mixed. Additional cold sucrose solution was added to fill the tube, and after mixing again, the tubes were centrifuged again at 2000× g for 20 min. Supernatant containing the top of the solution (20 mL) was collected into a new 50 mL tube and 30 mL of deionized water was added, followed by centrifugation at 2000× g for 20 min. The sediment was retained and washed with deionized water in a 15 mL tube. After centrifugation at 2000× g for 20 min, the pellet was transferred to a Fastprep TM (MP Biomedicals, Santa Ana, CA, USA) tube and centrifuged at 14,000× g for 4 min, after which the supernatant was removed and the pellet was kept at −20 • C until DNA isolation was carried out. DNA extraction was performed on the two different types of soil using a commercial kit (FastDNA Spin Kit for Soil from MP Bio) in conjunction with a FastPrep-24 Instrument (MP Biomedicals, Santa Ana, CA, USA), following the previously outlined methodology [8]. The original elution volume was 75 µL/sample.

Protocol Modifications for Processing (Seeding, Flotation Method, and DNA Extraction) in Commercial Potting Mix
Samples from the commercial potting mix were not autoclaved prior to the experiments. Modified approaches were needed to optimize the recovery and detection of the oocysts from the commercial potting mix samples. The top layer of the commercial potting mix containing peat moss, wood chips, and perlite greatly increased the difficulty of the oocyst collection during the flotation step. Therefore, the extra particulate material was removed by carefully pouring off the excess material after centrifugation at 2000× g for 20 min in the first wash step of the original protocol, before the sucrose solution flotation step was performed. Once the top layer was removed, the flotation protocol was performed as previously indicated by Shipley et al. [8].
Initially, 16 samples (10 g) seeded with 50 oocysts each were analyzed using two different DNA extraction kits: (1) a FastDNA Spin Kit for Soil from MP Bio and (2) a ZymoBIOMICS DNA Miniprep Kit from Zymo Research (Irvine, CA, USA). Similar rates of positive samples and average Cq values were detected using the two DNA kits. High average Cq values and a low percentage of positive samples were detected when the 10 g samples were seeded with 50 C. cayetanensis oocysts. Therefore, a smaller sample of commercial potting mix was analyzed (5 g) to compare the levels of detection. Reducing the potting mix sample size to 5 g enabled better detection when the samples were seeded with the same number of oocysts (50 oocysts). To confirm the detection limit in the potting mix, 10 additional 5 g samples were then seeded with 20 oocysts. After the DNA was extracted from the farm soil and the commercial potting mix, real-time PCR was performed as per Shipley et al. [8]. Briefly, the molecular detection of C. cayetanensis was performed via a duplex reaction targeting both the C. cayetanensis multicopy 18S ribosomal RNA gene (Cyclo250F C. cayetanensis TAGTAACCGAACGGATCG-CATT, Cyclo350RN AATGCCACGGTAGGCCAATA, and Cyclo281T C. cayetanensis 6-FAM-CCGGCGATAGATCATTCAAGTTTCTGACC) and an exogenous internal amplification control (IAC) (dd-IAC-f IAC CTAACCTTCGTGATGAGCAATCG, dd-IAC-r GATCAGCTACGT-GAGGTCCTAC, and dd-IAC-Cy5 IAC Cy5-AGCTAGTCGATGCACTCCAGTCCTCCT) using the QuantiFast Multiplex PCR Kit as the master mix [26]. The qPCR was performed on an Applied Biosystems 7500 Fast Real-Time PCR System (ThermoFisher Scientific, Waltham, MA, USA). A commercially prepared synthetic gBlocks gene fragment (Integrated DNA Technologies, Coralville, CA, USA) (HMgBlock135m) was used as a positive control for the amplification of the C. cayetanensis 18S rRNA gene [26,27], and a non-template control (NTC) was included in each run. The amplification protocol consisted of an initial step of 95 • C for 5 min followed by 45 cycles of 95 • C for 30 s and 67 • C for 30 s. Runs were only considered valid if all three replicates of the positive control reactions produced the expected positive result and the NTC was negative. Samples were only considered positive when one or more of the replicates produced a positive result with a cycle threshold (Cq) ≤ 38.0 for the 18S target. According to this method, undetermined reactions would be considered inconclusive if the IAC reaction failed or produced an average Cq value more than three cycles higher than that of the NTC for the same assay, or if the IAC reaction failed completely [26,27].

Quantitative Real-Time PCR for Farm Soil and Commercial Potting Mix Samples Using Cox3 (MIt1C qPCR) as Target
A second recently validated (SLV) method for a multi-copy specific target in the mitochondrial genome of C. cayetanensis (Mit1C assay) [28] was evaluated in both soils using the lower oocyst levels for comparison with 18S rRNA. The new TaqMan realtime PCR duplex assay targets both the new mitochondrial C. cayetanensis gene (Mit1Cf TCTATTTTCACCATTCTTGCTCAC, Mit1C-r TGGACTTACTAGGGTGGAGTCT, and Mit1C-P 6-FAM-AGGAGATAGAATGCTGGTGTATGCACC) and the same exogenous IAC target as described in Section 2.4.1. using PerfeCTa DNA polymerase and multiplex qPCR Toughmix Low ROX reaction cocktail (Quantabio, Beverly, MA, USA) as the master mix [28]. The mitochondrial target in the Cox3 gene was identified in silico using BLAST searches against C. cayetanensis and other genera/species (e.g., Eimeria spp. and Isospora spp.) in the Apicomplexa phylum. The target was a 205 bp region with a 100% consensus to all reported C. cayetanensis sequences in the NCBI database, being species-specific. The amplification protocol consisted of an initial step of 95 • C for 3 min, followed by 40 cycles of 95 • C for 15 s and 61 • C for 1 min, with a cycle threshold (Cq) ≤ 38.0 as a cut off [28]. We tested the new method on selected DNA samples previously analyzed using the 18S rRNA. The samples analyzed were the samples seeded with the lowest levels of oocysts, i.e., silty and sandy clay loam seeded with 10 oocysts from and commercial potting mix seeded with 20 oocysts. Results obtained using 18S rRNA and Mit1C as targets were compared.

Sequencing of Selected Samples
Selected qPCR positive amplicons from samples of soil spiked with 100 C. cayetanensis oocysts (2 samples, 6 positive replicates), 20 C. cayetanensis oocysts (2 samples, 6 positive replicates), and 10 C. cayetanensis oocysts (1 sample, 2 positive replicates) were combined according to seeding level and sequenced to confirm that C. cayetanensis was detected in the presence of the soil background in the samples. The amplicons for sequencing were generated by performing the Mit1C qPCR on these samples without the IAC to avoid the interference of the IAC probe and the primers on the sequencing. The qPCR products were purified using the QIAquick PCR purification Kit (Qiagen, Germantown, MD, USA), following the manufacturer's instructions. The purified amplicons were sequenced on both strands (Psomagen Inc., Rockville, MD, USA). The sequencing data were then compared with the NCBI nucleotide BLAST suite to confirm the specificity of the amplicons for the detection of C. cayetanensis in the soil background.

Statistical Analysis
Positive detection rates were obtained by calculating the percentage of inoculated samples which gave a positive result in the samples processed via each method and/or experiment. Statistically significant differences in Cq values between two different soil detection levels or between two methods used on the same samples were analyzed using a t-test. Statistically significant differences in Cq values between more than 2 soil types were analyzed via one-way analysis of variance and multiple comparisons using Tukey's multiple comparison test.
Contingency tables for the number of positive samples (percentages) in each type of soil were compared using Fisher's exact test (two sided). Statistical analyses were performed using GraphPad version 9.1 (GraphPad, San Diego, CA, USA), with a p value of ≤0.05 indicating statistical significance. A trend towards statistical significance was considered when the p value was ≥0.05 but ≤0.1.
When samples were undetermined in the qPCR reaction, the negative samples were excluded from the calculation of the average Cq values. The flotation method followed by an 18S rRNA qPCR was able to detect as few as 10 oocysts in 10 g samples of silt loam soil as well as in sandy clay loam samples (Tables 1 and 2). All of the negative control samples were negative, and no inhibition was observed in any of the processed samples. While the IAC values varied among the samples, they did not diverge from the IAC values of the NTC by 3 Cq values ( Table 1). The positivity rates of the silt loam samples were 80% (n = 10), 100% (n = 10), and 100% (n = 5) in the samples seeded with 10 oocysts, 20 oocysts, and 100 oocysts, respectively. In the sandy clay loam soil samples, the respective positivity rates were 20%, 50%, and 80% (Table 2). Table 1. qPCR detection data (number of positive qPCR replicates, individual Cq values) for 18S rRNA C. cayetanensis detection in individual samples of sandy clay loam and silt loam soil (10 g), both unseeded and seeded with different numbers of C. cayetanensis oocysts (10, 20, and 100) using the flotation in dense sucrose solution method.  Statistically insignificant differences in the average Cq values for C. cayetanensis 18S rRNA were observed between the two types of soil at each of the seeding levels, although a trend towards statistical significance (p < 0.1) was observed in the samples seeded with 100 oocysts (35.3 in silt loam versus 36.11 in sandy clay loam, Table 1, p = 0.052).
A significantly higher number of samples seeded with 10 oocysts were detected in the silt loam soil (80% of the 10 samples) compared with the sandy clay loam soil (20% of the 10 samples, p = 0.02), and a significantly higher number of samples seeded with 20 oocysts were also detected in the silt loam soil (100% of the 10 samples) compared with the sandy clay loam soil (50% of the 10 samples, p = 0.032). In the samples seeded with 100 oocysts, the rate of positive detection was higher in the silt loam (100% of the 5 samples) than in the sandy clay loam (80% of the 5 samples), but the differences were not statistically significant, showing only a trend towards statistical significance (p = 0.06). The new Mit1C target qPCR was analyzed using samples of both types of soil seeded with 10 oocysts (limit of detection). For each set of samples seeded (10) in each type of soil, an unseeded soil sample was included in the analysis as a negative control. Among the silt loam samples, a higher number of positive samples seeded with 10 oocysts were detected using the Mit1C qPCR (100% of the 10 samples) than were detected using the 18S rRNA qPCR (80% of the 10 samples) ( Table 2), but the difference was not statistically significant (p = 0.47). However, statistically significant differences in average Cq values were observed between both targets (37.6 for the 18S rRNA qPCR versus 34.5 for the Mit1C qPCR, p = 0.0005).
Similarly, among the sandy clay loam samples, a higher number of positive samples seeded with 10 oocysts were detected using the Mit1C qPCR (60% of the 10 samples) than were detected using the 18S rRNA (20% of the 10 samples) ( Table 2), but difference was not statistically significant (p = 0.17). Statistically significant differences were observed in in the average Cq values of both qPCR target methods (36.9 for the 18S versus 32.3 for the Mit1C, p = 0.003).
The average Cq values of the samples of silt loam and sandy clay loam seeded with 10 oocysts analyzed using the Mit1C were statistically significant (34.5 in the silty loam samples versus 32.3 in the sandy clay loam samples, p = 0.007).   To determine the fractional level of detection and the limit of detection of C. cayetanensis in commercial potting mix samples, 10 samples were seeded with 20 C. cayetanensis oocysts. The positive detection rate for the samples seeded with 20 oocysts was 20% when using the 18S rRNA qPCR, and therefore 20 oocysts was the detection limit for this type of soil using the modified flotation in sucrose method.

Method Modification and Detection Levels of C. cayetanensis DNA in Commercial Potting Mix Using Mit1C qPCR
Samples seeded with 20 oocysts (the lowest level of detection) were also analyzed using the Mit1C qPCR. An unseeded potting mix sample was included in the analysis as a negative control, and it was found to be undetermined when using the Mit1C. A small increase in the rate of positive detection was observed in the commercial potting mix samples seeded with 20 oocysts using the Mit1C qPCR (30% versus 20% detection rate using the 18S rRNA method, p > 0.05), and significantly decreased average Cq values were observed when using the Mit1C method (an average Cq value of 35.2 when using the Mit1C and an average Cq value of 37.2 when using the 18S rRNA qPCR, p = 0.015, Table 3). To determine the fractional level of detection and the limit of detection of C. cayetanensis in commercial potting mix samples, 10 samples were seeded with 20 C. cayetanensis oocysts. The positive detection rate for the samples seeded with 20 oocysts was 20% when using the 18S rRNA qPCR, and therefore 20 oocysts was the detection limit for this type of soil using the modified flotation in sucrose method.

Method Modification and Detection Levels of C. cayetanensis DNA in Commercial Potting Mix Using Mit1C qPCR
Samples seeded with 20 oocysts (the lowest level of detection) were also analyzed using the Mit1C qPCR. An unseeded potting mix sample was included in the analysis as a negative control, and it was found to be undetermined when using the Mit1C. A small increase in the rate of positive detection was observed in the commercial potting mix samples seeded with 20 oocysts using the Mit1C qPCR (30% versus 20% detection rate using the 18S rRNA method, p > 0.05), and significantly decreased average Cq values were observed when using the Mit1C method (an average Cq value of 35.2 when using the Mit1C and an average Cq value of 37.2 when using the 18S rRNA qPCR, p = 0.015, Table 3). Table 3. Average results for detection of C. cayetanensis using 18S RNA and Mit1C qPCRs in commercial potting mix (5 g) seeded with different numbers of oocysts of the parasite.

Sequencing of qPCR Amplicons from Samples of Soil Seeded with C. cayetanensis Oocysts
Sequencing results from the soil samples seeded with 100 C. cayetanensis oocysts showed a partial sequence with a 100% alignment to the reference C. cayetanensis mitochondrial genome KP231180 (between 4417bp and 4540bp of the genome). No successful sequencing was achieved using the soil samples seeded with 10 oocysts and 20 oocysts.

Discussion
The transmission of C. cayetanensis oocysts from contaminated soil is still not well understood. Methods for the detection of the parasite in soil are needed, and such methods need to be shown to work in different types of soil. To our knowledge, there are no previous studies comparing methods for the detection of C. cayetanensis in different types of soil. The present study evaluated a previously existing sensitive procedure for the detection of C. cayetanensis in large soil samples (5-10 g) [8] using two main types of farm soil, silt loam soil and sandy clay loam soil, as well as a commercial potting mix. Sandy clay loam and silt loam are probably the most common types of farm soil found in the USA in areas where fresh produce linked to C. cayetanensis infections are grown. Produce grown in the soil collected in OH included berries and leafy greens, as well as other produce, indicating that vegetable and fruit crops will thrive in silty soils if they have adequate drainage. A commercial potting mix was included in the evaluation of the method since it could be used to grow fresh produce (herbs and berries) in climate-controlled growth chambers or a controlled lab environment and thereby enable us to monitor how detection is affected by time and environmental factors such as temperature, humidity, and/or photoperiod.
There are still only a few studies concerning the detection of C. cayetanensis in soil [14,[29][30][31], and to our knowledge, only two of them were carried out using farm soil [27,29], which shows relatively high percentages of positive samples. In the first of these studies, which involved berry farm soils in Mexico, C. cayetanensis was detected in 20% (1/5) of soil samples analyzed using nested PCR [31]. Although the soil samples collected at the base of the plants weighed approximately 50 g, only around 250 mg of each soil sample was used for direct DNA extraction and PCR analysis. In the second study, conducted in Apulia, Italy, soil samples were collected from different vegetable/fruit crops planted in succession (cucumber, lettuce, fennel, celery, tomato, melon, endive, and chicory) [29]. Upon the harvesting of each vegetable crop, pooled soil samples (1 kg) were collected from under the water dripper from each plot, and 10 g soil samples were then processed using Percoll-sucrose gradients, commercial genomic DNA isolation kits, and qPCR and melting curve analyses. Cyclospora DNA was detected in 11.8% of the 51 soil samples, with the number of Cyclospora oocysts estimated (using qPCR) to be from 230 to 661 per 5 µL of genomic DNA. In both studies, the composition of the collected farm soil was not indicated. A recent study involving indigenous people in Venezuela [14] collected soil samples from around huts housing Cyclospora cases (2 soil samples from each of the 25 huts) rather than farm soil. The soil was analyzed using UV epifluorescence and phase-contrast microscopy, and 18% (9/50) of the soil specimens were found to be positive for Cyclospora sp. [14]. The authors indicated that all the positive soil specimens had a clay texture. Because Cyclospora infections predominated in the months of higher rainfall, the authors suggested that mean annual rainfall and the consequent moisture in the soil was the epidemiologically significant factor in the survival of the oocysts and the maintenance of their density in the soil of this semiarid region [14]. In the present study, both of the farm soils analyzed had similar percentages of clay.
The method/procedure used in the present study was able to detect as few as 10 oocysts/10 g of sample in silt loam soil, as was found in a previous study [6]. Similarly, the silt loam soil analyzed in the present study was collected from the same area and had a similar composition to that used in [6], and the same limit of detection was observed.
On the other hand, sandy clay loam with a very different composition also showed the same limit of detection (10 oocysts/10 g of soil). Therefore, the method was consistent in the detection of small numbers of C. cayetanensis oocysts in the main types of farm soil, and no inhibition was observed in any of the samples analyzed, regardless of soil type, a finding consistent with the previous study [8]. However, important differences in the detection of the parasite were observed between the two types of farm soil analyzed. Significant differences were not observed in the average Cq values of the two types of farm soil, although a trend towards statistical significance (p < 0.1) was observed in the samples seeded with 100 oocysts. The similarity of the average Cq values indicate that, when detected, the parasite was found in similar levels/numbers in the positive silty loam and sandy clay loam samples. However, lower rates of positive samples were observed in the sandy clay loam compared with the silty loam. The sandy clay loam showed lower rates of positive samples at all seeding levels (10 oocysts, 20 oocysts, and 100 oocysts), with statistically significant differences in the lower seeding levels (10 oocysts and 20 oocysts) and a trend toward lower rates at 100 oocysts compared with those of the silt loam samples. Although there were no previous comparison studies on the detection of C. cayetanensis in different types of soil, our results confirm previous findings concerning the differences in soil composition as they relate to the detection of protozoan parasites using flotation in soil samples [9,20]. Previous studies on Toxoplasma gondii detection in soil reported lower detection/recovery of the parasite in soil with high proportions of sand [9,21,22]. However, in parasites with larger parasitic forms, such as some Taeniid species, Toxocara canis, or soil-transmitted helminths eggs, including those of A. lumbricoides and Trichuris trichiura helminths, higher detection was observed in soil rich in sand [20,23,24]. Therefore, the parasite wall structure and specific gravity of each parasitic form needs to be considered.
When detection using the flotation method was analyzed in a commercial potting mix, the method needed modifications to enable the elimination of large particles of the potting mix before the proper flotation procedure could be carried out. Potting mix is made up of many less dense components than are found in normal soil, including peat moss, wood chips, and perlite, and these float even in aqueous solutions. These components produced an additional large layer in the supernatant which required removal before the sucrose solution flotation step. In this type of soil substrate, a reduction in the size of the samples, from 10 g to 5 g, significantly increased the detection of C. cayetanensis DNA. Therefore, 5 g samples are recommended when processing commercial potting mix. Two commercial kits for DNA extraction after flotation were compared (the MP Bio FastDNA Spin Kit for Soil and the ZymoBIOMICS DNA Miniprep Kit) in this type of soil substrate and were found to have similar detection levels (both in terms of percentage of positive rates and Cq values), but the ZymoBIOMICS DNA Miniprep Kit requires less time to complete the DNA extraction. No inhibition was observed in any of the processed potting mix samples, and after modifications as few as 20 oocysts/5 g of potting mix were detected after the flotation procedure. The ability to detect C. cayetanensis oocysts in potting mix will facilitate studies of C. cayetanensis detection in growth chambers mimicking environmental field conditions in the laboratory, and this in turn will increase our knowledge of the epidemiology of this important parasite.
Recently, a mitochondrial target for the detection of C. cayetanensis using real-time PCR was validated in produce [28]. In the present study, this new target, Cox3, was assayed in the sandy clay loam, silt loam, and potting mix samples with the lowest seeding levels and compared with C. cayetanensis 18s RNA real-time PCR detection in the same samples to assess sensitivity. Although the rates of positive samples detected were not significantly higher when using the Mit1C qPCR, significantly lower Cq values were observed when using the Mit1C qPCR. In addition, the amplicon of the Mit1C (205 bp) could enable the sequencing of the qPCR fragment (without IAC) to confirm the positivity of the samples. In the present study, the sequencing of positive qPCR samples confirmed the presence of C. cayetanensis in the presence of soil background in a sample seeded with 100 oocysts, although we were not able to do so at low seeding levels (10 oocysts and 20 oocysts). Therefore, the use of the second mitochondrial marker could be useful for the detection and confirmation of the presence of C. cayetanensis in soil.

Conclusions
In conclusion, a method involving flotation in saturated sucrose solution was evaluated using different types of farm soil and showed the same limit of detection in sandy clay loam and silt loam soil, although the sandy clay loam had lower rates of positive detection than the silt loam soil. Method modifications were needed for the commercial potting mix. The modifications evaluated in the present study enabled the detection of small amounts of oocyst DNA in the different types of soil analyzed, and the method could be used to confirm whether soil is a source of C. cayetanensis contamination in foods. Since each soil sample can vary in its composition, and since sand content clearly plays a role in the detection of C. cayetanensis oocysts, our results should be interpreted with caution. The ability to detect C. cayetanensis oocysts in different types of soil will facilitate outbreak investigations and increase our knowledge of the epidemiology of this important parasite. Future investigations could use the present protocol to study the effects of environmental climatic factors in controlled experiments in soil inoculated with C. cayetanensis.