**3. Results**

348 Environmental Monitoring

Each sample was taken in a clean 10-L polyethylene container from one point in the center of the bodies of water approximately 20 cm under the surface and sent within 2 h to the Laboratório de Genética Molecular e Citogenética (Genetics and Molecular Diagnostic Laboratory) of the Universidade Federal de Goiás, and concentrated according to Silva et al.

Briefly, water samples were pre-filtered in a vacuum filter with qualitative paper filter, a process also called clarification, aiming to remove excessive amounts of organic matter, such as algae, plants, and other organisms, and immediately submitted to microfiltration using a positively nylon membrane with 0.45µm porosity with 47 mm of diameter (Hybond TM-N+, Amersham Pharmacia). The material adsorbed to the membrane was eluted by 5 ml of TE buffer (10 mM Tris-HCl, pH 8.0; 1mM EDTA) and 0.02% Tween-20, aliquoted and stored at -

Aliquots of 10 µL of concentrated material were employed to prepare smears in two series of two slides each using the modified Ziehl-Neelsen-stain technique and the Kinyoun hot staining method, fixed in alcohol 70%, and processed for specific detection of Coccidea (*Cryptosporidium* sp., *Isospora belli*, and *Cyclospora caytanensis*).In order to detect enteral Microsporidia, the modified hot-chromotrope technique was used (Kokoskin et al., 1994). All the slides were analyzed in duplicate using a common optical microscope with a 100x oil

The modified method of Boom et al., (1990) was used to extract the genetic material, based on cationic exchange resin processes, simultaneously with the phenol/chloroform method

The detection of DNA was performed using Nested-PCR, a variation of the polymerase chain reaction (PCR). The literature was searched to find primers flanking site-specific regions of these opportunistic protozoan genomes (Table 1). The Nested-PCR method was applied only to the positive and/or doubtful samples detected by parasitological

Three primer pairs were used: XIAF/XIAR (*Cryptosporidium* sp. and *C. parvum*), flanking a region of approximately 1325 bp; AWA995f/AWA1206R *(Cryptosporidium* sp.*),* amplifying a region of approximately 211 bp; LAX469F/LAX869R (*C. parvum),* amplifying a

A conventional PCR was carried out using primers XIAF/XIAR and two aliquots were taken from the resulting product, one for detection of protozoan genera via Nested-PCR, using primers AWA995f/AWA1206R, (Awad-el-Kariem, 1994) and the other for the

PCR using primers XIAF/XIAR and 28 μL extracted DNA was performed in a final volume of 50 μL with the following reagents: 5.0 μL buffer 10X, 2.0 mM Mg, 200 μM dNTP (dATP, dCTP, dTTP, and dGTP), 0.5 μM of each primer, and 1.25 U Taq DNA polymerase. The reaction conditions were an initial denaturation step for 4 min followed by another denaturation step of 35 cycles of 94°C for 1 min, annealing at 55°C for 45 s, extension at 72°C

**2.2 Sample concentration** 

**2.3 Parasitological analysis** 

**2.4 DNA extraction and amplification** 

chromosomal region of approximately 451 pb.

detection of *C. parvum/C. hominis* using primers LAX469F/LAX869R.

for 1 min, and final extension at 72°C for 7 min (Xiao, et al., 1999).

of Sambrook & Russel (2001).

immersion objective.

methods.

(2010).

20°C.

Among the 72 samples processed, 8.33% (6/72) were positive for the protozoa researched. Using the MERIFLUOR® direct immunofluorescence assay kit, we found six positive

and PCR (Santos et al., 2010).

**Sampling site** 

samples.

Opportunistic Protozoa in Rivers and Lakes: Relevance to Public Health in the Neotropics 351

Fig. 7. *Cryptosporidium parvum* oocyst stained by the modified Ziehl-Neelsen technique (magnitude 100x) and confirmed by the MERIFLUOR® direct immunofluorescence assay kit

**Ziehl-Neelsen Kinyoun Hot-**

MP1 *C. parvum*\* Negative Negative *Giardia* sp. MP2 Negative Negative Negative Negative JL1 Negative Negative Negative *Giardia* sp. JL2 Negative Negative Negative *Giardia* sp.\*\* VB *Cryptosporidium* sp.\* Negative Negative Negative BB Negative Negative Negative Negative MP1: Meia Ponte river, at 16°37'40.94"S latitude and 49°16'13.41"W longitude; MP2: Meia Ponte river at 16°38'22.39"S latitude and 49°15'50.68"W longitude; JL1: João Leite river, at 16°37'40.18"S latitude and 49°14'26.08"W longitude; JL2: João Leite river, at 16°19'37.52"S latitude and 49°13'24.53"W longitude; VB: Vaca Brava Park lake, at 16°42'31.18"S latitude and 49°16'15.67"W longitude; BB: Bosque dos Buritis lake, at 16°40'58.51"S latitude and 49°15'38.35"W longitude. \*Confirmation by PCR; \*\* Two positive

Table 2. Results according to the six sampling sites and the methods used to analyze the 12

samples in each site monitored, in a total of 72 samples (2006/2007)

**Method** 

**chromotrope MERIFLUOR®** 

samples: two at JL2 in September and November, one at JL1 in August, two at MP1 in July, and one at VB in September.

Using the modified Ziehl-Neelsen-stain technique, 2.7% (2/72) samples were positive for Coccidea, and the presence of *Cryptosporidium* sp. was detected in two samples and confirmed by the MERIFLUOR® direct immunofluorescence assay kit Figure 6 shows a *Cryptosporidium* sp. oocyst and Figure 7 displays a *Cryptosporidium parvum* oocyst, which is approximately 5 µm in diameter, whereas *Cryptosporidium hominis* oocyst is approximately 4 µm in diameter.

Fig. 6. *Cryptosporidium* sp. oocyst stained by the modified Ziehl-Neelsen (magnitude 100x)technique and confirmed by the MERIFLUOR® direct immunofluorescence assay kit and PCR (Santos et al., 2010).

Using primers AWA995f/AWA1206R we demonstrated that the samples belonged to the genus *Cryptosporidium* sp., and using primers LAX469F/LAX869R, we showed that just the sample collected in July was identified as *Cryptosporidium parvum*. As we detected only two positive samples for *Cryptosporidium* sp., the molecular detection was processed exclusively for them.

Using the Kinyoun hot staining method and the hot-chromotrope method for the detection of protozoa, no samples were found to be positive. Table 2 shows the results of each test carried out for the six sampling sites. Table 3 presents the frequency of protozoa detected in each sampling site.

samples: two at JL2 in September and November, one at JL1 in August, two at MP1 in July,

Using the modified Ziehl-Neelsen-stain technique, 2.7% (2/72) samples were positive for Coccidea, and the presence of *Cryptosporidium* sp. was detected in two samples and confirmed by the MERIFLUOR® direct immunofluorescence assay kit Figure 6 shows a *Cryptosporidium* sp. oocyst and Figure 7 displays a *Cryptosporidium parvum* oocyst, which is approximately 5 µm in diameter, whereas *Cryptosporidium hominis* oocyst is approximately 4

Fig. 6. *Cryptosporidium* sp. oocyst stained by the modified Ziehl-Neelsen (magnitude 100x)technique and confirmed by the MERIFLUOR® direct immunofluorescence assay kit

Using primers AWA995f/AWA1206R we demonstrated that the samples belonged to the genus *Cryptosporidium* sp., and using primers LAX469F/LAX869R, we showed that just the sample collected in July was identified as *Cryptosporidium parvum*. As we detected only two positive samples for *Cryptosporidium* sp., the molecular detection was processed exclusively

Using the Kinyoun hot staining method and the hot-chromotrope method for the detection of protozoa, no samples were found to be positive. Table 2 shows the results of each test carried out for the six sampling sites. Table 3 presents the frequency of protozoa detected in

and one at VB in September.

and PCR (Santos et al., 2010).

for them.

each sampling site.

µm in diameter.

Fig. 7. *Cryptosporidium parvum* oocyst stained by the modified Ziehl-Neelsen technique (magnitude 100x) and confirmed by the MERIFLUOR® direct immunofluorescence assay kit and PCR (Santos et al., 2010).


MP1: Meia Ponte river, at 16°37'40.94"S latitude and 49°16'13.41"W longitude; MP2: Meia Ponte river at 16°38'22.39"S latitude and 49°15'50.68"W longitude; JL1: João Leite river, at 16°37'40.18"S latitude and 49°14'26.08"W longitude; JL2: João Leite river, at 16°19'37.52"S latitude and 49°13'24.53"W longitude; VB: Vaca Brava Park lake, at 16°42'31.18"S latitude and 49°16'15.67"W longitude; BB: Bosque dos Buritis lake, at 16°40'58.51"S latitude and 49°15'38.35"W longitude. \*Confirmation by PCR; \*\* Two positive samples.

Table 2. Results according to the six sampling sites and the methods used to analyze the 12 samples in each site monitored, in a total of 72 samples (2006/2007)

Opportunistic Protozoa in Rivers and Lakes: Relevance to Public Health in the Neotropics 353

sampling sites were: clandestine sewage discharges, livestock and poultry farms,

Nonetheless, we detected low recovery efficiency of opportunistic protozoa cysts and/or oocysts, which might be related to environmental influence and physical-chemical factors, such as water pH and turbidity, among others, since the influence of physical-chemical factors on sampling was reported by other researchers (Fricker & Crabb, 1998, McCuin & Clancy, 2003). The influence of physical-chemical factors on sampling was reported by other researchers (Fricker et al., 1998; Clency et al., 2003). Adverse environmental factors have been proven to alter the morphology of cysts and oocysts (Orgerth & Stibbs, 1987) , ; thus justifying the low positivity found in the present study using parasitological methods. Other factors might have had influence as well, such as the concentration of *Cryptosporidium* sp. oocysts, based almost exclusively on particle size (Fricker, 1998). The parasitological techniques employed in our study are not specific and, consequently, concentrate a large amount of several materials that may be present in the water, such as organic and inorganic

particles, bacteria, yeast, and algae, which interfere in the detection of the parasites.

employed by technicians trained to monitor water for human consumption.

*Cryptospridium* sp., nor the American one (McCuin & Clancy, 2003).

method similar to the one used in this study.

However, the methods used in the present study are in accordance with those recommended for concentration and detection of microorganisms by the Standard Methods for the Examination of Water and Wastewater (Clesceri et al., 1998). They are easily applied, do not pose a great risk to the technician, and are low cost techniques, which can be

Hall and Croll (1997) evaluated the performance of some rapid gravity filters in England using turbidity measurement and particle counts in filtered water as parameters for monitoring and controlling *Cryptosporidium* sp. oocysts as an indicator microorganism, a

Some studies have demonstrated that *Cryptosporidium* sp. prevalence is approximately 6% in developed countries (6), around 2-6% in immunodepressed adults (Goldman & Ausiello 2004), and shows a great variation in underdeveloped countries (Casemore, 1990). In industrialized countries, the seroprevalence of oocyst antigens is between 17% and 32% (Goldman & Ausiello 2004). In Canada, a study showed that 21% of the water samples collected were contaminated with *Giardia* sp. cysts and 4.5% with *Cryptosporidium* sp. oocysts (Wallis, 1996). However, in the United States, the contamination of 65% to 97% of superficial water with *Cryptosporidium* sp. oocysts and *Giardia* sp. cysts was reported (Kirkpatrick & Green, 1985), and it was also estimated that 80% of superficial water and 26% of treated water contains oocysts, although their infectivity has not been investigated (Goldman & Ausiello 2004). Nevertheless, we found contamination of 8.33% (6/72) of the samples in the present study, much inferior to the American data, which might be explained by the method applied. Therefore, new methodologies should be tested in order to compare the results in terms of specificity and efficiency to be employed in environmental monitoring of protozoa of public health interest. Since our sampling points are located before the municipal wastewater treatment plant of Goiânia, the results of this study were considered within the tolerable levels, due to the low protozoan positivity according to the method used, in spite of the clandestine sewage discharges. It is worth mentioning that the water from all sources analyzed in this research is improper for usage *in natura*, because it meets neither the Brazilian standard (Brasil, 2004), which establishes that water for human consumption ought to be free from *Giardia* sp. and

The parasitological techniques employed in our study are not specific and, consequently, concentrate a large amount of several materials that may be present in the water, such as organic and inorganic particles, bacteria, yeast, and algae, which interfere in the detection of

slaughterhouses, meat processing plants, landfills, among others.


MP1: Meia Ponte river, at 16°37'40.94"S latitude and 49°16'13.41"W longitude; MP2: Meia Ponte river at 16°38'22.39"S latitude and 49°15'50.68"W longitude; JL1: João Leite river, at 16°37'40.18"S latitude and 49°14'26.08"W longitude; JL2: João Leite river, at 16°19'37.52"S latitude and 49°13'24.53"W longitude; VB: Vaca Brava Park lake, at 16°42'31.18"S latitude and 49°16'15.67"W longitude; BB: Bosque dos Buritis lake, at 16°40'58.51"S latitude and 49°15'38.35"W longitude.

Table 3. General distribution of samples in the six sites according to the presence of protozoa, from February 2006 to January 2007

Average temperature in the period of protozoa occurrence was 26.8ºC, while in the period showing no register of this pathogen, it was 25.6ºC. The logistic regression analysis for temperature revealed p = 0.262 and OR = 1.227 (Table 4).

Average relative humidity in the period of protozoa occurrence was 42.3%, whereas in the period showing no register of this pathogen, it was 56.3%, a not significant value since the logistic regression analysis for relative humidity revealed p = 0.060 and OR = 0.944 (Table 4).


Table 4. Mean and standard deviation of temperature and relative humidity according to the presence of protozoa in the bodies of water sampled in Goiania during February 2006 to January 2007

(logistic regression analysis)

#### **4. Discussion**

This study revealed that the water in all sampling sites monitored during the research is not suitable for human consumption. Despite this evidence, we could observe the presence of people collecting water for human consumption, bathing, washing clothes, and even fishing. This fact is highly worrying because various waterborne diseases, not only related to opportunistic protozoa, but also to several other biological agents, can be transmitted through these contaminated bodies of water. Some sources of pollution observed in the

Negative 12 100.0 10 83.4 11 91,7 10 83,3 11 91.7 12 100.0

sp. 0 0.0 0 83.4 0 0,0 0 0,0 1 8.3 0 0.0 *C. parvum* 0 0.0 1 8.3 0 0,0 0 0,0 0 0.0 0 0.0 *Giardia lamblia* 0 0.0 1 8.3 1 8,3 2 16,7 0 0.0 0 0.0 Total 12 100.0 12 100.0 12 100,0 12 100,0 12 100.0 12 100.0

MP1: Meia Ponte river, at 16°37'40.94"S latitude and 49°16'13.41"W longitude; MP2: Meia Ponte river at 16°38'22.39"S latitude and 49°15'50.68"W longitude; JL1: João Leite river, at 16°37'40.18"S latitude and 49°14'26.08"W longitude; JL2: João Leite river, at 16°19'37.52"S latitude and 49°13'24.53"W longitude; VB: Vaca Brava Park lake, at 16°42'31.18"S latitude and 49°16'15.67"W longitude; BB: Bosque dos Buritis

Average temperature in the period of protozoa occurrence was 26.8ºC, while in the period showing no register of this pathogen, it was 25.6ºC. The logistic regression analysis for

Average relative humidity in the period of protozoa occurrence was 42.3%, whereas in the period showing no register of this pathogen, it was 56.3%, a not significant value since the logistic regression analysis for relative humidity revealed p = 0.060 and OR = 0.944 (Table 4).

> **Protozoa n Mean Standard deviation p OR Temperature**

Positive 6 26.8 1.5 0.262 1.227 **Relative humidity** 

Positive 6 42.3 14.6 0.060 0.944 Table 4. Mean and standard deviation of temperature and relative humidity according to the presence of protozoa in the bodies of water sampled in Goiania during February 2006 to

This study revealed that the water in all sampling sites monitored during the research is not suitable for human consumption. Despite this evidence, we could observe the presence of people collecting water for human consumption, bathing, washing clothes, and even fishing. This fact is highly worrying because various waterborne diseases, not only related to opportunistic protozoa, but also to several other biological agents, can be transmitted through these contaminated bodies of water. Some sources of pollution observed in the

Table 3. General distribution of samples in the six sites according to the presence of

**Sampling site MP1 MP2 JL1 JL2 VB BB n % n % n % n % n % n %** 

Protozoa

*Cryptosporidium* 

January 2007

**4. Discussion** 

(logistic regression analysis)

lake, at 16°40'58.51"S latitude and 49°15'38.35"W longitude.

temperature revealed p = 0.262 and OR = 1.227 (Table 4).

Negative 66 25.6 2.5

Negative 66 56.3 16.0

protozoa, from February 2006 to January 2007

sampling sites were: clandestine sewage discharges, livestock and poultry farms, slaughterhouses, meat processing plants, landfills, among others.

Nonetheless, we detected low recovery efficiency of opportunistic protozoa cysts and/or oocysts, which might be related to environmental influence and physical-chemical factors, such as water pH and turbidity, among others, since the influence of physical-chemical factors on sampling was reported by other researchers (Fricker & Crabb, 1998, McCuin & Clancy, 2003). The influence of physical-chemical factors on sampling was reported by other researchers (Fricker et al., 1998; Clency et al., 2003). Adverse environmental factors have been proven to alter the morphology of cysts and oocysts (Orgerth & Stibbs, 1987) , ; thus justifying the low positivity found in the present study using parasitological methods. Other factors might have had influence as well, such as the concentration of *Cryptosporidium* sp. oocysts, based almost exclusively on particle size (Fricker, 1998). The parasitological techniques employed in our study are not specific and, consequently, concentrate a large amount of several materials that may be present in the water, such as organic and inorganic particles, bacteria, yeast, and algae, which interfere in the detection of the parasites.

However, the methods used in the present study are in accordance with those recommended for concentration and detection of microorganisms by the Standard Methods for the Examination of Water and Wastewater (Clesceri et al., 1998). They are easily applied, do not pose a great risk to the technician, and are low cost techniques, which can be employed by technicians trained to monitor water for human consumption.

Hall and Croll (1997) evaluated the performance of some rapid gravity filters in England using turbidity measurement and particle counts in filtered water as parameters for monitoring and controlling *Cryptosporidium* sp. oocysts as an indicator microorganism, a method similar to the one used in this study.

Some studies have demonstrated that *Cryptosporidium* sp. prevalence is approximately 6% in developed countries (6), around 2-6% in immunodepressed adults (Goldman & Ausiello 2004), and shows a great variation in underdeveloped countries (Casemore, 1990). In industrialized countries, the seroprevalence of oocyst antigens is between 17% and 32% (Goldman & Ausiello 2004). In Canada, a study showed that 21% of the water samples collected were contaminated with *Giardia* sp. cysts and 4.5% with *Cryptosporidium* sp. oocysts (Wallis, 1996). However, in the United States, the contamination of 65% to 97% of superficial water with *Cryptosporidium* sp. oocysts and *Giardia* sp. cysts was reported (Kirkpatrick & Green, 1985), and it was also estimated that 80% of superficial water and 26% of treated water contains oocysts, although their infectivity has not been investigated (Goldman & Ausiello 2004). Nevertheless, we found contamination of 8.33% (6/72) of the samples in the present study, much inferior to the American data, which might be explained by the method applied. Therefore, new methodologies should be tested in order to compare the results in terms of specificity and efficiency to be employed in environmental monitoring of protozoa of public health interest.

Since our sampling points are located before the municipal wastewater treatment plant of Goiânia, the results of this study were considered within the tolerable levels, due to the low protozoan positivity according to the method used, in spite of the clandestine sewage discharges. It is worth mentioning that the water from all sources analyzed in this research is improper for usage *in natura*, because it meets neither the Brazilian standard (Brasil, 2004), which establishes that water for human consumption ought to be free from *Giardia* sp. and *Cryptospridium* sp., nor the American one (McCuin & Clancy, 2003).

The parasitological techniques employed in our study are not specific and, consequently, concentrate a large amount of several materials that may be present in the water, such as organic and inorganic particles, bacteria, yeast, and algae, which interfere in the detection of

Nested-PCR, respectively

**6. Concluding remarks** 

Cryptosporidium limited.

several studies should be adopted.

immunosuppressed patients.

**7. References** 

Opportunistic Protozoa in Rivers and Lakes: Relevance to Public Health in the Neotropics 355

 During seasonal monitoring of opportunistic protozoa, with emphasis on Coccidia *Cryptosporidium* sp., *Cyclospora cayetanensis, Isospora belli* and Microsporidia, it was possible to detectic *Cryptosporidium parvum* and *Cryptosporidium* sp. using PCR and

 The parasitological and molecular techniques applied are quick, low-cost, and can be employed in laboratories that monitor the microbiological quality of water for human consumption. Considering that the microorganisms studied herein are opportunistic, their persistent contact with humans may generate new parasites able to breach the immune barrier of normal individuals and to produce more aggressive cycles. Our results point to the need for efficient programs to prevent, treat, and monitor the presence of these parasites in rivers and lakes used for abstraction of water intended for human consumption and/or for recreational purposes all over the world. Furthermore, more efficient parasitological techniques, such as PCR, should be adopted in routine analyses in the laboratories of environmental monitoring, water for human consumption should be purified with UV radiation, and the activated sludge generated by wastewater treatment

*Cryptosporidium* is considered a coccidia resistant (Carey et al. 2004), because oocysts have characteristics that favor its rapid spread in the environment, such as the ability to withstand the action of commonly used disinfectants (formaldehyde, phenol, ethanol, lysol), able to cross some water filtration systems due to its small size, the ability to float, remain in the environment by a few weeks or months and tolerance in certain temperatures and salinity (Fayer et al. 2004). Given the scope of the aquatic environment coupled with the wide distribution of different species in Brazilian waters, make the control measures of

Therefore, to minimize the risks inherent in the spread of cryptosporidiosis in the populations of free-living mammals, it is of fundamental importance to environmental control, through the adoption of agricultural practices to prevent pollution of rivers by the faeces of animals (Graczyk et al. 2000), as well as encouraging the adequacy of sanitation facilities, protection of water sources, education and guidance on waste discharges from vessels during nautical activities. Regarding the control measures of captive aquatic mammals, so as to minimize or eliminate the risks inherent in the spread of coccidian,

Finally, it must be remembered that currently monitoring systems treated water are based on the frequency of fecal coliforms and Escherichia coli as indicators of pollution, and that this methodology is insufficient to predict the presence of other pathogens such as parasites. Thus, it is imperative the use of alternative methods for the diagnosis, investigation and monitoring of large amounts of water of these pathogens. For in this way can be proposed reorganization measures that contribute to reducing the incidence of opportunistic diseases emerging in water of human use, especially for children, elderly, immunocompromised and

Alves, T. M., & Chaveiro, E.F. Metamorfose urbana: a conurbação Goiânia-Goianira e suas

implicacões sócio-espaciais [Urban metamorphosis: the conurbation of Goiânia and

plants and intended for use in agriculture should be monitored.

the parasites. However, the methods used in this study are in accordance with those recommended for concentration and detection of microorganisms by the *Standard Methods for the Examination of Water and Wastewater* (Clesceri, 1998). They are easily applied, do not pose a great risk to the technician, and are low cost techniques, which can be employed by technicians trained to monitor water for human consumption.

The performance of some rapid gravity filters was evaluated in England, using turbidity measurement and particle counts in filtered water as parameters for monitoring and controlling *Cryptosporidium* sp. oocysts as an indicator microorganism (Geldreich, 1996), a method similar to the one used in our study.

The *in vitro* amplification of DNA fragments of *Cryptosporidium* sp. obtained sensibility and specificity. Nevertheless, the amplification was only possible using Nested-PCR primers (AWA995f/AWA1206R and LAX469F/LAX869R). The primer LAX469F/LAX869R amplifies the regions of *C. parvum*/*C. hominis*, but *C. parvum* diagnosis was confirmed by the difference in diameter, since its oocyst is approximately 5 µm in diameter, while *C. hominis* oocyst is approximately 4 µm in diameter.

Nested-PCR presents the advantage of concentrating a smaller quantity of PCR inhibitors (Kirkpatrick & Green, 1985). In environmental samples, there are several Taq DNA polymerase inhibitors, such as fecal hemoglobin and phenolic compounds, and it might have been the case of the samples processed in the present research.

It was possible to obtain satisfactory amplification with the two methods of DNA extraction applied. Furthermore, they are quick and low-cost, although close attention should be paid to the phenol/chloroform method since it is toxic and corrosive.

As adverse environmental factors have been proven to alter the morphology of cysts and oocysts (Hsu BM, 2001), making their detection more difficult, this may justify the low positivity found in the present study using parasitological methods. Other factors might have had influence as well, such as the concentration of *Cryptosporidium* sp. oocysts, based almost exclusively on particle size (Hsu, 2001). Also, the level of protozoa may vary according to the season, and an increase in their resistant forms in rainy periods, winter and beginning of spring has already been reported (Atherholt 1998, Ong et al. 2002)

Temperature has also been considered a factor that influences protozoa and autochthonous microorganism survival in rivers (Howe, 2002). In this study, we observed just small variations of water temperature in the rivers and lakes sampled during the period of study, although within the limits that allow the survival and viability of protozoa. Using univariate logistic regression (p = 0.066), we demonstrated that temperature was not a statistically significant variable, whereas humidity (p = 0.958) was. In the region of sample collection there are two well-defined seasons, the dry (from April to September) and the rainy (from October to March) seasons, the latter characterized by torrential rain and runoff, which certainly makes the detection of parasites more difficult.

Due to the low number of protozoa found in this work, i.e. two *Cryptosporidium* sp. and four *Giardia* sp., we could not infer if the protozoan levels vary by season, but only observe the qualitative inference of their presence in the bodies of water monitored.

### **5. Conclusion**

