Results
T. gondii -specific IgG antibodies were found in 59.6% (28/47) of captive NPs and 11.6% (23/198) of wild NPs. In captive NPs, seroprevalence was high for both species: 59.1% (26/44) of A. geoffroyi and 66.7% (2/3) of C. imitator . There were no significant differences in species or sex with 58.6% (17/29) ofA. geoffroyi females and 60.0% (9/15) males being positive, while for C. imitator , 50.0% (1/2) females and 100% (1/1) male had antibodies (Table 3, Table 4).
In contrast, there was a significant difference in seroprevalence among wild NP species (X2 = 20,072; df = 3;P = 0.0002): A. geoffroyi had the highest seroprevalence with 40% (2/5), followed by C. imitator with 30% (11/37) andA. palliata with 6.6% (10/151), while no S. oerstediiwere positive. No significant differences were found regarding sex for any of the species, however, greater seroprevalence was observed for females in the three positive species (Table 3, Table 5).
Antibody titers tended to be relatively low (Median = 8) in wild NP (<128), with the exception of four individuals: one A. geoffroyi and two C. imitator with antibody titers of 262,144 and one A. geoffroyi with a titer of 1,048,576). In contrast, antibody titers were considerably higher (Median = 262144) in captive NP, from 8,192 to 33,554,000 (Table 6). These two groups of NP were statistically different in their titer values (Mann-Whitney U= 303, n1= 52, n2=34, p<0.00001 two tailed).
The model that best accounted for seropositivity among A. palliata (minimizing AIC) was forest cover and annual precipitation (Figure 1.B., Table 2), with a positive relationship between forest cover and seropositivity and inverse relationship with annual precipitation. However, evidence weights (wi) for the second and third best-supported models were 1.02 and 1.94 times lower than model 1 respectively, which suggests that no one model is strongly supported as the best candidate. The second best model was annual precipitation (ΔAIC = 0.04) and the third was for human population density and annual precipitation (ΔAIC = 1.32). Notably, the precipitation parameter appeared among the top eight models.
For C. imitator, annual precipitation alone best accounted for seropositivity with a significant inverse relationship (Figure 1.C., Table 2). The next best models were forest cover and annual precipitation (ΔAIC = 1.09) and human population density and precipitation (ΔAIC = 1.5), with evidence weights 1.73 and 2.12 times lower than model 1, respectively. Again, the effect of precipitation on seropositivity was important for C. imitator when comparing the sum of the weights of Akaike (wi) in the models that contained this variable (0.98) with respect to the others, such as forest cover (0.35), population density (0.3) and sex (0.29).