RESULTS AND DISCUSSION

Figure 1 shows the four collecting sites of A. nancymai on the east bank of the Marañón river, where the O and B phenotype numbers were: (11) = 9 and 2; (22) = 18 and 4; (26) = 21 and 5; (15) = 13 and 2, respectively. No statistical differences were observed in the distribution of B and 0 phenotypes among these sites (X 2 = 3.948; df = 4; P > 0.05) or even when the farthest site (15) was compared with the others (X2 = 3.082; df = 1; P > 0.05).

As is indicated in Table 11, only two phenotypes (B and 0) were observed in the two species by saliva tests. In A. nancymai, 72 specimens were classified as 0, and 21 as B. In A. vociferans, 10 belonged to the 0 group, and 10 to the B group. These results suggest that only two alleles (I0 and I3) were present in these populations, producing two different phenotypes whose frequencies were significantly different (Fisher's test; P = 0.012). The allele frequencies observed were: IO = 0.88 and 0.71; IB = 0.12 and 0.29 in A. nancymai and in A. vociferans, respectively. Hardy-Weinberg equilibrium could not be tested due to the presence of only two phenotypes and two alleles, which results in a number of degrees of freedom equal to zero. Moreover, we could not extend our comparisons because no previous reports on Aotus are available. A review of the literature, however, indicates that the family Cebidae is very heterogeneous, with species showing one, two or even three alleles, though several of these reports are based on very few animals. It is therefore necessary to extend these studies and to include animals from natural populations to learn more precisely how these genes are distributed.

Table 11 - ABO phenotypes, antigen titers and gene frequencies in night monkeys.

* The inhibition titers were converted to logarithms, base two.
1-Two samples without H substance

The H substance was not detected in all saliva samples. Twenty-one samples of A. nancymai did not presented ABH substances and 11 presented only B substance. In A. vociferans 2 samples presented only B substance. So, the criteria used to classify the 0 phenotype was the absence of A and B substances in saliva. Somewhat similar results have been reported for other New World primates (Gengozian, 1964; Wiener et al., 1964). Wiener et al. (1967), when testing specimens of the genus Saguinus showed that saliva samples inhibited the antibody only when the anti-H reagent was highly diluted, suggesting that the H substance was qualitatively or quantitatively different in these species in comparison to man.

Comparison between the substances titers in distinct phenotypes did not show statistically significant differences.

Landsteiner's rule states that if a given antigen is present in one individual its corresponding agglutinin should be absent. This rule is commonly used when testing serum agglutinins to confirm detection in secretions and/or red cells. When testing the available sera from 89 A. nancymai and 13 A. vociferans, we found that all samples presented anti-A, agglutinating both Al and A2 red cells, while some showed very weak anti-B reactions. These results are presented in Table III, showing a high proportion of cases where Landsteiner's rule is not valid: 72% in A. nancymai and 54% in A. vociferans.

Several discrepancies between sera and secretions phenotypes had been described in previous reports testing non-human primates (Wiener et al., 1942; Gengozian, 1964; Moor-Jankowski et al., 1964; Wiener et al., 1966; Wiener et al., 1967 Nakajima et al., 1970; Socha et al., 1972; Downing et al., 1973; Wiener et al., 1974 Froehlich et al., 1977; Socha et al., 1977, 1981; Terao et al., 1981; Dracopoli and Jolly, 1983; Schneider et al. 1985; Harada-Hamel et al., 1988).

Some authors have proposed alternative models to explain these differences. However, an investigation dealing with this subject is necessary involving a more detailed analysis of the nature and stability of the antibodies after collection as we as its monitoring through the time.
To test whether these specimens had an anti-A fraction capable of reacting specifically with Al red cells, we selected 5 serum samples from A. nancymai and 3 from A. vociferans which were strongly reactive with human A red cells. After extensive absorption with A2 erythrocytes, these sera were retested with Al red cells and three samples of A. nancymai produced a positive reaction, suggesting that anti-Al should exist in this species. A similar result was obtained by Schneider et al., (1985) in Cebus apella, leading Us to propose that anti-Al must be present in other species of New World monkeys. Though A. vociferans sera did not produce a positive reaction, -the existence of anti-Al antibody in this species can't be excluded because our results are based on a small number of specimens.

Table III - Distribution of serum agglutinins and salivary phenotypes in night monkeys.

G = 0.167; df = 1; P > 0.05 (A. nancymai)
G = 0.023; df = 1; P > 0.05 (A. vociferans)

The expected number of random antigen and agglutinin association are in parentheses. If results perfectly fitted Landsteiner's rule, all individuals would lie on the diagonal B-anti A/0-anti A anti B. nose in O-anti A were classified as negative anomalous (N = 60 and 5), and those in B-anti A anti B (N = 4 and 2) as positive anomalous.