Blood Groups and H-Lea Salivary Secretion of Brazilian Cayapo Indians F. M . SALZANO, H . GERSHOWITZ, P. C. JUNQUEIRA, J . P. WOODALL, F. L. BLACK A N D W. HIERHOLZER Departamrnto d e Geneticu, Universidade Federal d o Rio Grande d o S u l , PBrto Alegre, Brazil; Department of Human Genetics, Unibersity of Michigan, A n n Arbor, Michigan, Cosa d r Saude Santu Lticia, Rzo d e Juneiro, Guannbara, Brazil; u n d Department of Epidemiology a n d Public Health, Y a l e University, New H a v e n , Connecticut KEY W O R D S Blood groups . Brazil Indians . Genetic taxonomy Multidisciplinary studies . Salivary secretion. . ABSTRACT Five hundred and twenty-six individuals from four populations were studied in relation to the ABO, MNSs, P, Rh, Lutheran, Kell, Lewis, Duffy, Kidd, Diego and I systems, as well as for the Wright antigen of blood groups. The H-Lea salivary secretion of 406 of them was also investigated. Considering the gene markers which show variation in South American Indians, the Cayapo frequencies are in the middle of the distribution range for genes LMS , R1 ( C D e ) , R2 (cDE), PI, Jka, Di" and Se but present high values of Le and Fya and low ones of L M S , LNS , Ro ( c D e ) and/or T ( c d e ) , LNS and R* (CDE). Unusual findings i n relation to previous studies are the high prevalence of Le (a ) persons (which, however, could be expected since the frequency of gene Se is not too high) and the presence of one Lu (a ) and one PPI (Tja)( - ) individuals. Comparison of the pattern of inter-village variation in relation to these polymorphisms with those furnished by historical, demographic and morphological data gives in general reasonable concordance, but some results are difficult to interpret. New approaches and further multidisciplinary studies are needed to obtain a clearer picture of the genetic relationships present among different tribes of South American Indians. as well as to understand how polymorphisms are created and maintained in human populations. + + Although as many as 43,500 South American Indians have already been studied in relation to the ABO system, a series of questions are still unanswered concerning the distribution of blood group polymorphisms in these populations. Two of the most general ones are: in what ways does the range of gene frequency variation differ in this major subgroup of our species, as opposed to those observed in other races? and How can this variability be related to their population structure, disease patterns and cultural characteristics? Important problems like the average degree of heterozygosity at several loci, and interaction between systems remain untouched. The present communication is the fourth in a series of the Brazilian Cayapo Indians, previous papers having dealt with the AM. J. PHYS.ANTHROP.,36: 4 1 7 4 2 6 . health status of one group (Nutels, Ayres and Salzano, '67) as well as the demography (Salzano, '71) and anthropometric variation (Da Rocha and Salzano, '72) of three of them. They are related to a longterm project aiming at the elucidation of the questions posed in the first paragraph (Salzano, '61a; Neel and Salzano, '66, '67; Neel, '71). MATERIALS A N D METHODS The Cayapo Indians speak a Ge language and live at present in eight still quite isolated, semi-independent communities in the Brazilian States of Para and Mato Grosso. Details about their history, exact location and demographic characteristics can be found in Salzano, '71. Four populations have been studied in 417 418 SALZANO E T AL. relation to their blood groups; brief information about them follows. 1. KubenKran-Kegn, location: lat. 8 " lo's., long. 52'8'W.; population: 310; pacified in 1952; blood collected i n 1965 and 1968; 2. Txuhahamae, lat. 1Oo20'S., long. 53" 5'W.; population: 190; pacified: 1953; blood collection: 1966; 3. Mehranoti, lat. 8"40'S., long, 54" W.; population: 240; pacified: 1958; blood collection: 1969; 4. Xihrin, lat. 5"55'S., long. 51'11'W.; population: 150; pacified: 1954; blood collection: 1970. The most acculturated are the Xikrin followed by the Kuben-KranKegn, while the Txukahamae and Mekranoti are the least disturbed by Neo-Brazilian influences. Blood was collected in 10 or 15 ml Becton-Dickinson vacutainers with EDTA or ACD; every effort was made to obtain samples from all individuals living in the villages at the time, with the exception of the very young (less than 2 years of age). Specimens were refrigerated shortly after collection and carried by air to the laboratory. Those obtained during the first visit to the Kuben-Kran-Kegn were tested in P8rto Alegre and Rio de Janeiro; all other determinations were performed in P6rto Alegre andlor Ann Arbor. The blood typing reagents with which tests were performed in duplicate included anti-A, A,, B, M, M g , M Y , N, S, s , PI, C (rh'), C" (rh"), c (hr'), D (Rho), E (rh"), e (hr"), K, k, Kp", Fya, Fyb, Di", Wra. Tests with the following antigens were done only once: Porto Alegre: anti-I; Rio: anti-Kpb, J P , Jkb; Ann Arbor: anti-U, V', Mi", M u + H t , PPI (Tja), f (hr), Lu", Lub, Kpb, Jsa, Lea, Leb, Jk", Jkb, Di". All studies were carried out in tubes with 2% washed red cell suspensions. Discrepancies in the duplicate determinations were reconciled by repeat typings on cells preserved in glycerol a t - 2 0 ° C or in liquid nitrogen. The salivas were directly collected into glass tubes and boiled immediately for 15 to 30 minutes. They were transported under refrigeration and kept frozen in the laboratory until tested. The determination of the H and Le" secretor status was done in duplicate in Parto Alegre and Ann Arbor. Gene frequencies related to the MNSs blood groups were calculated according to Mourant's ('54) method. In the Rh system a gene counting procedure was adopted, assuming the absence of r y ( C d E ) , r' ( c d E ) and r" (Cde) (details about the assumptions and limitations of this and other methods can be found in Gershowitz et al., '70). The other gene frequencies were obtained by standard gene counting or square root calculations. In all systems in which it was ,possible to identify the heterozygote and both homozygotes we verified if their numbers were those expected assuming Hardy-Weinberg equilibrium. The deviations from the theoretical values were generally very small and in only one case (the Duffy distribution among the Txukahamae) reached the 5% level of statistical significance. Due to the number of comparisons made this result may have occurred by chance alone. Restrictions can be presented to the use of ~2 methods and other "large sample" statistics in these populations (Nee1 et al., '64; Gershowitz et al., '67). Therefore, no attempts were made to verify statistically the significance of intertribal variations. RESULTS Table 1 shows the MNSs blood group phenotypes observed in the four populations and table 2 the corresponding gene frequencies, which are compared with those of two other Ge groups and of South American Indians in general. The largest amount of intratribal variation occurred with genes LMS and LNS.The Txukahamae and Mekranoti show very similar gene frequencies in relation to the four genetic markers. Gene LNS is not present in the Kuben-Kran-Kegn sample. As for the intertribal comparisons, the Cayapo do not show much departure from the prevalences observed i n the two other Ge tribes for the LNS and L N s genes but the frequency of LMS is higher and that of LMS lower than those found among the Caingang. Taking into consideration the whole range of South American Indian values, the Cayapo frequencies are in the upper half of the distribution in relation to gene LMs but in the lower third for the other markers. Phenotype prevalences for the Rh system are given in table 3 and the respective gene frequencies in table 4. There is no clear pattern in the intratribal dis- 419 BLOOD GROUPS AND SECRETION OF BRAZILIAN INDIANS TABLE 1 M N S s blood group phenotypes observed in four Cayapo Indian populations Phenotype Txukahamae Mekranoti Kuben-Kran-Kegn Xikrin Total ~~~~ no. MS MSs Ms MNS MNSs MNs NS NSs Ns Total 10 32 59 2 22 30 1 4 2 162 % no. % no. % no. % 6 20 36 1 14 19 1 2 1 1 28 31 1 9 14 1 31 34 1 10 10 37 35 6 8 31 7 28 26 4 6 23 2 17 20 1 18 25 2 18 22 1 20 27 1 8 92 - - - 2 5 91 15 _ - 8 135' 2 6 - _ _ 6 - - 1 9 ~~ no. 23 114 145 10 57 100 1 7 23 480 % 4.8 23.7 30.2 2.1 11.9 20.8 0.2 1.5 4.8 - I Forty-six bloods were tested with anti-M. -N, and - S only, yielding the following results: MS ( 10;MS(-): 11;MNS(+): 9;MNS(-): l I ; N S ( + ) : 2 ; N S ( - ) : 3 . + ); TABLE 2 M N S s blood group gene frequencies observed in four Cayapo Indian populations compared with those ofother Ge groups and South American Indians in general Gene frequencies Population Txukahamae Mekranoti Kuben-Kran-Kegn Xikrin Total Cayapo Xavante ( N = 537) Caingang ( N = 76) * S. American Indians Range(N=11,517)3 0.18 0.18 0.28 0.23 0.220 0.366 0.545 0.61 0.61 0.49 0.43 0.541 0.408 0.129 0.08 0.06 0.00 0.03 0.036 0.098 0.065 0.13 0.15 0.23 0.31 0.203 0.128 0.261 0.000 0.857 0.093 0.944 0.000 0.321 0.000 0.685 Nee1 et al., '64;Gershowitz et al., '67. Fernandes et al., '57. 3 Data from 97 surveys performed by different investigators. 1 2 tribution and it is curious that the Txukahamae and Mekranoti who showed similar gene frequencies in the MNSs blood group now present very different prevalence figures. Since no cDe (Rho) phenotypes were observed and at least one Indian proved to be rr (cdelcde), the gene frequency calculations were performed assuming the absence of Ro ( c D e ) and presence of r ( c d e ) . It is possible, however, that gene R" ( c D e ) may have been present. The Cayapo values are again similar to those observed in other Ge groups; the prevalence of R" (CDE) is very low but those of R' ( C D e ) and R2 ( c D E ) are in the middle of the range given by previous investigations in South American Indians. The results concerning the P, Duffy, Kidd, Diego and Lewis blood groups, as well as the H and Lewis secretor saliva phenotypes are given in tables 5 and 6. Previous results with gene P1 show an enormous range of variation, going all the way from 9 to 100%. We have reasons to suppose, however, that at least part of this variability is related to technical problems. The anti-PI reagents are somewhat irregular in their degree of association with the quantitative antigenic differences which occur in this system. For most of the tests we have used a potent antiserum prepared in Ann Arbor and the prevalences of PI( ) thus found were much higher than the one obtained in 46 bloods collected during the first expedition to the Kuben-Kran-Kegn and tested with commercial antisera. There exists little variation among the Cayapo villages in relation to this gene marker; the total Cayapo frequency is similar to the Xavante value, but the Caingang present a much lower prevalence. + 420 SALZANO ET AL. TABLE 3 R h blood g r o u p p h e n o t y p e s obserued in four C a y a p o Indian pop7tlations Phenotype Txukahamae CDe(RhlRhl) CcDe ( R h , r h ) cDE (RhzRh?) cDEe ( R h g h ) CcDEe ( R h l R h z ) CDEe ( R h z R h l ) CCDE (RhzRhs) CDE (RhzRhz) ce ( r h ) Total Mekranoti Kuben-Kran-Kegn no. % no. % 710. 50 10 20 6 60 8 8 31 6 12 4 37 5 5 14 4 36 4 30 16 4 40 4 33 1 1 38 8 34 12 79 4 4 _ _ _ 162 _ - -1 _ 1 1 1 91 - 1 1 181 76 21 4 19 7 43 2 2 1 1 - Xikrin 1 95 32 1 19 3 40 4 1 _ _ 92 - no. 29 1 17 3 37 4 Total >lo. 131 23 107 25 206 17 14 1 2 526 0 , 24.9 4.4 20.3 4.7 39.2 3.2 2.7 0.2 0.4 - TABLE 4 R h blood g r o u p gene f r e q u e n c i e s observed in f o u r C a y a p o I n d i a n p o p u l a t i o n s c o m p a r e d w i t h those of other Ge grozips unrl S o u t h A m e n c a n Indians i?i general Gene frequencies Population Txukahamae Mekranoti Ku ben- Kran-Kegn Xikrin Total C a y a p o X a v a n t e (N = 539) 1 C a i n g a n g ( N =243) 2 S . A m e r i c a n Indians R a n g e ( N = 12,826) 3 RL (CDe) R? (cDE) R Z (CDE) Ro(cDe) r (cde) 0.54 0.35 0.45 0.54 0.475 0.588 0.529 0.34 0.59 0.44 0.41 0.429 0.329 0.344 0.06 0.01 0.04 0.03 0.039 0.044 0.056 - 0.06 0.05 0.07 0.02 0.057 0.040 0.071 _ 0.123 0.958 0.025 0.732 0.023 0.308 0.000 0.207 0.000 0.233 Nee1 et al., '64; Gershowitz et al., '67. Salzano, '61b, '64a. Data from 104 surveys made by different investigators; in 94 of them there was no indication of the occurrence of gene r (cde). The evidence for the presence of this gene among the Cayapo rests in only one r r (cdeicde) individual whose parents are putatively "pure blood." The other Rh negative result was obtained from the blood of a White captured in h i s infancy and who has adopted the Indians' way of living, having married and procreated in the village. 1 2 3 The Fy" frequencies also have a wide range in South American Indians. There is more inter-village variability here, the Kuben-Kran-Kegn and Xikrin showing, however, similar results. Considering the Cayapo as a whole, they present a higher frequency i n relation to this gene than the Xavante or Caingang. As for Jh", the Xikrin show a somewhat low incidence; the Cayapo average is very similar to the Xavante one. The Di" gene frequencies separate the four villages into two groups: Txukahamae/Mekranoti and Kuben-KranKegn/Xikrin. There is a gradation of values among the Ge, from a low 0.17 found in the Xavante reaching 0.21 i n the Cayapo and 0.34 in the Caingang. Gene Se shows little variation in frequency among villages. The Cayapo present a value similar to the Caingang, which is two times higher than the Xavante frequency. Gene Le is practically invariant among the three villages from which data are available, the Cayapo now showing a frequency more similar to the Xavante than to the Caingang; the latter present a very low incidence of Le. Comparing the Cayapo results on these six gene markers with the range of gene frequencies observed in South American Indians, we see that they generally are in the middle of the observed distributions. In accordance with their somewhat low Se frequency (0.71 -table 6 ) the Cayapo show one of the highest prevalences of Le (a + ) individuals found so far in South American Indians (8% -table 5 ; surpassed only by the Xavante, who show a n unusual figure of 30%). The frequencies of Le ( a - b + ) and Le ( a - b - ) ob- 421 BLOOD GROUPS AND SECRETION OF BRAZILIAN INDIANS TABLE 5 P , Duffy, Kidd, Diego, a n d Lewis blood group, H a n d Lewis secretor phenotypes observed in f o u r Cayapo Indian populations Phenotype Txukahamae Mekranoti Kuben-Kran-Kegn Xikrin no. no. no. no. 127 162 76 91 112 135 1 Fy (a b - ) Fy (a b +) Fy(a-b+) Fy(a-b-) Total Fy (a + I ) * Total Grand Total 46 84 15 1 146 16 16 162 - Jk(a+b-) Jk (a b + ) Jk (a - b + ) Jk (a - b - ) Total Jk (a + ) Total Grand Total - - Pl(+) Total + + + Di(a+b-) Di (a b +) Di (a- b +) Total Di ( a +) 3 Total Grand Total + L e (a - b + ) L e (a + b - ) Le (a - b - ) Total H.Sec. Total L e a Sec. Total - - 89 91 91 - - Tot a1 no. % 76 92 391 480 82 81 42 12 49 36 1 48.0 44.1 7.6 0.3 135 45 46 181 86 6 6 92 176 162 28 1 367 156 159 526 21 40 31 3 95 128 180 180 - 22.1 42.1 32.6 3.2 47 86 86 21 40 31 3 95 360 518 518 - 11 59 107 177 198 515 515 6 33 61 - - - - - - 98.1 - - 69.5 118 161 161 67 91 91 - 43 152 152 5 24 62 91 29 91 91 84 180 180 6 35 45 86 42 92 92 109 11 42 162 62 5 24 91 95 10 30 135 52 11 23 86 318 37 119 4 74 67.1 7.8 25.1 136 149 85 91 150 166 - 371 406 91.4 114 148 68 91 125 164 - 307 403 76.2 - - - - 38.4 - - - 1 A significantly lower value (43%) was obtained with other antisera in 46 bloods collected in the first expedition. 2 Tests with anti-Fyb not performed. 3 Individuals with this phenotype in tests performed with one or two antisera. served in their red cells are respectively in the lower third and middle of the distribution found in 48 other surveys conducted by different investigators and involving 6,169 South American Indians. Theoretically it is possible to obtain further estimates of Le using data from these blood surveys. This was not attempted because we are not sure if the relationships between the Lewis blood and saliva phenotypes in these populations are those usually found in Caucasian populations (Gershowitz et al., '67). Table 7 presents the results concerning antigens which showed little or no varia- tion in the populations studied here. They are generally in accordance with investigations performed in other South American Indian groups (table 8); the majority of these antigens also had a limited variation in populations of other ethnic groups as well. The ABO results deserve a special comment. As is indicated in table 7, the AIB individual observed among the Mekranoti is a White man captured in his infancy and who has adopted the Indians' way of living. The one Al and two B individuals found are his children, born from an 0 Indian woman. A purist would be inclined to disregard these individuals, 422 SALZANO ET AL. TABLE 6 P, Duffy, Kidd, Diego, H and Lewis Secretor gene frequencies observed in four Cayapo Indian populations compared w i t h those of other Ge groups and South American Indians in general Gene frequencies Population P' Txukahamae Mekranoti Kuben-Kran-Kegn Xikrin Total Cayapo Xavante ( N = 539) 1 Caingang ( N = 288) 2 South American Indians Range 3 Fy" Se Le 0.70 0.74 0.68 0.50 0.50 0.51 - 0.15 0.19 0.27 0.27 0.215 0.166 0.342 0.707 0.365 0.772 0.512 0.632 0.252 0.1130.723 0.0000.483 0.3651.000 0.2520.632 Di a Jka 0.53 0.60 0.59 0.59 0.570 0.639 0.223 0.61 0.86 0.76 0.78 0.702 0.540 0.541 0.48 0.49 0.46 0.33 0.448 0.421 0.090- 0.131- 1.ooo 1.ooo - - Nee1 et al., '64; Gershowitz et al., '67. Salzano, '61b, '64a,b. The number of individuals studied was not the same for all systems. on data from several authors, a s follows: PI: 98 surveys, 13,499 individuals tested; Fy": 109 and 12,891; /ha: 78 and 9,347; Di": 135 and 14,224; Se: 6 and 1,374; Le: 7 and 1,285. 1 2 3 Based TABLE I Antigensfrom eight blood group systems which showed little or no variation in four Cayapo Indian populations Population System Type ABO 0 A1 U(+)2 VW(-) M"-) Mi" ( - ) MV(-) + P Rh Kell Mekranoti Kuben-Kran-Kegn 162 87 1 2 1 181 - B A1B 1 MNSs Txukahamae Mu Ht ( - ) PPl (Tj") ( + ) 3 PP1 (Tj 9 ( - ) Cw (rhw) ( - ) 4 f (hr) ( - ) 5 f (hr) ( +) K(-) k(+) Kpa(-) Kpb(+) Js"( - ) Lutheran Lu (a - b + ) Lu ( a + b + ) Wright Wrk2(- ) I I(+) 34 49 162 49 162 - - 135 135 181 135 135 135 - 12 1 162 17 3 162 50 162 50 - - - 51 181 91 - 48 1 162 162 Xikrin Total 92 522 1 2 1 169 270 429 184 449 135 24 1 486 52 3 526 271 429 271 22 1 269 1 435 171 86 86 61 12 92 35 - - 181 181 181 181 135 135 92 86 86 86 86 86 181 9 - - 92 1 The person whose blood had this reaction is a White captured in h i s infancy and who has adopted the Indians' way of living. The one A1 and two B individuals are his children born by a n 0 Indian woman, 2 Among the Txukahamae only the S ( - ) bloods were tested for U. 3 Of the 12 PPI (Tj ") ( + ) nine were PI ( - ) and three P1 ( ). 4 Among the Mekranoti only C (rh') ( ) bloods were tested for this antigen. 5 Only CcDEe (RhlRhz) bloods were tested for this antigen. + including only "putatively pure blood' persons. We have not done this for the following reasons: 1. The A I B man is culturally as Indian as any other member + of the tribe, having in fact led punitive expeditions against other White persons; 2. Such events probably occurred repeatedly in the past, involving children of 423 BLOOD GROUPS AND SECRETION OF BRAZILIAN INDIANS TABLE 8 Previous studies in South American Indians with antigens showing limited variation Gene or phenotype System ABO 10 IA IB MNSs UJ+' v (-) VW(+) Mg(-) Mi" ( - ) Mi" ( + ) MV(-> P Rh Kell + MU Ht ( - ) PPI (Tja) ( +) Cw (rhw) ( -) k KP"(-) Kpb ( + ) Js"(-) Lutheran Wright I 1 Js" ( + ) Lu (a - ) Lu (a +) Wr" ( - ) Wr" ( + ) I(+) 1 (-) No. of individuals tested No. of surveys 43,671 43,671 43,671 500 6,150 1 538 5,143 220 220 220 2 48 48 1 47 47 5 282 345 198 1,108 17,329 963 1,623 1,392 8 3,788 6 5,724 3 1,691 6 Range 0.797-1.OOO 0.000-0.203 0.000-0.102 - - - 1 1 2 11 101 3 7 10 10 24 24 27 27 1 1 - 0.952-1.000 - - 1 Pooled data from several authors. other tribes and ethnic groups. The detection of these genes has historical interest but there is no other reason for isolating them from the rest of the gene pool. As a matter of fact, they have to be considered in any effort to understand the factors which affect the genetic variability of these populations. Two somewhat unusual findings recorded in table 7 are the presence of one Lutheran ( a + ) and one PPI (Tj") ( - ) among the Txukahamae. In 24 other surveys conducted in South American Indians involving 3,794 persons, only six were found to be Lu (a +) - four Quechuas from Hacienda Vicos, Peru (Allen, '59), one Paraujano from the Sinamaica lagoon, Venezuela (Layrisse, Layrisse and Wilbert, 'SO), and one Aymara from Santa Fe, Bolivia (Matson, Swanson and Robinson, '66). The usual explanation for these findings is admixture with non-Indian persons. The isolation of the Txukahamae makes this hypothesis unlikely but an event like the capture of the AIB man mentioned above could have occurred in the past. In only two other surveys PPI (Tja) tests have been performed; among 198 Brazilian and Venezuelan Indians all proved to be PPI (Tja) ( + ) (Gershowitz et al., '67, '70). DISCUSSION In what way can the Cayapo frequencies be related to the known distribution of blood group and secretor polymorphisms in South American Indians? For seven of the 14 gene markers studied which show variation in these populations the prevalences are in the middle of the distribution range (LM", RI (CDe), R2 ( c D E ) , PI, Jk", Dia and Se), for two others the values are high (Le: 0.51 and Fya: 0.70) while for five they are low (LMS: 0.22; LNs: 0.20; RD ( c D e ) and/or T ( c d e ) : 0.06; L N S and R Z (CDE): both 0.04). Unusual findings in relation to previous studies are the high prevalence of Le ( a + ) persons (which however could be expected since the frequency of gene Se is not too high) and the presence of one Lu ( a + ) and one PPI (Tj") ( - ) individuals. Comparing now our results with those obtained among two other Ge tribes, we see that in general the Cayapo figures are more similar to the Xavante than to the Caingang. In relation to the former the 424 SALZANO ET AL. main difference occurs at the Se locus, the Cayapo (Ca) showing a prevalence of 0.71 and the Xavante (X) 0.36. Other differences are not as large; the three most marked concern Fya (Ca: 0.70; X: 0.54), LMS (Ca: 0.22; X: 0.37), and L M s (Ca: 0.54; X : 41). The four main differences between the Cayapo and Caingang (Cg) are related to LMS (Ca: 0.54; Cg: 0.13), P' (Ca: 0.57; Cg: 0.22), LMS (Ca: 0.22; Cg: 0.54), and Le (Ca: 0.51; Cg: 0.25). If the pattern of inter-village variation concerning these polymorphisms is related to those revealed by historical, demographic and morphological data, some interesting observations result. On the basis of the available historical information (see Salzano, '71) the Txukahamae and Mekranoti should show the lowest degree of genetic differentiation; dissimilarities between them and the KubenKran-Kegn should be more marked, while the Xikrin, who separated earlier from the others, should present the highest degree of differentiation. Detailed demographic information was not available for the Xikrin but the data from the three other localities showed in fact that the Kuben-Kran-Kegn could be set apart from the other two subgroups. On the other hand, the anthropometric analysis (Da Rocha and Salzano, '72j yielded some curious results. The Kuben-Kran-Kegn and Txukahamae presented the lowest degree of morphological differentiation (as measured by the D2 statistic), the Txukahamae and Mekranoti intermediate and the Kuben-Kran-Kegn and Mekranoti the largest amount of dissimilarity. Inspection of the blood group and secretor results indicate that as expected the Xikrin figures deviate from those observed among the Mekranoti and Txukahamae (especially in relation to genes L M s , LNs, Jh" and Di") but not from those observed among the Kuben-KranKegn (the only difference higher than 10% found among them occurred in relation to the prevalences of gene Jh"). On the basis of blood group and secretor data alone the Kuben-Kran-Kegn cannot be clearly separated from the Mekranoti and Txukahamae. The latter show very similar frequencies in the MNSs and Diego systems but this is not true for the Rh blood groups. The distribution of the frequencies in the 14 genes showing variation in South American Indians and studied here shows ranges of different orders of magnitude, some of them very wide. A constant feature in these distributions is that despite the fact that they have different shapes the values do not show distinct discontinuities. Therefore, despite the vast array of variation found, no classification of the populations studied according to their genetic homology is possible at the moment. Fitch and Nee1 ('69), using three different methods, have studied the relationships of 12 Indian tribes of Central and South America using data from seven polymorphic systems, which were considered simultaneously. Analyses such as these will help in the establishment of a meaningful genetic taxonomy of the South American Indian. But for this kind of approach it is necessary to study a population in relation to many polymorphisms and to have concomitant data on other variables, to check if the associations made are the correct ones. Therefore, at the present pace of investigation, it is unlikely that we will soon have a clear picture of the main pattern of genetic variation among South American Indians on a continental scale. However, multidisciplinary studies aiming at the answer to the questions posed in the introduction, besides furnishing the main source of data for this approach, may also provide some significant clues for the understanding of how polymorphisms are created and maintained in human populations. ACKNOWLEDGMENTS Thanks are due to F. J. da Rocha, G. V. Simoes and M. Ayres for help in the field work and to the FundaGBo Nacional do Indio and Parque Nacional do Xingu which gave permission to study the Indians and provided all facilities. Transportation to and from the villages was made possible by Asas de Socorro. The skillful assistance in the laboratory determinations given by G. V. Simbes, M. Ayres, Margarete S. Mattevi, Maria C. Mallmann and Sidia M. Callegari is gratefully acknowledged. This work is part of the Brazilian contribution to the International Biological Programme. Our researches are BLOOD GROUPS AND SECRETION OF BRAZILIAN INDIANS supported by the Conselho Nacional de Pesquisas, Conselho de Pesquisas da Universidade Federal do Rio Grande do Sul, Coordenaciio do AperfeiGoamento do Pessoal de Nivel Superior, FundaCiio de Amparo a Pesquisa do Estado do Rio Grande do Sul, FundaGPo ServiGos de Saude Publica, Pan American Health Organization, Rockefeller Foundation, U.S. AEC AT (11-1) 1552, and the WennerGren Foundation for Anthropological Research. LITERATURE CITED 1959 Summary of blood Allen, F. H., Jr. group phenotypes in some aboriginal Americans. Am. J. Phys. Anthrop., 17: 86. Da Rocha, F. J., and F. M. Salzano 1972 Anthropometric studies in Brazilian Cayapo Indians. Am. J. Phys. Anthrop., 36: 95-102. Fernandes, J. L., P. C. Junqueira, H. Kalmus, F. Ottensooser, R. Pasqualin and P. Wishart 1957 P.T.C. thresholds, colour vision and blood factors of Brazilian Indians. I. Kaingangs. Ann. Hum. Genet., 22: 16-21. Fitch, W. M., and J. V. Neel 1969 The phylogenic relationships of some Indian tribes of Central and South America. Am. J. Hum. Genet., 21 : 384-397. Gershowitz, H., P. C. Junqueira, F. M. Salzano and J. V. Neel 1967 Further studies on the Xavante Indians. 111. Blood groups and ABHLea secretor types i n the Simbes Lopes and S5o Marcos Xavantes. Am. J. Hum. Genet., 19: 502-513. Gershowitz, H., M. Layrisse, J. V. Neel, C. Brewer, N. Chagnon and M. Ayres 1970 Gene frequencies and microdifferentiation among the Makiritare Indians. I. Eleven blood group systems and the ABH-Le secretor traits: a note on Rh gene frequency determinations. Am. J. Hum. Genet., 22: 515-525. 425 Layrisse, M., 2. Layrisse and J . Wilbert 1960 Blood group antigens among the Paraujano. Am. J. Phys. Anthrop., 18: 131-139. Matson, G. A., J . Swanson and A. Robinson 1966 Distribution of hereditary blood groups among Indians i n South America. 111. In Bolivia. Am. J. Phys. Anthrop., 25: 13-33. Mourant, A. E. 1954 The Distribution of the Human Blood Groups. Charles C Thomas, Springfield. Neel, J. V. 1971 Genetic aspects of the ecology of disease in the American Indian. In: The Ongoing Evolution of Latin American Populations. F. M. Salzano, ed. Charles C Thomas, Springfield, pp. 561-590. Neel, J. V., and F. M. Salzano 1966 A prospectus for genetic studies on the American Indians. In: The Biology of Human Adaptability. P. T. Baker and J. S. Weiner, eds. Clarendon Press, Oxford, pp. 245-274. 1967 Further studies on the Xavante Indians. X. Some hypotheses-generalizations resulting from these studies. Am. J . Hum. Genet., 19: 554-574. Neel, J. V., F. M. Salzano, P. C. Junqueira, F. Keiter and D. Maybury-Lewis 1964 Studies on the Xavante Indians of the Brazilian Mato Grosso. Am. J. Hum. Genet., 16: 52-140. Nutels, N., M. Ayres and F. M. Salzano 1967 Tuberculin reactions, x-ray and bacteriological studies in the Cayapo Indians of Brazil. Tubercle, 48: 195-200. Salzano, F. M. 1961a Studies on the Caingang Indians. I. Demography. Hum. Biol., 33: 11C130. 1961b Studies on the Caingang Indians. 111. Blood groups. Am. J. Phys. Anthrop., 19: 391404. 1964a Blood groups of Indians from Santa Catarina, Brazil. Am. J. Phys. Anthrop., 22: 91-106. 1964b Salivary secretions of Indians from Santa Catarina, Brazil. Am. J. Hum. Genet., 16: 301-310. 1971 Demographic and genetic interrelationships among the Cayapo Indians of Brazil. Social Biol., 18: 148-157.