Volume 107 1998 > Volume 107, No. 2 > Estimation of stature for the prehistoric/protohistoric Rapanui, by Patrick M. Chapman and George W. Gill, p 187-194
ESTIMATION OF STATURE FOR THE PREHISTORIC/PROTOHISTORIC RAPANUI
Information on the stature of the prehistoric Easter Island (Rapa Nui) population is scarce and sometimes difficult to acquire. Most of the information comes from the early historical accounts of the first European visitors to the island (see Ayres and Ayres 1995, Beechey 1831, Cook 1777, Eyraud 1866, Forster 1777, Langdon 1995, Loti 1991, Sharp 1970). Other than these scant and generally subjective reports, one has to rely on the material presented by Murrill (1965, 1968), reanalysed by Houghton (1996), and a recent study by Orefici and Drusini (1997). Unfortunately, these studies have small sample sizes, and all but Houghton (1996) use stature formulae developed for non-Polynesian populations. This paper addresses these problems by providing stature estimates for late prehistoric/protohistoric Rapa Nui males and females based on a large sample of osteological remains. We will also present comparisons between different Easter Island sites as well as with other East Polynesian populations.
The earliest available information on the stature of the late prehistoric and protohistoric Rapanui comes from the accounts of the first European visitors to the island. Easter Island was first encountered by Europeans on Easter Sunday 1722 by the Dutchman Jacob Roggeveen, who noted: “These people are well-proportioned in limbs, having very sturdy and strong muscles, are generally large in stature, and their natural colour is not black, but pale yellow or sallow” (in Sharp 1970:97).
Senior Lieutenant Alberto Olaondo, from a 1770 Spanish expedition led by Captain Felipe González, indicated that the Rapanui were “mostly of good build, very robust, well made, extremely lively and lithe” (in Langdon 1995:115).
In 1774 George Forster, a naturalist on Captain James Cook's expeditions, observed that the Rapanui “were of a middle stature, but rather thin” (Forster 1777:558). He also noted that they “were inferior in stature to the natives of the Society and Friendly Isles [Tonga], and to those of New Zeeland [sic]” (1777:564). On the same voyage, Cook noted: “In general the people of this isle are a slender race. I did not see a man that would measure six feet [182.9 cm]” (Cook 1777:290).
Frederick William Beechey, who captained the H.M.S. Blossom in 1825, stated: “our estimate of the average height of the people [assumed males] was 5 feet 7 1/2 inches [171.5 cm]” (Beechey 1831:51).
The prehistoric period on Easter Island ended with the establishment of the first Christian mission, headed by the Frenchman Eugène Eyraud. In 1864 Eyraud observed: “These natives are tall, strong and well built” (Eyraud 1866:55, our translation).- 188
Pierre Loti, from the French naval ship La Flore, noted in 1872: “that they are Polynesians is incontestable. Having become a little paler than their ancestors, due to the cloudy climate, they remain just as tall” (Loti 1991:15, our translation).
A more detailed account of the Rapanui was provided by Captain Wilhelm Geiseler of the Hyäne in 1882. Among other observations he noted:
The average height of a man seems to be 1.60-.70 m, though one of the men was found to be 1.82 m tall. In comparison the women are smaller just as on other South Sea islands and in this respect they remind one of Japanese women (in Ayres and Ayres 1995:50).
In brief, most of the observers agreed that the Rapanui were of “average” to “tall” stature with only Geiseler providing specific estimates. Previous osteological studies have come to a similar conclusion. Murrill (1965:299-300) determined that
correct mean stature estimate [for males] may lie between the figures 169.6 cm. (L) and 168.2 cm. (R) derived from the Trotter and Gleser formula and 172.7 cm. (L) and 171.5 cm. (R) derived from the Dupertuis and Hadden formula.
The same can be said for the females; 153.6 cm. (L) and 152.4 cm. (R) derived from Trotter and Gleser formula, and 156.9 cm. (L) and 155.8 cm. (R) derived from the Dupertuis and Hadden formula.
Murrill (1965:298) emphasised that “in this sample, the number of femora used is small. For the males, left or right, the number is 11. For the females, left 8, right 5. Thus, any interpretation of stature estimation in this case should be a cautious one.”
Two main problems are evident with Murrill's estimates. First, both formulae mentioned above are based on the living stature of whites since no formula for the estimation of stature for Polynesians had as yet been derived. Houghton (1990) illustrates the significant morphological differences between Polynesians and Caucasians, especially in the ratio between leg and upper body length. This difference prevents accurate estimation of Polynesian stature through formulae derived for other groups. Second, as Murrill himself indicated, the sample size used for his estimates was small and therefore non-representative.
Houghton (1996) addressed the first of these problems in a recent publication. Using Murrill's data, Houghton recalculated Rapanui stature using stature formulae derived from New Zealand Maori (Houghton et al. 1975). He estimated that Rapanui males averaged 172.8 cm, while females averaged 160.0 cm. However, once again the sample size is small with n = 14 for males and n = 8 for females.
Orefici and Drusini (1997) analysed the osteological material from the recent Tongariki Archaeological Project led by Cristino. They calculated average stature from each of the six long bones using formulae derived by Trotter and Gleser (1952, 1977) for Whites and Mongoloids. The largest sample size for the different long bones was 15 for males and 7 for females. They obtained mean stature by averaging - 189 the derived values for each of the six long bones (no comment was made on whether this methodology increased sample size). They estimated that Rapanui stature averaged 170.61 cm for males and 156.12 cm for females, with the tallest individual standing at 177.23 cm. Both of these estimates are similar to estimates provided by Murrill (1965, 1968) but are noticeably lower than the revised estimates calculated by Houghton (1996). The use of formulae derived from White and Mongoloid populations instead of one derived from Polynesians has likely caused some of the difference.
The data which comprise the base samples for the current study are from the extensive data base assembled by Gill and associates at the University of Wyoming. The sample consists of 92 Rapanui representing 54 males and 38 females. Assessment of sex is based on accepted osteological and forensic procedures detailed in Stewart (1979) and Bass (1987), among others. Gill (1986) has shown that the Rapanui skeletal material used in this study is late prehistoric (A.D. 1680-1722) and protohistoric (A.D. 1722-1868). Further details regarding provenance may be found in Gill and Owsley (1993) and Owsley et al. (1994). Long bones not associated with other osteological material are not included in this study to ensure that individuals are not represented more than once. For the same reason the stature estimates are based on bones from the right side of the body only.
Gill and Owsley (1993) also demonstrate that the skeletal material represents a Polynesian population with, at most, minimal gene flow from non-Polynesian groups. In addition, mitochondrial DNA analysis (Hagelberg et al. 1994) demonstrates that the Rapanui skeletal material has the 9 base pair deletion and characteristic Polynesian base substitutions confirming Polynesian ancestry. A recent cranial nonmetric analysis (Chapman and Gill 1998) also supports a Polynesian ancestry for the prehistoric Rapanui with minimal to no gene flow from South America.
Given the Polynesian ancestry of the Rapanui, estimation of stature is calculated using a regression formula derived from New Zealand Maori and developed by Houghton et al. (1975). Individual stature estimates are based on, in ranked order, data available for the femur, tibia, fibula, humerus, radius or ulna. Lower limb bones are given precedence over the bones of the arm since they correlate more closely with stature (Stewart 1979). Mean stature is then calculated for Rapa Nui males and females, as group entities, as well as for males from five site locations. Intra-island variance is examined using the single-factor ANOVA statistical method in which alpha is .05. The single-factor ANOVA is an analysis of sample variance based on the hypothesis that the means from the samples analysed are equal.
The mean stature for Rapa Nui late prehistoric/protohistoric males is determined to be 1726 mm and 1595 mm for females. As is common among modern populations, the range of individual stature for males and females displays some overlapping: the range for males is from 1664 - 1817 mm, and for females 1541 - 1672 mm (Table 1), making 1817 mm the maximum stature with a minimum of 1541.- 190
Table 1: Stature estimates of Rapa Nui females and males
We divide Rapa Nui males into five groups based on site location to determine possible within-group variation. Included are two north coast sites (Anakena and Mahatua), two south coast sites (Akahanga and Oroi) and one west coast site (Kihi Kihi Rau Mea; see Fig. 1). Of the five site locations, males from the north coast have the highest means (Mahatua =1737 mm and Anakena = 1736 mm). The two south coast locations display the lowest means (Akahanga = 1720 mm and Oroi = 1715 mm), with Kihi Kihi Rau Mea on the west coast intermediate at 1724 mm (Table 2). ANOVA analysis indicates that the variation is not significant (p = .649).- 191
Table 2: Stature estimates of Rapa Nui males by site location
Discussion and Conclusion
The mean estimates of stature for Rapa Nui males and females presented here are close to the means presented in Houghton (1996), differing only by 2 mm for males and 5 mm for females. The estimates are in general agreement with the European observations of the protohistoric Rapanui presented above. Geiseler in 1882 (in Ayres and Ayres 1995) placed the Rapa Nui males as being between 160-170 cm, slightly shorter than the estimate obtained here, with the tallest at 182 cm equivalent to the tallest estimate obtained in this study. It is possible that the stature of the Rapanui decreased slightly from the prehistoric/protohistoric into the historic period, possibly due to disease, environmental stress and the population bottleneck which occurred in the 1870s (Bahn and Flenley 1992).
Compared to other Polynesian populations to which Houghton et al.'s (1975) formulae have been applied, the Rapa Nui males are shorter than all but two Hawaiian samples (Table 3), while the females exceed only the Marquesans in stature. A possible explanation for the reduced height of the Rapanui in comparison to other Polynesians was an inconsistent food supply (Pollock 1993). Pollock (1993) suggests that the Rapa Nui food supply was subject not only to seasonal tapus but also periodic shortage and famine. Nutrition is a critical factor contributing to a person achieving their genetic potential for height (Relethford 1994). Therefore, the lack of a consistent food supply may be responsible for the Rapanui being among the shortest of the Polynesian populations. However, it should be noted that the actual differences between the various Polynesian populations are not great, with the majority of the means for both males and females falling within 25 mm (about 1 inch) of each other. It is interesting to note that among all groups included in the comparison the Keopu males are the shortest and the Keopu females are the tallest. This suggests that there may be a problem in the determination of individual sex with some females mistakenly included in the male sample.
Chapman (1993), Gill (1988), Gill and Owsley (1993) and Gill et al. (1983) suggest there was biological variation among Easter Island regional groups. While this study demonstrates that the Rapa Nui males from the north coast do tend to be taller, the difference is not significant. Thus, if regional variation did exist among the prehistoric Rapanui, it is not evident in the stature estimates.- 192
Previous studies of Rapanui stature have been based on small samples and have generally used methods devised for non-Polynesian populations. This study corrects both problems by using statistically significant sample sizes with stature estimates based on formulae developed specifically for Polynesians. The mean stature estimates for the Rapa Nui males (1726 mm)and females (1595 mm) fall within the range found in other Polynesian populations, although both tend to be among the shortest. Mean stature estimates for late prehistoric/protohistoric Rapa Nui males demonstrate non-significant regional variation, where north coast populations are the tallest and south coast populations the shortest.
Table 3: Comparison of East Polynesian stature estimates using formulae derived by Houghton et al. (1975)
The authors would like to thank the Sebastian Englert Museum on Easter Island for the use of the Rapa Nui skeletal material. Chapman would like to thank Dr. Thor Heyerdahl and the Kon-Tiki Museum for providing grant money assisting with the data analysis and Martin Fisher for creating the map.
Gill's excavation and data collection on the Easter Island skeletal sample was supported by the National Georgraphic Society, Center for Field Research + Earthwatch, the government of Chile, the Pacific Foundation, World Monuments Fund, the Kon-Tiki Museum, the Smithsonian Institution, the University of Wyoming and the University of Chile. Gill would also especially like to thank Dr. Douglas W. Owsley of the Smithsonian Institution, Scott James Baker of the University of Pittsburgh, and Sonia Haoa C. of Easter Island for their help in osteological data collection. Gill would also like to thank the various scientists and museum curators who have made the Easter Island research possible. Among these are: Sergio Rapu H., Claudio Cristino, Dr. Andrea Seelenfreund, Dr. Leslie Snow, Gonzalo Figueroa G.-H., Silvia Quevedo K., Dr. Ian Tattersal, Claudio Gomez and José Ramirez.- 193