Highland Guatemala was inhabited by antagonistic Maya and other ethnic groups in the fifteenth and early sixteenth centuries (Figure 1). Kaqchikel Maya were vassals of the K'iche’ Maya, whose capital was Q'umarkaj, but fled that polity's internal conflicts and founded their capital of Iximche’ between AD 1463 and 1470 (Borg Reference Borg, Roger Nance, Whittington and Borg2003). Iximche’ was built on a strategically located and defensible hilltop (elevation 2,260 m) surrounded by steep ravines on three sides. Iximche’ means “maize-tree” and is the Kaqchikel name for the ramón or breadnut tree (Brosimum alicastrum) that grows around the site (Fox and Smith Reference Fox and Smith2006).
Highland and Pacific coastal Guatemala showing places mentioned in the text and their predicted human stable oxygen isotope compositions (δ18Op) based on results from the Online Isotopes in Precipitation Calculator (Bowen Reference Bowen2006) converted using an equation published by Daux and others (Reference Daux, Lécuyer, Héran, Amiot, Simon, Fourel, Martineau, Lynnerup, Reychler and Escarguel2008), as explained in the text. Associated ethnic groups for the sites are Alotenango (Kaqchikel); Beleh (Poqomam?); Chiaa (Tz'utujil); Cotzumalguapa (Pipil [non-Maya]); Mixco / Chinautla Viejo (Poqomam); Escuintepeq (Pipil); Iximche’ (Kaqchikel); Kaminaljuyu (Poqomam?); Mixco Viejo / Jilotepeque Viejo (Aquajal / Chajoma’); Patulul (Tz'utujil); Petapa (Poqomam); Pinula (Poqomam); Q'umarkaj (K'iche’); Siquinalá (Pipil); Todos Santos Chuchumatán (Mam); Tzolola’ / Tz'olojya (Kaqchikel); Ucubil (Los Sacatapéquez); Yampuc Viejo (Los Sacatapéquez).
Warfare soon became endemic both between Iximche’ and surrounding groups and internally (Borg Reference Borg, Roger Nance, Whittington and Borg2003; Recinos Reference Recinos1991). The ruler of the Aqajal / Chajoma’ Kaqchikel branch, his principal lords, and warriors were killed when they came to Iximche’, and Aqajal / Chajoma’ Mixco Viejo / Jilotepeque Viejo came under early Iximche’ control. A coalition of Iximche’ and Los Sacatapéquez, an Aqajal / Chajoma’ branch centered at Ucubil, countered a Poqomam incursion northeast of Mixco / Chinuatla Viejo in 1475. The Kaqchikel and K'iche’ constantly fought beginning in 1475. Iximche’ conquered Mixco / Chinuatla Viejo in 1480 and took many captives. Iximche’ resisted an attack by the K'iche’ in 1491, during which the Kaqchikel captured and executed two K'iche’ rulers, many court dignitaries, and warriors. The Tuquche’, a Kaqchikel branch, revolted and attacked Iximche’ in May 1493 but were defeated. Iximche’ attacked Mixco / Chinuatla Viejo again in 1497, while also conquering the Poqomam towns of Pinula and Petapa. In 1517, Iximche’ again defeated the K'iche’, taking many prisoners, and conquered Tzolola’ / Tz'olojya, home of another Kaqchikel branch. Los Sacatepéquez revolted and Iximche’ eventually conquered their new stronghold at Yampuc Viejo between 1517 and 1524. By 1524, an Iximche’–Tzolola’ coalition wrested control of Pacific lowland Patulul from the Tz'utujil, whose capital was Chiaa, and other Pacific lowland territory around Escuintepeq, including Cotzumalguapa and Siquinalá, from the non-Maya Pipil. After the cessation of hostilities with the Pipil, Iximche’ founded a settlement at Alotenango.
Spaniards and their allies from Mexico arrived in 1524 under the command of Pedro de Alvarado and conquered the K'iche’. The Kaqchikel allied themselves with the invaders, hoping to defeat their other enemies. Alvarado founded the first colonial capital of Guatemala at Iximche’, but hostilities between the Kaqchikel and Spaniards and a revolt within the Spanish ranks led to the destruction and abandonment of Iximche’ in 1526 (Fuentes y Guzmán Reference Fuentes, Guzmán and Antonio de1932 [1690]).
George Guillemin (Reference Guillemin1967, Reference Guillemin1969, Reference Guillemin and Hammond1977) excavated elite sections of Iximche’ between 1958 and 1972. His main interest was architecture, and his crews exposed and then restored plazas, palaces, altars, and temples (Figure 2). Many high-quality photographs of the work exist. Guillemin's notes, photographs, and drawings document 68 complete or partial human burials but do not describe the provenience, associated artifacts, or basic layout for some of these burials. Only cryptic notes or tags bagged with bones exist for some human remains. Guillemin died in 1978 before he analyzed the excavated materials or wrote a comprehensive monograph on Iximche’.
Iximche’ looking west from the top of Temple 5. Structure 104 is the small square structure located at the southern corner of Temple 4 (on the right) near the bottom of the photograph. (Color online)
After 1978, while the remains were stored at the site, bags broke open, artifacts and bones from multiple contexts became mixed together, and tags with provenience information rotted or became separated from their bags. C. Roger Nance endeavored to organize the contents of the bags and started to analyze the artifacts in 1990.
In 1992, Whittington and Reed began to clean and study the human skeletal remains, some still in soil matrix. Whittington returned in 1993 and 1995 to complete analyses of the remains’ sex, age at death, trauma, and diseases. Bone and tooth samples were taken for stable carbon and nitrogen isotope analyses to explore the highland Maya diet. Reed sorted all ribs found together in each storage bag by morphology, preservation, size, side, and location within the rib cage. For stable isotope analysis, he took 18 samples from fragments of the first rib, when available, to avoid sampling individuals twice. For skeletons that appeared not to have a first rib, fragments of another rib were taken. All 43 teeth chosen for analysis were third molars, in which the crown typically forms between 9 ± 2 and 12 ± 3 years of age in Native Americans (Ubelaker Reference Ubelaker1989).
This study is a product of a particular time and place. Bone and tooth samples were collected and analyzed during the 1990s and 2000s, long before the Society for American Archaeology (2021) issued the “Statement Concerning the Treatment of Human Remains.” The research was performed ethically and with permission of Guatemalan authorities but did not involve consultation with Indigenous descendant communities, which was not an expectation then. Relatively little comparable work has been done in the Highlands. For example, multi-isotopic surveys to produce faunal and plant baselines for subsistence and migration studies, available for the Maya Lowlands (Rand et al. Reference Rand, Freiwald and Grimes2021) and Peru (Szpak et al. Reference Szpak, White, Longstaffe, Millaire and Vásquez Sánchez2013), have not been done in the Highlands.
Decisions about stable isotope analysis of the bones were made that reflected biogeochemical practice and not necessarily paleodiet practice at the time the work was performed. The choices of which isotopes to analyze were based on the literature and research interests current at the time. Dietary studies of the ancient Maya were of prime interest, and analyses of stable carbon and nitrogen isotopes in bones and teeth were standard. Studies of migrations of ancient Maya individuals and groups based on bones were just beginning to be conducted, and the best methods for doing so were still being determined.
Oxygen isotopes were one source for studying movements of individuals and populations (e.g., White et al. Reference White, Spence, Longstaffe and Law2000). Strontium isotope analysis, first described in the mid-1980s and now considered by many to be the measure of choice for human provenience profiling, was just beginning to be done in Mesoamerican skeletal studies (e.g., Price et al. Reference Price, Manzanilla and Middleton2000). Because of decisions made long ago, oxygen isotope data are available for Iximche’ (Olsen Reference Olsen2006) but not strontium isotope data. In what follows, we present isotopic evidence to test hypotheses that most of the individuals buried at Iximche’ were of high status and that many were foreigners who ended up there because of warfare.
Materials
At least 66 whole and partial crania were present in the human skeletal sample stored at Iximche’, based on elements that still formed units (e.g., occipital-parietal-temporal). Postcranial bones from primary and secondary burials were present for at least 26 individuals. Cranial elements representing distinct individuals were each assigned an IX label. Except for cervical vertebrae, it was rarely possible to match cranial with postcranial material for two reasons: provenience information became mixed or lost during storage, and many crania were from decapitations (see the later discussion). When it was possible to associate postcranial remains with a cranium, they were assigned the same IX label. Groups of postcranial bones that either articulated or came from opposite sides of the body but matched in terms of preservation, surface color, size, morphology, robustness, and degree of epiphyseal union were assigned two-letter designations, unless they were found bagged with cranial elements assigned an IX label. Relatively complete single postcranial bones independent of groupings were also assigned a two-letter designation. Single cranial elements unassociated with any others and small postcranial fragments were identified by the bag in which they were found.
Guillemin's notes and publications indicated that at least 50 crania were from decapitations. Seventeen of the 66 crania could be visually matched with excavation photos of crania that Guillemin identified as decapitations. Sixteen came from a group of 48 crania deposited together adjacent to Structure 104 in Plaza C that Guillemin (Reference Guillemin1969, Reference Guillemin and Hammond1977) identified as a skull rack (Figure 3). One came from a deposit of two crania associated with Structure 74 in Plaza A, which he also identified as a skull rack (Guillemin Reference Guillemin1967, Reference Guillemin1969). The ritual context of these structures is clear. Each was a small, low platform located adjacent to the southern corner of the western pyramidal structure, identified by Guillemin as a temple, in the two largest plazas at Iximche’. Structure 74 was decorated with polychrome murals showing stylized skull-and-crossbones motifs. Structure 104 was badly preserved and lacked features demonstrating its function.
Iximche’ site plan with known burial locations of individuals in this study circled.
There is strong evidence that the crania deposited around these structures were decapitations. Most crania included both the calvarium and mandible, and many were deposited with cervical vertebrae, indicating that soft tissue still adhered to the bones when they were buried. All but one of the 17 decapitations identified by Guillemin had damage on the cranial bones, associated cervical vertebrae, or both, related to separating the head from the body (Whittington Reference Whittington, Roger Nance, Whittington and Borg2003). Beck and Sievert (Reference Beck, Sievert, Rakita, Buikstra, Beck and Williams2005), Ruiz (Reference Ruiz, Tiesler, Suzuki and Pereira2021), and Tiesler and Miller (Reference Tiesler and Miller2023) reported damage to the base of the skull near the foramen magnum or angles of the mandible like that found at Iximche’ for sacrificial victims at Chichen Itza, Toniná, and Tula. However, the crania at Iximche’ lacked any bilateral or basal damage associated with display on vertical or horizontal poles, as Tiesler and Miller reported for Chichen Itza. If Structures 74 and 104 functioned to display heads, they were more likely skull stands than skull racks.
The crania were deposited over an extended period. Structure 104 crania differed remarkably in their states of preservation. Some had been buried together in small lots, but the rest had been set individually in pits cut into the plaza floor. Poorly preserved crania lacking mandibles and cervical vertebrae might have been exposed to the elements on the adjacent structure for long periods before burial. Well-preserved crania accompanied by mandibles and vertebrae might have been placed in the ground soon after they were removed from the rest of the body. For stable isotope analysis, we selected third molars from 15 crania from Structure 104 and 1 from Structure 74.
Two crania were associated with Structure 27 at the western edge of Plaza A. One came from a group consisting of a complete skeleton (unavailable for study) crowned by a gold “diadem” and three other skeletons. A third molar was taken for analysis from an individual with an apparently complete postcranial skeleton whose cranium exhibited decapitation damage. Guillemin's drawing of the burial shows that the torso of this individual was prone, but the cranium was facing upward. The three skeletons could represent sacrificed retainers or family members of the one wearing the diadem. Another cranium from Structure 27 was part of the primary burial of an individual. A rib from this individual was selected for stable isotope analysis.
Structure 38 is a large structure defining the eastern side of Plaza C. Two individuals were interred together in a burial pit. One cranium may have come from the pit, but it was not sampled. Rib fragments from one individual accompanied by gold beads, shell beads, jade, and copper ear ornaments were selected for stable isotope analysis.
Structure 39, located at the far eastern edge of Plaza C, was associated with cranial elements from six individuals. One cranium was part of a primary burial accompanied by a jade pendant; a rib and third molar from the individual are included in this study. The burial types of the remaining five individuals are unknown. Third molars from them were collected for analysis.
Structure 45 is a relatively small construction located at the far eastern edge of the site, behind structures that define Plaza D. Elements from four crania whose burial types are unknown were associated with Structure 45; third molars from two of them were analyzed for this study.
Other third molars and ribs in this study cannot be associated with specific structures or skeletal remains described by Guillemin.
Methods
Crania were assigned an age based on bone characteristics, dental development, and tooth eruption. Adult crania were sexed by the presence or absence of 17 traits. Adult postcranial remains were sexed based on the presence or absence of 15 traits on the innominate and assigned an age based on auricular surface seriation. Subadult postcranial bones were identified by their small size and lack of epiphyseal union. Sex determination resulted in confirmed male and female and probable male and female assignments. Age-at-death determinations for the current study are grouped into subadults (<15 years old), young adults with incompletely formed third molar roots (15–21 years old), and adults (>21 years old). See Whittington (Reference Whittington, Roger Nance, Whittington and Borg2003) for the detailed methodologies used to determine age and sex.
Confirmed decapitations (n = 17) are crania that were visually matched to excavation photos that Guillemin identified as decapitations. Probable decapitations (n = 14) are based on the presence of decapitation damage, documentation, or both. Probable nondecapitations (n = 3) are based on documentation. Confirmed nondecapitations (n = 4) are crania from nearly intact skeletons lacking decapitation damage.
We used standard methods to analyze the stable isotopes (Supplemental Text 1).
Stable Isotopes
Stable carbon and nitrogen isotope compositions, together with paleobotanical, paleopathological, and paleosocial interpretations, provide a method of determining diet directly from the remains of an individual. Oxygen isotope signatures provide a way to reconstruct individual migration and mobility. Isotopic analyses of bones and teeth can differentiate the crania of venerated ancestors from sacrificial victims, victims of political upheaval from war captives, and foreign enemies from local criminals; they also can determine the origins of decapitated crania (e.g., Cheung et al. Reference Cheung, Jing, Tang, Weston and Richards2017; Forgey and Williams Reference Forgey, Williams, Rakita, Buikstra, Beck and Williams2005; Price et al. Reference Price, Burton, Wright, White, Longstaffe, Tiesler and Cucina2007; White et al. Reference White, Spence, Longstaffe, Stuart-Williams and Law2002).
Collagen is the nonmineral portion of bone. The stable isotopes of carbon (δ13C, ‰), reported relative to Vienna Pee Dee Belemnite (VPDB), and nitrogen (δ15N, ‰), reported relative to ambient air (AIR), are used to infer diet (DeNiro and Epstein Reference DeNiro and Epstein1978, Reference DeNiro and Epstein1981). The collagen isotopic compositions can be related to terrestrial or aquatic resources (DeNiro Reference DeNiro1987). Terrestrial plants have distinct stable carbon isotope signatures (Coleman and Fry Reference Coleman and Fry1991), based on the photosynthetic pathway. C3-based plants (around –26‰), for example, are leafy and include legumes. C4-based plants (around –12‰) include tropical grasses such as maize (Ambrose and Norr Reference Ambrose, Norr, Lambert and Grupe1993). Nitrogen isotope compositions differ among plants based on their nitrogen source (e.g., symbiotic bacterial fixation or soil nitrate), climate, and environment. These differences are transferred to primary consumer collagen, albeit enriched in 13C by ~5‰ and 15N by ~3‰ (Tykot Reference Tykot, Staller, Tykot and Benz2006). Bone collagen is produced from dietary protein, with an upward shift of about 1‰ for δ13C and about 3‰ for δ15N for each consumer trophic level (Ambrose and Norr Reference Ambrose, Norr, Lambert and Grupe1993; Hedges and Reynard Reference Hedges and Reynard2007). Plant δ13C also shifts toward less negative values with increasing altitude (e.g., Körner et al. Reference Körner, Farquhar and Roksandic1988).
Bioapatite forms a portion of bone and tooth enamel. In the biomineral equivalent of hydroxyapatite (Ca10(PO4)6(OH)2), carbonate (CO3) can replace phosphate (PO4) or hydroxyl (OH) (Sillen Reference Sillen and Douglas Price1989). The stable carbon isotope composition of this structural carbonate (δ13Csc) in bone reflects the whole diet (Ambrose and Norr Reference Ambrose, Norr, Lambert and Grupe1993).
Oxygen isotopes (δ18O, ‰)—reported relative to Vienna Standard Mean Ocean Water (VSMOW) or VPDB—of most bone and enamel are derived from environmental water (δ18Oprecip). Values of δ18Oprecip are influenced by geographic and climatic factors, including latitude, elevation, temperature, and humidity (Yurtsever and Gat Reference Yurtsever, Gat, Gat and Gonfiantini1981), yielding characteristic values of δ18Oprecip for many localities. Values of δ18Oprecip are incorporated into bioapatite PO4 and CO3. Body water—derived from consumed environmental water—isotopically equilibrates with body phosphate at body temperature (Longinelli Reference Longinelli1984), which controls the oxygen isotope composition of bioapatite phosphate (δ18Op). There is generally a good correlation between δ18Op and the oxygen isotope composition of bioapatite structural carbonate (δ18Osc; Iacumin et al. Reference Iacumin, Bocherens, Mariotti and Longinelli1996). Values of δ18Osc, however, are more prone to postmortem changes than those of δ18Op (Nelson et al. Reference Nelson, DeNiro, Schoeninger and DePaolo1986).
Turnover of bone collagen means that δ13Ccol and δ15Ncol reflect the diet averaged over several years before death (Schwarcz and Schoeninger Reference Schwarcz and Schoeninger1991). The amount of time averaged depends on which bone is sampled. At one extreme, turnover takes three to five years in ribs (Cheung et al. Reference Cheung, Jing, Tang, Weston and Richards2017); at the other extreme, the cortical bone of the femur may never completely turn over and may partially reflect the adolescent diet in adults (Hedges et al. Reference Hedges, Clement, Thomas and O'Connell2007). Unlike bone, enamel bioapatite is not subject to continued remodeling after initial formation. Tooth enamel δ13Csc, δ18Osc, and δ18Op, therefore, reflect diet and water consumption at the age of formation, regardless of age at death. Hence, enamel isotopic composition can signal changes in diet, residence location, or both between the time of tooth formation and the remodeling of bone that occurred over the last years of life. When comparing stable isotope compositions to expose dietary or mobility differences between skeletal populations, the most valid comparisons are those between the same bone or tooth. However, there is little consistency in which bones or teeth have been studied in the Maya area, even within single skeletal samples. For δ13Ccol and δ15Ncol, we did not attempt to standardize results between sites because of the great variation in which bones were sampled and the ages of sampled individuals. In comparing δ13Csc, δ18Osc, and δ18Op from Iximche’ to other samples, we try to make it clear what is being compared.
Results
The δ13Ccol and δ15Ncol data suggest that the diet of both adult and subadult residents of Iximche’ was high in maize content (Table 1; Supplemental Text 2). The area around Iximche’ was especially well suited to the production of C4-based maize (McBryde Reference McBryde1947). C3-based plants possibly augmenting the diet and affecting isotope compositions include beans, chile, pumpkin (or squash), avocado, chayote, pacaya, yucca, purslane, and black nightshade, as well as ramón, but the last item is considered a famine food (Standley Reference Standley1946). Possible animal sources of protein include marine and freshwater fish, birds such as ducks and quail, turtles, white-tailed deer, pocket gophers, cottontail rabbits, and domesticated dogs, although some may have been obtained through long-distance trade and primarily used in ritual contexts (Emery et al. Reference Emery, Thornton, Cannarozzi, Houston, Escobedo, Götz and Emery2013; Sharpe et al. Reference Sharpe, Arroyo, Estrada, Ajú and Serech2022).
Stable Isotope Ratio Measurements of Carbon and Nitrogen from Iximche’ Bone Collagen.
Of the ribs analyzed, three were set aside as duplicate samples. Two others were extreme outliers, which might represent some group but share no obvious social, demographic, or pathological characteristics. For the remaining 15 ribs, δ15Ncol has a mean value of +7.9 ± 1.4‰, and δ13Ccol has a mean value of –7.7 ± 0.8‰. For the 13 samples without the two outliers, δ15Ncol has a mean value of +8.0 ± 0.6‰, and δ13Ccol has a mean value of –8.0 ± 0.6‰.
Based on bone collagen data, diets of presumed residents of Iximche’ were different from those of most other tested Maya sites (Figure 4); they especially differed from Postclassic diets at five lowland sites. The mean δ15Ncol for Iximche’ is low relative to most other sites (range = +7.4 to +11.6‰), which indicates that residents of Iximche’ ate a terrestrial diet consisting of less meat and more maize than residents of most lowland sites. The specimens from Iximche’ have a less negative mean δ13Ccol than most other Maya sites (range = –14.8 to –6.5‰). Thus, individuals from Iximche’ seem to have been dependent on maize to a greater degree than most lowland Maya. This probably reflects dietary differences among the sites but perhaps regional differences as well, such as the 2,260 m altitude at Iximche’, higher than any other Maya site yet sampled.
Mean bone δ13Ccol and δ15Ncol for Maya localities for which five or more results are reported. Geographic identifications are based on Scherer (Reference Scherer2015), but with additional categories. Exclusively Postclassic samples are circled. Seibal and Ceibal are alternate spellings used by different researchers for one site. Iximche’ compositions are for ribs, but results for other sites are from various bones. See Supplemental Figure Data 1 for sources of data in this figure.
1 Aguateca
2 Altar de Sacrificios
3 Altun Ha all
4 Altun Ha Postclassic
5 Baking Pot
6 Barton Ramie
7 Belize Valley region
8 Blackman Eddy
9 Cahal Pech
10 Caledonia
11 Campeche Historic
12 Ceibal
13 Chac Balam
14 Chau Hiix all
15 Chau Hiix Postclassic
16 Chinikihá
17 Copán
18 Copán region
19 Cuello
20 Dos Barbaras
21 Dos Hombres
22 Dos Pilas
23 Floral Park
24 Holmul
25 Itzán
26 Iximche’ Postclassic
27 K'axob
28 La Milpa
29 Lamanai all
30 Lamanai Postclassic
31 Marco Gonzalez
32 Mayapán Postclassic
33 Minanha all
34 Minanha Postclassic
35 Mojo Cay
36 Nakum
37 Pacbitun
38 Petén region
39 Piedras Negras
40 Pook's Hill
41 San Juan
42 San Pedro
43 Saturday Creek
44 Seibal
45 Tamarindito
46 Tikal
47 Tipu
48 Uaxactún
49 Xcambó
50 Xunantunich
51 Yaxuná
Based on δ13Csc from tooth enamel bioapatite, the juvenile diet of adults buried at Iximche’ was high in maize content (Table 2). However, in contrast to the rib samples, most of the tooth samples are from decapitations, at least some of which may not have been residents of the site (see the next section). The 43 tested tooth samples have a mean δ13Csc of –2.1 ± 1.1‰. Four individuals (IX-2, IX-4, 224-v, and 270-i) have extreme δ13Csc relative to the median of –2.0‰ (Figure 5a).
Box and whisker plots showing the distribution of tooth enamel δ13Csc. The horizontal line dividing the boxes is the median, and x is the mean. The boxes show the upper and lower quartiles. Horizontal lines at the end of the whiskers are the upper and lower quartiles plus or minus 1.5 times the interquartile range. Points outside those lines are outliers. (a) All 43 third molars; (b) pooled decapitations versus pooled nondecapitations (explained in the text); (c) pooled females versus pooled males; (d) adults versus young adults; (e) comparison of decapitations by age and sex.
Stable Isotope Ratio Measurements of Carbon and Oxygen from Iximche’ Enamel Bioapatite.
The average δ13Csc of teeth (–2.1‰) should be adjusted by –12.0‰ to estimate the average δ13C (–14.1‰) of foods that people buried at Iximche’ consumed (Lee-Thorp and van der Merwe Reference Lee-Thorp and van der Merwe1987). With maize averaging about –9.5‰ when adjusted for pre-anthropogenic atmospheric levels (Wright Reference Wright1994), C3-based plants and animals (δ13C of –26 to –19‰) formed a minor portion of the diet. Using theoretical end points of –14‰ and +2‰ in bioapatite (Lee-Thorp and van der Merwe Reference Lee-Thorp and van der Merwe1987) for pure C3 and pure C4 diets, respectively, maize or maize-fed animals contributed 75% to 85% of the whole diet at the time of tooth formation. This interpretation is consistent with that based on bone δ13Ccol, which indicates that dietary protein was derived primarily from maize, a poor source of protein.
Only one individual that could be evaluated for evidence of decapitation (IX-1) was tested for δ13Csc. Pooling individuals classified as decapitation with those classified as probable decapitation, and the individual classified as a nondecapitation with the one classified as a probable nondecapitation, permits more exploration of the data (Table 3). The decapitation / probable decapitation group has a wider range, more positive mean, and larger variance (Figure 5b). Three individuals categorized as decapitation or probable decapitation have extreme δ13Csc.
Comparison of δ13Csc and δ18Op Compositions in Groupings of Crania Buried at Iximche’.
Notes: n = number of observations; pU = two-tail p-value for Mann-Whitney U statistic; pF = right-tail p-value for Levene's F statistic; * = statistically significant at least at the 0.05 level.
aDecapitation and decapitation?, male and male?, and female and female? from Table 2 are combined.
When tooth δ13Csc are pooled for individuals classified as male and probable male, and for individuals classified as female and probable female, the range of values for the male / probable male group is wider, although the means are almost identical (Figure 5c). The variance for the male / probable male group is greater than for the female / probable female group. Two of the extreme values of δ13Csc are associated with probable males. The adult category has a wider range, more negative mean, and larger variance than the young adult category (Figure 5d). All four extreme values of δ13Csc are associated with adults.
The Mann-Whitney U test compares the mean ranks and tests whether one variable tends to have values higher than the other (Conroy Reference Conroy2012). Levene's F test statistic for homogeneity of variance tests whether ranges of variation between the samples are statistically different (Levene Reference Levene and Olkin1960). The tests indicate no significant difference in the distributions or variance of δ13Csc between groupings of crania based on pooled burial type, pooled sex, or age.
When we analyzed pooled decapitations / probable decapitations for differences based on age and sex, some statistically significant differences in variance emerged (Figure 5e). The adult female / probable female group has significantly more homogeneity than the adult male / probable male group (p = 0.049) and the young adult female / probable female group (p = 0.005).
In comparison to mean δ13Csc reported from elsewhere in the Maya area (range = –9.6 to –1.0‰; Figure 6), Iximche’ has one of the most positive values. Other sites with similar isotopic compositions are either located at altitudes approaching that of Iximche’ (Kaminaljuyu and Beleh) or mostly date to the Postclassic (Chichén Itzá, Chau Hiix, and Beleh).
Mean, minimum, and maximum tooth enamel δ13Csc for Maya localities for which five or more results are reported. Locations in the Guatemalan Highlands have darker bars. Names of localities for which exclusively Postclassic results are presented in the figure are in bold italics. Names of localities for which results are based exclusively on third molars in the figure are followed by (M3). See Supplemental Figure Data 1 for sources of data in this figure.
In the subset of 23 teeth whose phosphate portion was analyzed, the mean δ18Op is +15.6 ± 1.2‰ (VSMOW; Table 2; Figure 7a). The mean δ18Osc for those teeth is –6.9 ± 1.2‰ (VPDB). The correlation between the results of the two independent analyses is high (Pearson's r = 0.86, p < 0.00001), indicating only minor diagenesis of the enamel structural carbonate.
Box and whisker plots showing the distribution of tooth enamel δ18Op (see Figure 5 for explanation). (a) All 23 third molars with the predicted local values for Iximche’; (b) pooled decapitations versus pooled nondecapitations; (c) pooled females versus pooled males; (d) adults versus young adults; (e) comparison of decapitations by age and sex.
Different methods have been used to determine the baseline oxygen isotope composition for specific locations in the Maya area. Ideally, animal remains are available for direct comparison, but that is not the case for Iximche’. The Online Isotopes in Precipitation Calculator (OIPC) can be used to estimate the expected modern mean annual δ18Oprecip at a particular location (Bowen Reference Bowen2006). An alternative is to use “isoscapes”—maps of spatial distributions of oxygen isotope compositions derived from sampling local water sources. Lachniet and Patterson (Reference Lachniet and Patterson2009), Locker and others (Reference Locker, Valdez, Breecker, Banner, Loewy, Luzzadder-Beach, Drake, Hyde, Trachman and Lewis2023), Moreiras Reynaga and coworkers (Reference Reynaga, Diana, Millaire, Balderas, Román Berrelleza, Lujan and Longstaffe2021), Scherer and coauthors (Reference Scherer, de Carteret and Newman2015), Stansell and others (Reference Stansell, Steinman, Lachniet, Feller, Harvey, Fernandez and Shea2020), and Wassenaar and others (Reference Wassenaar, Van Wilgenburg, Larson and Hobson2009) provide data relevant to the Maya area.
The OIPC value for Iximche’ δ18Oprecip is –9.4‰ (VSMOW), which is very close to the –9.2‰ value for surface water collected from the Río La Vega (sample #132), located 8.93 km from Iximche’ and 191 m lower in altitude (Lachniet and Patterson Reference Lachniet and Patterson2009). Daux and coauthors (Reference Daux, Lécuyer, Héran, Amiot, Simon, Fourel, Martineau, Lynnerup, Reychler and Escarguel2008) published an equation showing the relationship between meteoric water (w) and phosphate in human bone:
If δ18O of freshwater in the region has not changed through time, this can be used to predict bioapatite δ18Op for Iximche’ residents of +15.8 to +15.9‰, which is close to the sample mean of +15.6‰.
Analysis of δ18Op for groupings reveals variations within teeth from crania buried at Iximche’ (Table 3). The pooled decapitation / probable decapitation group has a wider range, less positive mean, and greater variance than the nondecapitation / probable nondecapitation group (Figure 7b). Pooled females / probable females have a narrower range, more negative mean, and smaller variance than males / probable males (Figure 7c). Adults have a wider range, more positive mean, and larger variance than young adults (Figure 7d). Levene's F test indicates that only the variance difference based on sex is statistically significant (p = 0.03).
Comparing pooled decapitations / probable decapitations grouped by both age and sex (Figure 7e), Levine's F test indicates statistically significant differences in variance between males / probable males and females / probable females (p = 0.04), between adult females / probable females and adult males / probable males (p = 0.01), between adult females / probable females and young adult females / probable females (p = 0.00006), and between the four groups based on both age and sex (p = 0.01). The Mann-Whitney U test indicates a significant difference in mean ranks between adult females / probable females and young adult females / probable females (p = 0.0498), with adult females / probable females higher than young adult females / probable females.
In comparison to mean oxygen isotope compositions from other tested Maya sites, Iximche’ lies at extremes for both δ18Osc (range = –7.0 to –1.3‰, VPDB) and δ18Op (range = +15.6 to +27.9‰, VSMOW; Figures 8a and 8b). Iximche’ has the most negative mean δ18Osc and the least positive mean δ18Op of any Maya site. Other highland sites (Kaminaljuyu and Beleh) approach the values for Iximche’, but except for Beleh, other Postclassic sites have much different values.
Mean, minimum, and maximum bioapatite δ18O for Maya localities with five or more reported values. (a) Bone and tooth enamel δ18Osc; (b) bone and tooth enamel δ18Op. Locations in the Guatemalan Highlands have darker bars. Names of localities where exclusively Postclassic results are presented in the figure are in bold italics. The material that is the basis for the reported values in the figure follows each locality's name in parentheses. See Supplemental Figure Data 1 for sources of data in this figure.
Discussion
Evidence related to diet, health, and social status appears in bones and teeth from Iximche’ (Supplemental Text 3). Low frequencies of anemia, infection, and dental caries suggest generally good health, despite the heavy reliance on maize. Possibly this is because maize was a high-status food (White Reference White, Whittington and Reed1997), with as much as 25% of the diet coming from nutritious plant and animal foods. A high frequency of enamel hypoplasia defects, which is evidence of severe metabolic stress during the first seven years of life, implies that weaning was a time of difficulty for children, but this pattern is common in many other Maya skeletal populations. Most individuals studied at Iximche’ survived this period of subadult stress. Like evidence of good health, labret polishing on the lower anterior teeth of decapitations, particularly in males, suggests generally high status for those buried at Iximche’ (Miller Reference Miller2012). There is no statistical support for health differences between groups buried at Iximche’, nor are there any significant correlations between pathological conditions and δ13Csc, δ18Osc, or δ18Op.
Stable carbon and oxygen isotope ratios of tooth enamel bioapatite show consistent and parallel patterns in comparisons between groups buried at Iximche’. This is despite or perhaps because δ13Csc reflects juvenile diet, and δ18Osc and δ18Op reflect juvenile water consumption. Some variations in stable isotopes between groups are statistically significant, but evidence points to age, sex, or social status differences being the factor in most cases. With a few exceptions, the effects of different locations of residence on these variations are masked because oxygen isotope variation is greater within different populations than between them (Freiwald Reference Freiwald and Douglas Price2023).
It is worthwhile to explore the possible origins of individuals buried at Iximche’ who have extreme values for δ18Op and δ13Csc. Several methods have been used to define data outliers for isotopes in Maya populations, such as a 2‰ spread around the mean (White et al. Reference White, Spence, Longstaffe, Stuart-Williams and Law2002), Tukey's interquartile range method (Locker et al. Reference Locker, Valdez, Breecker, Banner, Loewy, Luzzadder-Beach, Drake, Hyde, Trachman and Lewis2023), and three times the median absolute deviation from the median (3MADnorm; Lightfoot and O'Connell Reference Lightfoot and O'Connell2016). Because comparison of these methods showed no differences in results for Iximche’, we discuss the outliers identified in Figures 5 and 7. No individuals have extreme values for both δ18Op and δ13Csc.
Both outliers for δ18Op are adult male decapitations. IX-20 is from the group of 48 crania associated with Structure 104. IX-10 was interred with another cranium in a pit adjacent to Structure 74. Both outliers have δ18Op (+18.1‰) identical to that predicted by using Daux and others’ equation to convert Lachniet and Patterson's water sample #68 from near Patulul (+18.1‰) and are likely foreigners, perhaps Tz'utujil captives.
Two outliers for δ13Csc (IX-2 and IX-4) also are adult male decapitations from the group of 48 crania buried near Structure 104. The structures associated with the other δ13Csc outliers (224-v and 270-i) are unknown. Both of the latter two individuals were adults, and 270-i had decapitation damage. Outlier 270-i has δ18Op (+16.0‰) consistent with that calculated for the OIPC oxygen isotope water composition for Tzolola’ / Tz'olojya (+16.0‰), but the individual could be an Iximche’ resident whose δ13Csc was caused by eating a unique juvenile diet. IX-2 and IX-4 have δ18Op (+14.8‰ and +14.6‰) closely matching that calculated using water sample #40 collected near the Mam community of Todos Santos Chuchumatán (+14.6‰). However, there is no ethnohistoric evidence of conflict between Iximche’ and the Mam, so they could be Iximche’ residents whose δ13Csc is due to unusual juvenile diets. The predicted δ18Op (+13.6) for outlier 224-v, based on conversion of its δ18Osc to the phosphate equivalent (equations in Locker et al. Reference Locker, Valdez, Breecker, Banner, Loewy, Luzzadder-Beach, Drake, Hyde, Trachman and Lewis2023) is similar to the predicted δ18Op (+14.1) for Lachniet and Patterson's water sample #51 collected near Todos Santos Chuchumatán. As with IX-2 and IX-4, the outlying δ13Csc of 224-v may simply reflect an unusual juvenile diet.
In summary, the values of δ18Op are indicators that some victims of decapitation originated from outside Iximche’ and some from Iximche’ itself. This is consistent with the local and foreign origins of human sacrifices found at highland Kaminaljuyu (Price et al. Reference Price, Burton, Wright, White, Longstaffe, Tiesler and Cucina2007).
Decapitations cannot automatically be equated with human sacrifice. Forgey and Williams (Reference Forgey, Williams, Rakita, Buikstra, Beck and Williams2005) and McAnany (Reference McAnany1995) argued that some crania unassociated with postcranial remains could instead represent ancestor veneration. However, isotopic analyses, burial taphonomy, ethnohistoric documents, and archaeological research at contemporary Maya sites whose political circumstances were comparable to Iximche’ argue against that interpretation. Postclassic decapitations have been reported for Ixlú and Macanché around the Petén lakes in northern Guatemala, where interethnic group conflict was endemic (Duncan Reference Duncan, Rakita, Buikstra, Beck and Williams2005). In some cases, cervical vertebrae remained with the cranium. They have also been reported at Chichen Itza in northern Yucatán and Toniná in Chiapas (Beck and Sievert Reference Beck, Sievert, Rakita, Buikstra, Beck and Williams2005; Ruiz Reference Ruiz, Tiesler, Suzuki and Pereira2021; Tiesler and Miller Reference Tiesler and Miller2023). Las Casas's (Reference Las Casas and de Tutela Bueso1958:152) description of the display of the heads of sacrificial victims—the supreme lord “put the heads of the sacrificed on some poles on a certain altar dedicated only to this, where they had these for some time, after which they buried them”—fits the characteristics of Structures 74 and 104 at Iximche’, yet the associated crania bear no evidence of having been placed on poles. These two structures are located adjacent to the southern corner of the western temple in their respective plazas, with both plazas featuring ballcourts that form their western corners. Duncan (Reference Duncan, Rakita, Buikstra, Beck and Williams2005) stated that interments west or south of architecture reflect either an attempt to hinder the soul's progress or an association with the underworld. Decapitations play important parts in the K'iche’ Popol Vuh (Tedlock 1996), where they are sacrificial acts associated with the underworld and ballcourts.
Las Casas (Reference Las Casas and de Tutela Bueso1958) reported a preference for war captives as sacrificial victims. Ethnohistoric accounts stated that leaders and warriors from enemy polities attacking Iximche’ were captured and executed. Iximche’ also had opportunities to capture sacrificial victims during military excursions outside its territory and when defending itself against internal rebels. Generally, younger adults in the prime of life with relatively little evidence of pathological conditions were selected for decapitation. This is logical if they were chosen as offerings to deities, were enemy warriors, or both.
Conclusions
Despite continuous warfare with surrounding groups and an internal revolt, Iximche’ was politically strong and underwent a period of rapid and aggressive expansion into the territories of its neighbors. It resisted all attempts at conquest by its enemies before the arrival of the Spanish. Physical characteristics of deposits of decapitations and their stable isotope compositions indicate that the skulls placed on display at Iximche’ came from warfare victims, defeated ballplayers, or community members who participated in local uprisings (Duncan Reference Duncan, Rakita, Buikstra, Beck and Williams2005): they were used to intimidate both outside enemies and the local populace (Forgey and Williams Reference Forgey, Williams, Rakita, Buikstra, Beck and Williams2005; Kennett et al. Reference Kennett, Masson, Serafin, Culleton, Lope, VanDerwarker and Wilson2016).
The skeletons buried at Iximche’ contain evidence bearing on their social status. Variations in stable carbon and nitrogen isotope ratios were observed in samples of ribs and teeth, but all individuals exhibited a signature of high reliance on C4-rich food resources, most likely maize. The lack of association between high maize consumption and the frequencies of dental caries, porotic hyperostosis, and periosteal reactions supports the interpretation that the skeletons, including those of decapitations, generally came from high-status individuals.
Significant dietary differences were not observed based on groupings by age, sex, or burial type, but several individuals had bone stable carbon isotope compositions that were exceptionally different compared to the majority, indicating that they ate substantially different diets than most others. The large range of δ13Csc and δ18Op in teeth, given the short time span during which burials occurred, may reflect diversity in geographic origin. However, there is no overlap in individuals with extreme values of both isotopes, which fails to support our hypothesis that many decapitated individuals were foreigners, although some were.
This study demonstrates that the diet of people buried at Iximche’ had a substantial maize component and that some may have come from other locations in the Highlands. The fragmented and mixed state of the remains—and considering that most individuals’ heads were separated from their bodies—means that reconstructions of diet and migration are less clear-cut than is desirable. Analysis of multiple teeth from the same individual, combined with the correlation of isotopic results for tooth enamel samples with bone collagen from the same individual, could help resolve these issues, as would a study of the isotopic ecology and foodweb of the Iximche’ region, which should be a future research priority. Analysis of bioapatite strontium isotope composition could potentially test the evidence from oxygen isotopes that the lives of both residents of Iximche’ and foreigners sometimes ended with the cut of an obsidian knife.
Acknowledgments
We thank the anonymous reviewers for their comments, which improved this article. Many students and other collaborators contributed to the skeletal analysis. This is Western's Laboratory for Stable Isotope Science Contribution #408.
Funding Statement
Grants from the Wenner-Gren Foundation for Anthropological Research Inc., the Foundation for the Advancement of Mesoamerican Studies Inc. (FAMSI), and the Faculty Research Fund at the University of Maine funded Iximche's skeletal analysis. The National Science Foundation (EAR 85-11549) provided funds for equipment to Peter Deines that enabled David Reed's analysis at Penn State. A FAMSI grant supported Robert Tykot's analysis of enamel carbonate at the University of South Florida. Operating (SSHRC, NSERC, Canada Research Chairs) and infrastructure (Canada Foundation for Innovation, Ontario Research Fund) grants to Christine White and Fred Longstaffe supported Karyn Olsen's analysis of enamel phosphate.
Data Availability Statement
George Guillemin's original notes, photographs, and drawings are archived at the Centro de Investigaciones Regionales Mesoamericánas (CIRMA) in Antigua, Guatemala. Archaeology and Ethnohistory of Iximché (Nance et al. 2003), which built on Guillemin's earlier publications about excavations at the site, includes extensive data on paleopathology and paleodemography. The skeletal material not used for isotopic analyses was rehoused in stable containers after analysis and returned to storage at Iximche’. The current status of the bones is unknown. Requests for access to notes, forms, and photographs related to skeletal analyses can be made to the corresponding author.
Competing Interests
The authors declare none.
Supplemental Material
For supplemental material accompanying this article, visit https://doi.org/10.1017/laq.2024.24.
Supplemental Text 1. Stable Isotope Analysis Methods.
Supplemental Text 2. Preservation Determination Methods.
Supplemental Text 3. Iximche’ Paleopathology.
Supplemental Figure Data 1. Sources for Figures 4, 6, and 8.
