La Paz is a major city located in the highlands of west-central modern Bolivia, which is 16°30′S, 68°09′W, 3,650 masl. Dry conditions in this region and the adjacent Andes Mountain range have aided in the preservation of desiccated organic material. Our maize samples were found within a woven camelid fiber pouch that was one of many offerings associated with the mummified remains of a child reported as being found in a chullpa (stone tower tomb) south of La Paz, Plurinational State of Bolivia around 1890 (Figure 1). Although it is impossible to know the exact location of the chullpa or the origin of the individual, the assemblage was acquired by the U.S. Consul to Chile and was sent as a gift to the Michigan State University Museum, where she was curated until 2019. Following a series of recent analyses, she has since been voluntarily repatriated to Bolivia (Lovis, 2023). Given the reported archaeological context of the mummy and her abundant offerings recorded in the MSU Museum intake records, it has been proposed that she might represent an elite member of society. However, the nature of the coarse fabrics, simple design motifs, and restricted color palette of her textiles would suggest a more common individual, perhaps with an affiliation with a local high-status group. The vast majority of the funerary towers that exist in a radius of 60 km south of La Paz city are made of mudbricks. To our knowledge, the only exception is an Inca ‘cushioned’ (i.e. rounded edges) stone and mudbrick tower located in Uypaca, in the nearby Achocalla Valley, 8 km southwest of the city center (Lovis, 2023). Because of its characteristics, this tomb most likely belonged to a local (possibly non-Inca) high-status group and was part of a more extensive funerary site. Although it is impossible to know the exact provenience of the mummy, it would not be unreasonable to suppose that she originated from this or another similar tomb in the Achocalla Valley. Regardless, in the decades following her death, community members periodically visited the chullpa and left offerings. These included the pouch full of well-preserved maize from which we collected and analyzed samples from two kernels.
Location and appearance of archaeological Bolivian maize (aBM).
(a) Map of maize sample collection and location information. (b) Photographs of kernels showing morphological characteristics.
A single Accelerator Mass Spectrometer (AMS) date was obtained from one kernel, yielding a radiocarbon age of 400±26 BP (D-AMS-027148; Zea mays kernel; corrected for fractionation, no δ13C reported). For the date 400±26 BP, the two possible calibrated age ranges are 1455–1516 cal AD (p=0.489) and 1538–1627 cal AD (p=0.511), with a median age of cal AD 1534 (Calibrated at 2σ with Calib 8.2 [SHCal 20.14 c]). Choice of the earlier age range is supported by other absolute dates on the assemblage (Lovis, 2019; Hogg et al., 2020): 447±45 BP (D-AMS-32518; Cucurbitaceae, δ13C=–27.6) with two possible age ranges of 1443–1504 cal AD (p=0.848) and 1594–1616 cal AD (0.152) with a median age of 1474 cal AD; and 510±20 (PSUAMS-5947; Camelidae fiber, δ13C=–18.9) calibrated to 1420–1454 cal AD (p=1.00) with a median age of 1438 (Both samples calibrated for southern hemisphere at 2σ with Calib 8.2 [SHCal 20.14 c]). Ages on the squash and maize are statistically identical at p=0.95 (t=1.70, x2.05=3.84, df=1) with a pooled mean of the two ages of 424±18 BP calibrated to 1453–1505 cal AD (p=0.769) and 1593–1617 (p=0.231). Results on the squash, with the highest probability between 1443–1504 cal AD, and camelid fibers with an age range of 1420–1454, as well as the pooled mean, reveal a greatest likelihood for a mid- to late 15th c age for the assemblage, consistent with the earlier of the two calibrated age ranges on the maize kernel.
Two associated maize kernels from the same pouch were subjected to molecular analysis. Genomic base modification patterns associated with ancient DNA (aDNA), including cytosine deamination damage patterns, have been identified. Observed position-specific substitutions at the ends of the sequence reads (e.g. C→T/G→A Handt et al., 1996; Krings et al., 1997; Hansen et al., 2001; Gilbert et al., 2003; Dabney et al., 2013; Figure 2a), as well as an increase in adenine (A) and guanine (G) residues at the 5´ end, concomitant with a reduction of C and T residues (Dabney et al., 2013; Briggs et al., 2007; Stiller et al., 2006; Figure 2b). From this, we evaluated all substitution types and read length (Figure 3), including read length range (Pääbo, 1989) and substitution rate types, which are associated with aDNA genome data (Prüfer et al., 2010). In total, these data are consistent with the radiocarbon results, supporting that (1) with the maize is in fact consistent with an archaeological biological sample, and (2) with based on the geographic origin of the sample, the kernels potentially descended from the period of the last Inca empire before the Spaniard Francisco Pizarro’s conquest of the Inca capital, Cuzco, in A.D. 1533 (Wilson et al., 2007).
Ancient DNA (aDNA) damage pattern of archaeological Bolivian maize (aBM).
(a) Cytosine deamination damage patterns for the combined BAM file of six aBM sequence samples and individual six aBM sequence samples. The position-specific substitutions from the 5’ end (red) and the 3’ end (green) of a read. The red line corresponds to C to T substitutions, the green line corresponds to G to A substitutions, and the blue line represents other types of substitutions. (b) Ancient DNA damage profile. The four upper plots show the base frequencies inside and outside of a read, where the open gray box corresponds to a read. The two lower plots show the position-specific substitutions from the 5’ end and the 3’ end of a read. The red line corresponds to C to T substitutions, the blue line corresponds to G to A substitutions, and the fade line represents other types of substitutions.
Archaeological Bolivian maize (aBM) DNA read length.
(a) Read length distribution of combined aBM bam files prior to hard-masked 5´ thymine and 3´ adenine residues within 5 nt of both ends.
Our study focused on identifying the location from which aBM originally came, establishing and explaining patterns of genetic variability of maize, with a specific focus on maize strains that are related to our current aBM samples. We were especially interested in exploring how these patterns of variability found in both archaeological and modern samples reflect the influence of cultural systems of behavior on both the spread of variants across the ancient landscape and selection for phenotypic traits by local populations.