Dos Cabezas Giants

University of San Diego Palynological Investigation of the Dos Cabezas Giants Patrick Scott Geyer, University of San Diego Thomas S. Larson, Texas A & M University Laura Stroik, Arizona State University Special note: This paper was presented at the 2002 European Meeting of the Paleopathology Association at the University of Coimbra, Portugal. For the 2003 peer reviewed journal article of the same title see the International Journal of Osteoarchaeology, Volume 13, Issue 5, September-October. Abstract Beginning in 1999, coprolite soil samples were collected from the skeletal remains of Moche Giants excavated at the archaeological site of Dos Cabezas, northern Peru. In 2000, a pilot palynological study was undertaken to ascertain if pollen was recoverable from them. The results of this study encouraged us to form a student team to continue the laboratory research. It was also decided to sample as many of the excavated skeletal remains from the Giant‘s tombs as possible. Following the 2001 field season a palynological study was performed on the three individuals (one adult Giant and two children) recovered from Tomb 3 in 1999. Pollen was extracted and analyzed. The combined results of this archaeobotanical study are significant and in one instance have led us into insights into the possible causes of death of these three contemporaneous internments. This report represents only the preliminary results of this ongoing research. Keywords: archaeology, pollen, Peru, Moche, coprolites, palynology, osteology, botany. Introduction In the summer of 1997, Dr. Christopher Donnan and his team began excavations at Dos Cabezas, a Moche site on the north coast of Peru. The site yielded a complex of 4 tombs. In 1999 the occupants of Tomb 3, a man of unusually tall stature, and 2 juvenile offerings, had coprolite samples taken from their pelvic girdles. These samples were taken and returned to the University of San Diego by the project’s osteologist and USD professor, Dr. Alana Cordy-Collins. The samples provided the material for a pilot study, undertaken in the fall of 2000, the goal of which was to determine if fossil pollen could be recovered from these and subsequent Moche burials. In the fall of 2001 and the spring of 2002, undergraduate researchers under the tutelage of Patrick Geyer fully extracted and analyzed the pollen that the coprolite samples yielded. The following paper contains the preliminary results of the pollen analysis.

It is important to acquaint the reader with the occupants of Tomb 3. The principal occupant, the one with whom the burial accouterments are associated with, is known to the excavators as Giant #3 or Pinocchio. Pinocchio is flanked to the east by a 10-year-old child referred to as Chico. Above the main chamber of this tomb lay the third individual, also a juvenile, nicknamed Enriquetta by the excavators. In obtaining a coprolite sample, sediment is taken from the pelvic girdle, specifically where the colon would have descended. These samples preserve pollen, and what is remarkable about this pollen is that it was contained in the digestive system of the individual at the time of death. By examining this pollen, the students were able to determine the plants each individual had consumed mere days prior to his or her demise. In identifying the pollen, the students had the ability to ascertain which plant foods, medicines, or perhaps even poisons, each person had ingested, so long as these items produced pollen. Additionally, background pollens from native plants are inadvertently consumed with daily food intake. As it was common for burials to be clothed and/or wrapped, background pollens present at time of burial can be useful in determining the plants associated with, and perhaps the methods of, internment in a particular culture. All of this information is vital to a study, such as the one undertaken at the University of San Diego in the 2001/2002 school year. Procedures In order to identify the pollen present in a coprolite or soil sample, a chemical separation from the sediment matrix, to which the pollen has bonded over time, must be performed. This pollen separation was accomplished using a series of slightly modified, standard palynological procedures (see appendix for details). To be successful these procedures have to be adapted to the particular soil type encountered (Schoenwetter, 2000). Briefly stated, sedentary agricultural deposits (to include coprolites) are rich in humic acids and are less resistant to strong base chemicals, while desert soil deposits are primarily alkaline and are thus less resistant to strong acids. The development of such a customized extraction procedure for the Moche burials was the primary focus of the initial pilot study. Results and Analysis

2

After the pollen was extracted and slides mounted, the pollen was identified and counted using a standard biological microscope. Each burial received a 200-grain count, which is the current acceptable minimum for accurate statistical representation of pollen present in a palynological study. Since no complete South American pollen keys nor reference materials are currently available, identifications of individual grains were based solely upon comparison to the currently published photographic plates of contemporary pollen from this subregion (Heusser, 1971 and Roubik, 1991). Thirty-three pollen types were identified (see table), although each was not present in all three of the individuals. This number also includes results which were not statistically significant. After the native use of each of these species was determined, the pollen was categorized as either a food, herb, or background pollen. During analysis, it became clear that many of the species had several native usage’s (Duke, 1994), and were therefore placed in the category for which they were most commonly used. For instance, prickly pear or cholla (Cylindropuntia) is used in food and drink, in making dye, and in strengthening adobe. However, it is most frequently used as a food source, and was therefore labeled as such. In order to facilitate understanding of the results, the findings of each individual will be discussed separately. Pinocchio Donnan and his team have determined the primary burial figure to be Pinocchio, deeming the other 2 individuals as “offerings” to this giant. Because Pinocchio has been identified as the principal occupant of the burial chamber, with all of the accouterments attributed to him, the results of his coprolite sample shall be discussed first. Of the 23 pollen species comprising the ‘food” category, Pinocchio’s sample contained a mere seven. The three most abundant pollen species in this category were guava (Psidium guajava), wild cucumber (Cyclanthera pedata), and white carrot (Arracacia xanthorrhiza). Although these results indicate little more than the contents of Pinocchio’s last few meals prior to death, they are consistent with some of the major food sources thought to be consumed by the Moche either due to their depiction on artifacts (Donnan, 1978), or from macrobotanical evidence that has since been recovered (Campbell, 1989).

3

Perhaps the most interesting results were found within the “herbal’ category. Out of the six herbal pollen species found, four of them were significant in Pinocchio’s sample. One of the four, retama (Cassia reticulata) is administered as an infusion for internal discomforts associated with the liver or kidneys or simply for upset stomach (Duke, 1994). Two of these four species, barbasco (Lonchocarpus nicou) and botaka (Nealchornea yapurensis) were used indigenously as fish poison (Duke and Martinez, 1994), while the fourth, soap berry (Sapindus saponaria) is also a known poison. Obviously, pollen samples cannot determine the presence of fish in the colon, but if Pinocchio had eaten fish around the time of his death (not an unlikely conjecture, as fish were a staple in the Moche diet), then the presence of a fish poison would be explained. However, not the pollen of one of the plants used as a fish poison. These poisonous plants are scattered upon the waters for the purpose of stunning the fish (Plotkin, 1993). Chemicals, pollen and other residue from the plants could conceivably contaminate the scales, gills and digestive tract of the fish. Yet, prior to human ingestion these areas of the fish would be removed and along with them most if not all of the pollen residue. In addition, Donnan and his team determined that 4 of the 5 Moche giants, each exhibiting identical bone deformities, were buried, and presumably died contemporaneously. Armed with this knowledge, one might venture that these 4 individuals’ deaths were not all natural. This suggests that the Moche could have perceived these poisons as harmful to humans (although this might have been an incorrect perception on their part) and purposely incorporated them into the last meals of these individuals with the intent of causing a premature death. The researchers ask that this hypothesis at least be considered. The plurality of poisonous herbs present in the remains of these three individuals further suggests that they might have been administered in combination along with other herbs (e.g. Brunfelsia grandiflora) for the purposes of concocting a “curare” similar to those documented on the north west coast of South America (Plotkin, 1993). Well over 87 curare herbs have been documented and catalogued (Schultes, 1990). Of the six herbals present in the coprolites of Tomb 3, two, barbasco (Lonchocarpus nicou), and soapberry (Sapindus saponaria), are known as human poisons (ibid.). However, the Jamandi tribe of Brazil utilize soapberry (Sapindus saponaria) along with other well known poisonous

4

species in order to make their arrow poison (Schultes, p 308), which thus qualifies this species as an indicator of the presence of a curare. All three of the individuals from Tomb 3 contain trace amounts of soapberry, while Pinochio with 4% contains the most. Four pollen species were considered background material, meaning that they were not ingested as food, medicine, or drink, and their presence in the sample is most likely the result of accidental ingestion or after-burial contamination. Of the four background pollens identified, only one, sedge (Cyperus proxlixus), appeared as significant in Pinocchio’s coprolite. As Pinocchio’s social status suggests that he was wrapped in several layers of textiles upon interment, the lack of background pollen, can be easily explained as being due to minimal outside contamination. These wrappings could also explain the presence of sedge. Sedge is used as a rope to bind cotton wrappings to the burial (Donnan, 1995) and could have been introduced as a contaminant in this way. One would wonder why the pollen from the cotton wrappings wouldn’t be present as well, but plant fibers used in weaving textiles would (unlike those used in braiding rope) undergo repeated washings, which would in the process remove any contaminate cotton pollen. Chico The sample taken from Chico, the offering buried adjacent to the right leg of Pinocchio, contained 13 significant pollen species. Of those 13 species, nine of them were classified as food. Zea mays, or maize, comprised nearly 35% of the total pollen identified from Chico, contrasting with the 3% found in Pinocchio’s sample. This discrepancy is not uncommon, as there is a lack of patterning present among Moche burials of varying social status, gender, etc. with respect to amount of maize (Gumerman, 1994). Gumerman’s study, however, focused on burial offerings rather than coprolite samples, so it is possible that, although the burial offerings of maize may have no pattern, the actual food substances eaten before death may be consistent with social status, etc. There was also a comparatively low percentage of guava pollen (Psidium guajava) in Chico relative to Pinocchio. In Pinocchio’s sample, guava was the most abundant pollen, comprising roughly 18% of the total pollen, but guava only accounted for 9% of the total pollen for Chico. This incongruity may indicate a difference in diet between those of an elite status and those of a lower class, for example, a person given as an offering in a burial. Aside from the discrepancy in maize and guava 5

consumption between the two individuals, the relative amounts of the other food substances show a similar pattern. Three of the six pollen species identified in this study and classified as herbal’s were significant in Chico’s sample, botoka (Nealchornea yapurensis) being present in the largest amount. It was also the most abundant herbal identified in Pinocchio’s sample, perhaps further supporting the notion that the intended use of the plant may have been for poisoning these individuals. Of the four background pollen species, only one, bulrush (Scirpus californicus), was found in significant amounts in Chico’s sample. Bulrush is used in making mats and boat material, as well as funerary wrappings. Taken together with the knowledge that Chico was buried alongside Pinocchio, one could conjecture that the pollen from Pinocchio’s outer “cane wrappings” (Donnan, 1995) contaminated Chico’s coprolite sample. It is expected that this pollen would not have been found in abundance in Pinocchio’s sample due to the numerous layers of the burial wrappings, but it’s presence in Chico, who was not carefully interred, is more likely. However, when we consider that bulrush composes 16% of Enriquetta’s sample, a contrary interpretation is that the 2.5% of bulrush present in Chico’s sample is merely contamination raining down from above. What needs to be determined is if bulrush (Scirpus californicus) is in fact the plant species being utilized for the “cane wrappings”. Enriquetta Within the sample taken from Enriquetta, only 12 of the 33 pollen species identified were deemed significant. Enriquetta was buried above the burial chamber that houses both Pinocchio and Chico and therefore could not have been contaminated by the pollen from their burials. As a result of her unique burial location, the pollen found in her grave should be a good indication of the food substances a typical offering may have ingested in the last days of his or her life. It should be noted that the excavators consider Enriquetta a burial offering due to the lack of burial goods and the nontraditional internment position. Of all three identifications, the analysis of Enriquetta’s burial recovered the greatest number of maize (Zea mays) pollen grains. As noted above, this may not be

6

unusual, due to previous evidence of the lack of consistency in abundance of maize in Moche burials. It should be noted, however, that both Enriquetta and Chico, the two burial offerings, did contain much more maize than the main burial, Pinocchio (69 and 68 grains for Enriquetta and Chico, respectively, and a mere 7 for Pinocchio). Again, as this phase of the study, although currently limited, sampled the burial coprolites rather than the grave goods, it may be stated that more maize was traditionally fed to the burial offerings than was consumed by the main burial. This does not necessarily refute the previous evidence, but it does exhibit patterning in amount of maize in Moche coprolites. Additionally, Enriquetta’s burial contained considerably less carrot, chilipepper and guava (as did the burial of Chico), than did Pinocchio’s. This may be additional evidence of a dietary differentiation between the upper class and those who accompanied them to their grave. In addition, the absence of a species of plant, wild cucumber (Cyclanthera pedata), present in both Pinocchio and Chico (amassing 15% and 10% of the samples, respectively), yields an interesting hypothesis. It is possible that the presence of this pollen in Chico was due to contamination, however, the evidence shows that Pinocchio was tightly wrapped and such a large contamination is unlikely. The absence of this pollen in Enriquetta, and its presence in both Pinocchio and Chico, may indicate a special relationship between the two. Chico could have been a servant to Pinocchio, and therefore ate similar foods. Or, Chico and Pinocchio may have shared some of the same foods that simply would not have been available to a person of Enriquetta’s status. This may also help to explain the physical closeness the two share in the grave, as compared to Enriquetta, who was buried on the roof of the tomb. Enriquetta’s coprolite sample yielded only one significant herbal pollen: botoka (N. yapurensis). As mentioned above, this species was used as a fish poison and was the most abundant herbal pollen in all three. The presence of this pollen in significant amounts further begs the question: was it the intent of the Moche to poison these people, or did they simply ingest a large amount of fish prior to death? Another interesting aspect of the presence of this pollen is that it was present in amounts relatively proportionate to the individuals’ body size. Pinocchio, the largest of the three, contained nearly twice as much

7

botoka (N. yapurensis) as Chico and five times as much as Enriquetta. This information yields further evidence to an argument for intended poisoning, as the botaka could have been fed in specific dosages determined by the stature of the individuals. However, this proportionality could simply correspond to the greater amount of fish eaten by a larger individual. As is true of the rest of the results, further data is necessary to draw a definitive conclusion. Enriquetta’s sample contained the most background pollen species of the three burials. While only one background species was identified in both Chico and Pinocchio’s samples, Enriquetta’s sample was composed of three. And not only did she have the most background pollen, but it was also found in relatively large amounts, comprising more than 25% of her total pollen count. One reasonable explanation for this high percentage is her relative close proximity to the surface, as compared to either Chico or Pinocchio, or her body could have simply been left exposed longer than the other two burials. In fact, she may have died at the same time as Pinocchio and Chico, but was not buried until after Pinocchio was buried and his tomb completed. Conclusion The analysis of the coprolites taken from these three burials yielded some interesting and unexpected information. It could be argued that this study raised more questions than it answered. For example, were these three individuals poisoned, and if so, why? Is there a significant difference in the diets of the elite’s from that of the commoners? Why is there far more background pollen present in Enriquetta’s sample than in Pinocchio and Chico’s combined? Building on this, can the presence of these pollens in the coprolites be explained as merely a matter of contamination (control samples were recovered from the tombs, but were not made part of this preliminary study)? As of now, we are only able to look at a microcosm of Moche society. With such a fleeting glimpse back through the ages, one is left with no complete answers, only the desire to continue the research and learn more about how plants were used as foods, clothes, drinks, dyes, drugs, and perhaps even poisons among the Moche. The only way that these, and any other questions raised by this study, can be answered is by further coprolite studies of the tombs. In fact, it is the opinion of these authors that only with a complete

8

palynological study of the Moche culture, from the lowest commoners to the highest kings and priests, would one be able to confidently reconstruct the lives of these people who inhabited the north coast of Peru nearly two millennia ago. Special acknowledgements go to Mr. and Mrs. K. Clampit and Dr. A. Orona for their continued support. Additional credit must be given to our exceptional group of research interns: J. Alan, S. Delellis, A. DeYoung, J. Williman and R. Mandell, for their hard work.

9

Readings Cited: Campbell J. 1989. Historical Atlas Of World Mythology, Volume II The Way Of The Seed. Part 3 Mythologies Of The Primitive Planters: The Middle And Southern Americas. Harper & Row. De Feo V. 1992. Medicinal and magical plants in the northern Peruvian Andes. FITOTERAPIA Volume LXIII, No. 5. Donnan CB. 1978. Moche Art Of Peru. Regents of the University of California. Donnan CB. 1995. Moche Funerary Practice. In Tombs For The Living: Andean Mortuary Practice. Dillehay TD (ed.). Dumbarton Oaks. Duke JA and Vasquez R. 1994. Amazonian Ethnobotanical Dictionary. CRC Press. Gumerman G. 1994. Corn For The Dead: the significance of Zea mays in Moche burial offerings. In Corn & Culture in the Prehistoric New World., Johannessen S and Hastorf C (eds.). University of Minnesota Publications. Heuser CJ. 1971. Pollen And Spores Of Chile.University of Arizona Press. Plotkin MJ. 1993. Tales Of A Shaman’s Apprentice.Viking. Roubik DN. 1991. Pollen And Spores Of Barro Colorado Island. Missouri Botanical Garden. Schoenwetter J and Geyer PS. 2000. Implications Of Archaeological Palynology At Bethsaida, Israel. Journal of Field Archaeology Volume 27, Number 1, (Spring). Vasquez R. 1994. Plant As Utiles De La Amazonia Peruana I. USDA publication.

10

Table 1 Moche Pollen Counts 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Species/type names Cultivar (foods): Agave americana Amaranthus caudatus Arachis hypogaea Arracacia xanthorrhiza Bunchosia armeniaca Capsicum baccatum Chenopodiaciea Cylindropuntia Cyclanthera pedata Cucurbita maxima Inga Feuillei Lagenaria siceraria Lepidium meyenii Lucuma bifera Lupinus mutabilis Manihot esculenta Oxalis tuberosa Persea americana Phaseolus lunatus Psidium guajava Solanum tuberosum Vicia Faba Zea mays

English/native names

Native usages

24 25 26 27 28 29

Herbals: Brunfelsia grandiflora chiric sanango Cassia reticulata senna/retama Eethroxylon coca coca/ipayu/huangana Lonchocarpus nicou barbarosa Nealchornea yapurensisbotaka/huiri caspi Sapindus saponaria soapberry

hallucinagen/fish poison antibiotic/dye/purgative/diuretic stimulant/narcotic/antirheumatic fish poison/arrow poison fish poison soap/arrow poison

30 31 32 33

Background pollens: Cirsium alissimum Artemisia tridentia Cyperus eragrostis Scirpus californicus

rope mats, boat material

mexico/maiguey/agave edible fruit/fiber/decongestant inca wheat/kiwicha edible starch/dye peanut edible seed white carrot edible vegetable peanut butter fruit/ciruelaedible fruit/drink/flavoring chili pepper/aji edible vegetable/antirheumatic quinoa/quinua/kiuna edible starch/soup/flat bread prickly pear/cholla edible fruit/drink/dye/fiber wild cucumber/achochaedible vegetable winter squash/zapallo edible vegetable/container pacay/pacae edible vegetable bottle gourds/cucuzzi edible vegetable/container Peruvian ginseng/maca edible fruit/jam/antirheumatic egg fruit/lucumo edible fruit/wood lupine/tarwi edible starch yuca/cassava/manioc edible starch/extract fish poison occa/oca edible starch/antiinflammatory avocado/palta edible fruit/antidysenteric lima bean/pallar edible starch guava/guayaba edible fruit/diuretic potato edible starch/extract poison fava bean edible starch maize edible starch/burial food

tall thistle sage brush tall flat sedge bulrush/totora

B1 B2 B3 Enriq Chico Pino 1 1 3 6 1 8 6 17 4 3 2 2 2 8 8 3 9 1 2 20 30 1 1 1 4 1 1 3 2 3 1 7 1 1 5 10 4 1 1 4 4 3 13 18 35 3 1 5 2 69 68 7

1

5

7 2

16 7

1 5 1 5 33 8

3 5 9 200

1 9 2 6 200

5 2 17 32 Unidentifiable 1 Pollen Sum 200

11

Appendix Pollen Extraction Procedure Step One The first step in the process is termed deflocculation. It is the goal of this step to loosen and break the bonds that hold the pollen to the soil. 1) After preparing and labeling the appropriate number of 15 ml test tubes, 3 ccm of sample is scraped from the envelopes into each of the tubes. 2) 10 ml of a 20% Trisodium Phosphate (TSP) solution is then added to each tube in order to break down the physical bonds that cement the pollen to the soil matrix. 3) The contents of the tubes are then stirred to form a homogeneous solution. 4) The samples are now let to set for 3-5 days and are stirred occasionally. Step Two The next step is the reduction of the organic fraction. 1) The samples are first centrifuged at 3000 rpm for 5 minutes and decanted. They are then rinsed with distilled water to remove any remaining TSP and again centrifuged and decanted. 2) The volume in the tubes is now brought back up to 10 ml with a 20% solution of NaOH. The resultant mixture is stirred vigorously and placed in a boiling bathe of distilled water for 10 minutes. 3) Hot water is harvested from the bathe for the subsequent rinsing process. 4) Shortly after the tubes are removed from the bathe they are centrifuged and decanted. Samples are then rinsed with the hot distilled water, stirred, centrifuged and decanted again until the liquid portion clears. 5) After repeated rinses the samples are allowed to sit until the following week; a second boil might be needed. Step Three The third step is an optional repetition of Step Two, this is done until the decanted liquid is clear. Step Four The next step in separating out the pollen so that it can be analyzed is termed swirl separation. 1) The mixture in the tubes is again stirred to create a homogeneous mixture. 2) The mixture is then decanted through a #80 screen (180 mu) into fresh labeled 15 ml test tubes. 3) This process is repeated at least one more time to separate as much of the matrix from the large mineral fraction of the sample as possible. 4) The volumes of the tubes are now brought up to a level of 15 ml with distilled water. 5) The tubes are then centrifuged for five minutes at 3000 rpm. 6) The liquid from the tubes is then poured off as the matrix containing the pollen is forced to the bottom of the tube during centrifuging. Samples with multiple tubes are combined into a single tube. 7) Each pollen matrix should now have its own individual tube (e.g. three tubes for three samples). Step Five The fifth step in isolating the pollen is termed heavy liquid separation. This step uses a dense liquid to float the lighter pollen grains, while the heavier soil is separated out as it sinks to the bottom. 1) Samples are centrifuged at 3000 rpm for five minutes and then decanted to remove the unneeded water. 2) Ethanol is now added to the matrix as a drying agent in place of the water and the sample is stirred. 3) The mixture is then centrifuged and decanted with the ethanol solution. 4) After allowing for drying the test tube is filled with a solution of ZnCl2 with a specific gravity of 2.0, using three times as much volume of ZnCl2 as remaining matrix. 5) The solution is stirred vigorously until it is homogenous, then centrifuged. 6) While decanting, the top 2/3 of the liquid is poured off into a clean labeled 15 ml tube and the new solution is then diluted with water using three times the volume of water as the solution. It is now centrifuged and decanted. This then allows the recovered pollen to drop back down out of suspenion. 7) The new polleniferous matrix is then transferred with ethanol into freshly labeled dram vials for storage. Step Six Step six is the final in the progression. The pollen can now be mounted (glycerol) and stained (fuchsin) on microscope slides so that individual grains can identified and 200 grain per sample counts made.

12