In their critique of Emerson and colleagues (Reference Emerson, Hedman, Simon, Fort and Kelsey E. Witt2020), Hart and colleagues (Reference Hart, Lovis and Katzenberg2021) imply that we question their microbotanical and isotope research identifying early maize in the northern Eastern Woodlands. We do not. We accept their findings and cite their work as an example of the diverse regional and chronological variation that marks the history of maize in the Eastern Woodlands.
However, we take this opportunity to clarify several issues raised. Hart and colleagues cite the Icehouse Bottom and Edwin Harness sites as evidence for Middle Woodland maize. As noted in our article (Footnote 3, p. 224), reanalysis of these archaeobotanical samples was ongoing. Simon (Reference Simon2019), however, had reported preliminary results at the Plains Anthropological Conference as follows: two alleged maize samples from the Edwin Harness site returned δ13C values of <−21.0, indicating they were not maize, and two Icehouse Bottom samples returned δ13C >−10, but postdated AD 1000 in age. We would have been glad to share this information.
Hart and colleagues criticized us for ignoring the Ellege site sample date (1520 ± 70 RCYBP; Simon Reference Simon, Raviele and Lovis2014). We rejected this sample as being maize because the radiocarbon laboratory did not return a carbon isotope assay for it. A second alleged maize sample from the same feature returned a δ13C value of −25.3‰, indicating that it was not maize. We discount wood dates from associated contexts as unreliable. Simon (Reference Simon, Raviele and Lovis2014) demonstrated that with one exception, maize samples from early Late Woodland contexts (ca. AD 500–750) from western Illinois postdated standard dates obtained on associated materials by hundreds of years, suggesting that maize was intrusive in these feature contexts.
Compared to other Illinois data, the Late Woodland date (ca. AD 700) obtained on the Edgar Hoener site sample is “aberrant”—a conclusion corroborated by our extensive macrobotanical record (Simon Reference Simon, Raviele and Lovis2014:Table 5). We accept this date, however, and cite this as an example of early maize.
The application of stable isotope analysis to questions of maize consumption is particularly applicable where the pre-maize diet is almost exclusively C3 based and addition of maize, a C4 grass, would be evident in the δ13C values of skeletal tissues. Furthermore, because maize displays minimal geographic or landrace differences in isotopic values, comparisons are possible among different regions (Bender Reference Bender1968; Lowden Reference Lowden1969; Tieszen and Fagre Reference Tieszen and Fagre1993). The isotopic datasets available for the northeast; comparable methodology in sample preparation and analysis; and long-established, multifaceted research programs aimed at identifying early maize make the work of Hart and colleagues (Hart and Lovis Reference Hart and Lovis2013; Katzenberg Reference Katzenberg, Staller, Tykot and Benz2006) a particularly appropriate and valuable point of comparison to our own work.
Hart and colleagues took issue with our questioning bone carbonate data as reliable evidence for minimal maize consumption in southern Ontario. As summarized in Emerson and colleagues (Reference Emerson, Hedman, Simon, Fort and Kelsey E. Witt2020:245–246) and presented in detail by Harrison and Katzenberg (Reference Harrison and Katzenberg2003:228–230), the use of bone or enamel carbonate data for dietary reconstruction has been demonstrated, and lab pretreatment procedures have been developed and implemented to address concerns about diagenetic effects on bone carbonate (Harrison and Katzenberg Reference Harrison and Katzenberg2003:228). The (organic) bone collagen primarily reflects the isotope values of dietary protein, whereas (inorganic) bone and enamel carbonate reflect those of the whole diet. Because maize is a low-protein food, its consumption in small amounts would not be reflected in bone collagen but would be first reflected in bone or enamel carbonate (Ambrose and Norr Reference Ambrose, Norr, Lambert and Grupe1993; Harrison and Katzenberg Reference Harrison and Katzenberg2003; Hedman et al. Reference Hedman, Hargrave and Ambrose2002; Tieszen and Fagre Reference Tieszen and Fagre1993). As in our study area, Harrison and Katzenberg found that the C4 isotopic signature of maize is reflected in bone collagen at around AD 1000 in southern Ontario but not earlier (Harrison and Katzenberg Reference Harrison and Katzenberg2003:236). Also, as in our study area, the δ13C values of bone carbonate for some individuals who predate AD 1000 are slightly enriched in heavy carbon, but corresponding δ13C values of collagen are not (Harrison and Katzenberg Reference Harrison and Katzenberg2003:234, Table 2). This pattern is consistent with a mixed C3/C4 diet consisting primarily of C3 resources supplemented with a small amount of a C4 resource (e.g., maize). In the Northeast, this interpretation is supported by maize phytoliths and starch grains recovered from dated pot residues. Although, as Hart and colleagues have noted, we lack microbotanical data, we have an exceptionally robust macrobotanical record essentially lacking maize from pre-AD 900 contexts. Therefore, we are cautious about interpreting slightly enriched apatite δ13C values as evidence of maize consumption.
To address this concern, we analyzed enamel apatite carbonate δ13C for Middle Archaic and Middle Woodland individuals from several sites in Illinois. In nearly all cases, the enamel δ13C values are less enriched than that of bone carbonate for the same individuals, which both reflects a C3 diet and is consistent with the diet suggested by bone collagen δ13C results (Emerson et al. Reference Emerson, Hedman, Simon, Fort and Kelsey E. Witt2020:Supplemental Figure 2.3, Supplemental Table 2.3). If we accept Harrison and Katzenberg's interpretation of their bone carbonate δ13C levels as indicating maize consumption, then we might conclude that Middle Archaic people in Illinois consumed maize. We think that is unlikely. The divergence between bone carbonate δ13C and enamel carbonate δ13C from the same individuals, when preservation criteria for both collagen and apatite carbonate are met, raises the question of bone carbonate δ13C as a reliable indicator of slight maize consumption, particularly when quality of bone carbonate cannot be confirmed.
We recognize from macrobotanical and microbotanical evidence that the history of maize's chronological appearance as well as cultural and economic impact east of the Mississippi River varies dramatically. The seeming discrepancies between the Eastern Woodland macrobotanical and Great Lakes–Northeastern microbotanical evidence calls for additional research to help us understand the role that maize played in Native subsistence practices. We interpret the current evidence as an example of the tremendous variation in the history of maize among Native peoples across this region, and we call for our colleagues to continue to explore this variation.
In their critique of Emerson and colleagues (Reference Emerson, Hedman, Simon, Fort and Kelsey E. Witt2020), Hart and colleagues (Reference Hart, Lovis and Katzenberg2021) imply that we question their microbotanical and isotope research identifying early maize in the northern Eastern Woodlands. We do not. We accept their findings and cite their work as an example of the diverse regional and chronological variation that marks the history of maize in the Eastern Woodlands.
However, we take this opportunity to clarify several issues raised. Hart and colleagues cite the Icehouse Bottom and Edwin Harness sites as evidence for Middle Woodland maize. As noted in our article (Footnote 3, p. 224), reanalysis of these archaeobotanical samples was ongoing. Simon (Reference Simon2019), however, had reported preliminary results at the Plains Anthropological Conference as follows: two alleged maize samples from the Edwin Harness site returned δ13C values of <−21.0, indicating they were not maize, and two Icehouse Bottom samples returned δ13C >−10, but postdated AD 1000 in age. We would have been glad to share this information.
Hart and colleagues criticized us for ignoring the Ellege site sample date (1520 ± 70 RCYBP; Simon Reference Simon, Raviele and Lovis2014). We rejected this sample as being maize because the radiocarbon laboratory did not return a carbon isotope assay for it. A second alleged maize sample from the same feature returned a δ13C value of −25.3‰, indicating that it was not maize. We discount wood dates from associated contexts as unreliable. Simon (Reference Simon, Raviele and Lovis2014) demonstrated that with one exception, maize samples from early Late Woodland contexts (ca. AD 500–750) from western Illinois postdated standard dates obtained on associated materials by hundreds of years, suggesting that maize was intrusive in these feature contexts.
Compared to other Illinois data, the Late Woodland date (ca. AD 700) obtained on the Edgar Hoener site sample is “aberrant”—a conclusion corroborated by our extensive macrobotanical record (Simon Reference Simon, Raviele and Lovis2014:Table 5). We accept this date, however, and cite this as an example of early maize.
The application of stable isotope analysis to questions of maize consumption is particularly applicable where the pre-maize diet is almost exclusively C3 based and addition of maize, a C4 grass, would be evident in the δ13C values of skeletal tissues. Furthermore, because maize displays minimal geographic or landrace differences in isotopic values, comparisons are possible among different regions (Bender Reference Bender1968; Lowden Reference Lowden1969; Tieszen and Fagre Reference Tieszen and Fagre1993). The isotopic datasets available for the northeast; comparable methodology in sample preparation and analysis; and long-established, multifaceted research programs aimed at identifying early maize make the work of Hart and colleagues (Hart and Lovis Reference Hart and Lovis2013; Katzenberg Reference Katzenberg, Staller, Tykot and Benz2006) a particularly appropriate and valuable point of comparison to our own work.
Hart and colleagues took issue with our questioning bone carbonate data as reliable evidence for minimal maize consumption in southern Ontario. As summarized in Emerson and colleagues (Reference Emerson, Hedman, Simon, Fort and Kelsey E. Witt2020:245–246) and presented in detail by Harrison and Katzenberg (Reference Harrison and Katzenberg2003:228–230), the use of bone or enamel carbonate data for dietary reconstruction has been demonstrated, and lab pretreatment procedures have been developed and implemented to address concerns about diagenetic effects on bone carbonate (Harrison and Katzenberg Reference Harrison and Katzenberg2003:228). The (organic) bone collagen primarily reflects the isotope values of dietary protein, whereas (inorganic) bone and enamel carbonate reflect those of the whole diet. Because maize is a low-protein food, its consumption in small amounts would not be reflected in bone collagen but would be first reflected in bone or enamel carbonate (Ambrose and Norr Reference Ambrose, Norr, Lambert and Grupe1993; Harrison and Katzenberg Reference Harrison and Katzenberg2003; Hedman et al. Reference Hedman, Hargrave and Ambrose2002; Tieszen and Fagre Reference Tieszen and Fagre1993). As in our study area, Harrison and Katzenberg found that the C4 isotopic signature of maize is reflected in bone collagen at around AD 1000 in southern Ontario but not earlier (Harrison and Katzenberg Reference Harrison and Katzenberg2003:236). Also, as in our study area, the δ13C values of bone carbonate for some individuals who predate AD 1000 are slightly enriched in heavy carbon, but corresponding δ13C values of collagen are not (Harrison and Katzenberg Reference Harrison and Katzenberg2003:234, Table 2). This pattern is consistent with a mixed C3/C4 diet consisting primarily of C3 resources supplemented with a small amount of a C4 resource (e.g., maize). In the Northeast, this interpretation is supported by maize phytoliths and starch grains recovered from dated pot residues. Although, as Hart and colleagues have noted, we lack microbotanical data, we have an exceptionally robust macrobotanical record essentially lacking maize from pre-AD 900 contexts. Therefore, we are cautious about interpreting slightly enriched apatite δ13C values as evidence of maize consumption.
To address this concern, we analyzed enamel apatite carbonate δ13C for Middle Archaic and Middle Woodland individuals from several sites in Illinois. In nearly all cases, the enamel δ13C values are less enriched than that of bone carbonate for the same individuals, which both reflects a C3 diet and is consistent with the diet suggested by bone collagen δ13C results (Emerson et al. Reference Emerson, Hedman, Simon, Fort and Kelsey E. Witt2020:Supplemental Figure 2.3, Supplemental Table 2.3). If we accept Harrison and Katzenberg's interpretation of their bone carbonate δ13C levels as indicating maize consumption, then we might conclude that Middle Archaic people in Illinois consumed maize. We think that is unlikely. The divergence between bone carbonate δ13C and enamel carbonate δ13C from the same individuals, when preservation criteria for both collagen and apatite carbonate are met, raises the question of bone carbonate δ13C as a reliable indicator of slight maize consumption, particularly when quality of bone carbonate cannot be confirmed.
We recognize from macrobotanical and microbotanical evidence that the history of maize's chronological appearance as well as cultural and economic impact east of the Mississippi River varies dramatically. The seeming discrepancies between the Eastern Woodland macrobotanical and Great Lakes–Northeastern microbotanical evidence calls for additional research to help us understand the role that maize played in Native subsistence practices. We interpret the current evidence as an example of the tremendous variation in the history of maize among Native peoples across this region, and we call for our colleagues to continue to explore this variation.