Identification of trace metals and potential anthropogenic influences on the historic New York African Burial Ground population: A pXRF technology approach

The New York African Burial Ground (NYABG) is the country’s oldest and largest burial site of free and enslaved Africans. Re-discovered in 1991, this site provided evidence of the biological and cultural existence of a 17th and 18th Century historic population viewing their skeletal remains. However, the skeletal remains were reburied in October 2003 and are unavailable for further investigation. The analysis of grave soil samples with modern technology allows for the assessment of trace metal presence. Portable X-ray fluorescence (pXRF) spectrometry provides a semi-quantitative and non-destructive method to identify trace metals of this population and in the surrounding environment. Sixty-five NYABG soil samples were analyzed on a handheld Bruker Tracer III- SD XRF with 40 kV of voltage and a 30μA current. Presence of As, Cu, and Zn can potentially decipher the influence of the local 18th Century pottery factories. Elevated levels of Sr validate the assumed heavy vegetative diets of poor and enslaved Africans of the time. Decreased levels of Ca may be due in part to the proximity of the Collect Pond, the existing water table until the early 19th Century, and Manhattan’s rising sea level causing an elevated water table washing away the leached Ca from human remains. These data help us reconstruct the lives of these early Americans in what became New York City.

Grave soil, the decomposed substrate of living organisms, is characterized by the presence of mineral remains present in the living organism and as such is used as a resource for forensic and anthropological studies. By 1664 almost 300 enslaved Africans would call new Amsterdam their home. Over the next two centuries almost 17,000 of these enslaved people and their descendants would come to be buried in the "Negroes Burial Ground" located on the southern tip of the island of Manhattan. Human grave soil remains an understudied resource that could provide a better understanding of the lived experience of the interred human. Human remains allow researchers to investigate a population that once lived but has left behind little to no record of its existence. The rediscovered burial ground, later renamed the New York African Burial Ground (NYABG), represents a high profile, partially excavated gravesite located in lower Manhattan referred to as the most significant archaeological find of the 20 th Century 1 . This burial ground contains a population of free and enslaved Africans who were buried between the 1640s and the 1790s in what was then known as the New Amsterdam Colony. Currently, human remains have been investigated using genomic ancient DNA sequencing methods and assessments of microbial community. These techniques have provided novel insights into human ancestry, gender identification, and historical infectious disease prevalence [2][3][4] .
In this investigation, we used pXRF spectrometry to detect trace metals found in 64 human-associated burial samples and their corresponding concentrations. PXRF technology is typically used in a number of environmental capacities including archaeology field studies, art conservation, and waste management. It provides a rapid, and inexpensive way to detect elements in a short time without destructive and time-consuming acid digestion [5][6][7] . Additionally, studies have demonstrated the accuracy of pXRF suggesting its efficacy in trace metal detection to be consistent with chemical analyses such as ICP-MS (inductively coupled mass spectrometry) and AAS (atomic absorption spectrometry) [8][9][10][11][12][13][14][15][16][17][18] . This technology (pXRF) was chosen for its non-destructive nature and the finite availability of these 17 th and 18 th -century grave soils samples.
The main sources of trace elements for this investigation are decomposed human debris, their adornment and artifacts, the presence of a major water source, and multiple pottery factories. The time period for internment of NYABG inhabitants and anthropogenic contributions is ~1640 -late 1790s. Knowledge of this time period allows for determination of geologic time and subsequent soil profile based on an historical map ( Fig. 1) 19 .
Previous studies have observed the impact of human remains on the presence of trace metals in soils using skeletal remains and generally using invasive/destructive analyses. As such, chemical studies are used to complement noninvasive studies using pXRF when we want to better understand the interactions between humans and their surrounding environments [20][21][22] . However, with pXRF technology we are able to identify elemental species and quantitatively analyze the concentrations for identification of potential factors affecting this historical population. To our knowledge, little research has explored residual trace elements strictly in grave soils, particularly 400-year old historic grave soils, to reconstruct the as-lived experiences of the population, without complete destruction of the sample.
At the time of the 1991 NYABG excavation, the skeletal and archaeological remains of 419 individuals were discovered during the initial plan to build the 30-story Ted Weiss Federal Building at 290 Broadway. The excavation site is located 2 blocks away from City Hall with Duane St. along the north border, Elk St to the east, Reade St to the south and Broadway to the west. While historical maps show the burial ground spans 6.6 acres, its vast existence was only revealed at the time of excavation. The erection of the Ted Weiss building required sub-basements that reached 30 feet below ground. Descriptions of each burial were provided in the initial reports generated shortly after the excavation by Howard University and General Services Administration (GSA). These reports detailed the skeletal biological, archaeological, and historical analyses of the NYABG 23 . The greatest depths of the coffins were recorded at 25 feet. By exploring the NYBAG soil samples we are able to extrapolate living conditions and possible contaminants of the interred individuals through trace metal analysis. Consequently, the findings of this investigation enrich the fields of archaeology, anthropology, and soil chemistry. This paper describes a pXRF approach for the analysis of trace metals (i.e. As, Ca, Cu, Sr and Zn) in cadaveric NYABG soils and establishes a feasible protocol for similar research projects involving historic human remains. This paper also aims to describe the potential source of quantifiable trace metals in relation to anthropogenic influences during the time of internment.

Results
Using the Bruker Tracer pXRF, we were able to detect upwards of 15 elements per sample. These elements include Fe, Cr, Pb, K, Ti, Rb, Mg, Mn, Ni, Br, Rb, Cd and the five reported in this study, As, Ca, Cu, Zn, and Sr. We only reported on the five elements that had accurate or certified values from the NYABG samples, control samples and NIST SRMs data sets. The most abundant element found in NYABG samples is Sr with an average concentration www.nature.com/scientificreports www.nature.com/scientificreports/ of 307.2 μg/g and the least abundant element is Ca with an average concentration of 1.9 μg/g. The average NYABG sample concentration values for the Cu, Zn, and As are 82.3 μg/g, 117.9 μg/g, and 11.5 μg/g respectively. The most abundant element found in NRCS control samples is Ca with an average concentration of 2356.4 μg/g and the least abundant is As with an average concentration value of 4.9 μg/g. The average NRCS control sample concentration values of Zn, Cu, and Sr are 41.2 μg/g, 17.8 μg/g, and 47.2 μg/g respectively (Table 1).
Comparisons between the NYABG and control samples for each element can be seen in Fig. 2. Scatterplots are shown for As, Cu, Zn, Ca, and Sr respectively. The depth for our control samples ranged from approximately 0.6 feet below ground (FBG) to 6 FBG and for the purposes of this investigation are referred to as surface depths as compared to our NYABG samples or buried depths which were found between 10 and 25 FBG. At buried depths As is present in low concentrations (<10 µg/g). At surface depths As presence is also low in concentration (i.e. <10 µg/g excluding outliers). The comparison shows no significant difference between the control and NYABG soil As samples. At buried depths Cu is present in varying concentrations from 19-544 μg/g. At surface depths Cu is present in low concentrations (<26 µg/g). The comparison shows a significant difference between the control and NYABG soil Cu samples. Similarly, a significant difference is also present in Zn, Ca and Sr as shown in Table 2.
Statistical significance was determined using the Mann Whitney U-test. This method was performed using a two-tailed test to show the differences between medians of two independent samples which were not expected to display a normal distribution. An exact p-value (P < 0.0001) for Ca, Cu, Zn and Sr shows that the values of NYABG samples and values of control samples are indeed as hypothesized, significantly different. The values of NYABG samples and values of control samples of Ca, Cu, Zn and Sr are statistically significant in relation to depth. Conversely, the NYABG and control samples for As are not statistically significant in relation to depth.
To discern the relative elemental composition (Ca, Cu, Zn, As, and Sr) of NYABG grave soil samples generated using pXRF technology the concentration values were graphed to visualize the correlation of elemental concentrations and depth (Fig. 3). All depth information retrieved from the GSA archaeological reports 24 were reported in average mean feet above sea level (AMSL). Statistical significance was determined using linear regression analysis displayed by the regression coefficient. All regression data show no linear correlation between element concentration and depth. P-values for each element (all p > 0.05) suggest there is no significant depth distribution of trace metals. Consequently, elemental deposits originate from the NYABG time period.

Discussion
We use pXRF analysis to non-destructively assess the trace metal composition of cadaver-associated grave soil derived from the cranium, torso, and lower body of previously interred enslaved African Americans living in New York City during the 17 th and 18 th centuries. We found evidence of potential influences from 18 th Century pottery factories, the rising sea level of New York City, and the vegetative diets of the NYABG population. Previous heavy metal research on the NYABG has been conducted on the skeletal remains. The NYABG grave soil samples have never been explored. Limited studies on elemental concentrations of the skeletal remains have been performed on this population including elemental signature analysis (ESA) on enamel, dentine, and cementum of teeth using laser ablation, inductively coupled plasma-mass spectrometry (ICP-MS), and 87 Strontium to 86 Strontium isotopes 25 . However, NYABG grave soils have never been systematically studied for their trace metal composition, until now. This research provides new and complementary insights into the source of trace metals and their ability to serve as indicators into the lifestyles of this historical community. We contribute elemental concentration data for five elements (Ca, Cu, Zn, As, and Sr), some of which support previous ESA findings, about the NYABG population and the surrounding environment 25 . Arsenic, copper, and zinc. We infer that the presence of As, Cu, and Zn concentrations in the NYBAG samples reflect the anthropogenic contribution of 18th Century pottery factories, two of the first kilns built in the Colonies. This inference is based on the approximate locations of the factories and the chemical gradient of Arsenic, Copper, and Zinc that show higher concentrations closer to the pottery production sites. The Crolius Pottery factory and Corselius Pottery stoneware factory were opened in 1730 and 1735 respectively. The Corselius factory later came to be known as the Remmey and Crolius Pottery factory in 1742 26 . These pottery factories along with a potash factory were located just north of the northern border of the excavation site along today's Duane St. 27 . They regularly used the site as a dumping ground for their kiln waste as they saw this area as "undesirable land" 1 . We found burials with the highest concentrations of As, Cu, and Zn in the northern section of the excavation site in close proximity to Duane St. In a 1998 study by Hirtle et al., of kiln emissions measured at 50 sites, the presence of As, Cu, and Zn were always present 28  www.nature.com/scientificreports www.nature.com/scientificreports/ heavy metals found in contaminated soil at industrial waste sites. These metals can be detected years after they've been introduced into the environment since soil serves as a reservoir and they cannot be biodegraded [29][30][31] . In the case of Cu, we see increased concentrations varying throughout the burial ground. In few burials we do not see increased concentrations of As or Zn, particularly in burials where copper artifacts or adornments were reported in the initial reports 24 . Therefore, we posit the existence of these elements is evidence of kiln waste and burial artifacts including copper rings, buttons, and decorations of interred individuals deposited during the and 17 th and 18 th Centuries.    www.nature.com/scientificreports www.nature.com/scientificreports/ Calcium. The abundance and availability of Ca, unlike the other metals analyzed in this investigation, is highly dependent upon other geochemical processes occurring in soil. More frequently, Ca is present in its soluble form and adsorbed to the soil complex as a positively charged cation 32 . Additionally, Ca and Mg cations leach most frequently in soils leading to a decreased concentration at varying depths 33 . The factors affecting the availability of Ca in soil are pH, cation exchange capacity (CEC), cation competition, sodium content, and sub-soil/ parent material. Excess sodium (Na) competes with Ca and other cations in areas composed of shale or sandstone, the soil type described in the initial reports of the NYBAG 24 , resulting in lower levels of Ca 32 . Calcium, the least abundant element found in NYABG samples, may exist in low quantities as a result of one or several of the previously mentioned factors. We suspect Ca leached from NYABG skeletal remains to the soil and were desorbed by the wetland environment of the New Amsterdam Colony of the 17 th and 18 th Centuries. Historical maps of NYC show the Collect Pond with streams flowing east to the East River and west to the Hudson River 34 . This body of water was New York's main water source and responsible for much of the wetland environment until 1811 when it was filled in by leveling Bayard's Hill 35 . By the 1990s we see a similar contribution of water beneath Lower Manhattan, a result of the rising sea level due to increasing temperatures in the atmosphere. It has been reported that one foot of sea-level rise has occurred since 1900 due in part to warming of the ocean water 36 . We suppose this rising water table may be desorbing Ca from the soil and washing it away west to the Hudson River or the East to the East River located approximately 1.5 miles away.
Strontium. Sr is commonly used when investigating ancient populations to infer characteristics about their biological and cultural existence. Generally, in studies of human bone remains, Sr is studied using isotope analysis. This method was used to generate Sr data reported in the initial reports on the NYABG 26 . In a 2000 study by Beard and Johnson, they report Sr isotope composition measured from skeletal remains can be used to infer their geographic habitat 37 . Biogenic 87 Sr/ 86 Sr analysis of enamel, dentine, and cementum was studied to distinguish the New World Africans from the Old-World Africans buried in the NYABG. As expected, the juveniles were born in New York and the majority of adult individuals, were born in Africa 26 . Additionally, researchers used ESA to explore Sr and Ba to Ca ratios to evaluate trophic level of diets. It has been well documented that ratios of Sr and Ba in relation to Ca are indicators of meat presence in diets 26,[38][39][40][41][42][43][44] . It is fundamentally important to understand the human and biological culture of the NYABG to archaeologists and anthropologists 45 . Under normal metabolic circumstances Sr/ Ca ratios help researchers to investigate the tendency of human populations towards meat vs. vegetable consumption 44 . Elevated Sr suggests the interred individuals of the NYABG maintained a largely vegetative diet as opposed to a meat-based diet [37][38][39][40][41][42][43][44][45] . This aligns with the historical assumptions about the socio-economic status of these people. As enslaved or poor free Africans in the New Amsterdam Colony, this population lacked access to robust, protein-rich foods. Consequently, they consumed only the leafy and root crops they cultivated from the surrounding land.

Conclusion
We demonstrate the utility of pXRF technology to successfully and conservatively detect trace elements in burial soil allowing for the deduction of major events, daily routines, and cultural influences in the lives of the enslaved population of 17 th and 18 th Century New York City. PXRF technology helps us extrapolate elemental signatures of the remaining fraction of the burial ground and present-day lower Manhattan that is now covered with concrete and cannot be sampled. Our conclusions indicate heavy metal distribution in the NYABG, indicators for predominant vegetative diet in a historical population, and possible signs of local climate change. This research establishes evidence for the value of pXRF technology to reconstruct past human activities in soil. Understanding these findings is relevant for similar studies of cemetery soil investigations, ancient human populations, forensic science, and anthropology. PXRF technology provides a mechanism for discovering details of the historical presence of populations that are currently inadequately documented. Researchers can now make connections between past populations and contemporary ones. PXRF makes these discoveries possible with minimal starting material and low-cost experimentation. We intend to follow this pXRF examination of grave soil remains up with most conventional chemical analyses such as ICP-MS and ICP-OMES. These additional analytic techniques are useful in determining the speciation and source of heavy metal deposition of these elements within a sample. These destructive methods were avoided as they require consumption of limited grave soil samples of significant historical interest.

Materials and Methods
Study design. This investigation looks at 325 pelleted soil subsamples analyzed from the decomposed remains of 44 burials to elucidate anthropogenic and environmental conditions of the NYABG population and Lower Manhattan. These 44-cadaver associated burials resulted in the collection of 65 soil samples chosen by opportunistic sampling. The collection is on loan to the W. Montague Cobb Research Laboratory (CRL) at Howard University from the National Park Service. All burial samples containing >20 g of soil were chosen for analysis. Burial samples containing <20 g of soil were not used for pXRF analysis. We collected subsamples of each burial sample for pXRF analysis. The remaining sample will be used for future chemical analysis. The subsamples were particle size sifted using a 250 μm and formed into a cylindrical pellet with a urea binder. The pellet process creates a uniform weight, shape, and size for each sample reducing the chances of variation in element detection. Each pellet was analyzed on a non-invasive, semi-quantitative, portable Bruker Tracer SD-III XRF spectrometer 46 to detect the wavelength of each element in each sample. Bruker Tracer S1PXRF and ARTAX 47 software packages were used to interpret the atomic spectrum identified by the device. The output from these programs were analyzed for statistical significance and interpreted based on reports from the initial excavation, historical context, and pre-existing scientific literature.

Data availability
The New York African Burial Ground soil sample dataset belongs to the W. Montague Cobb Research Laboratory at Howard University. This dataset is not publicly available yet, since it's actively being used for the completion of a doctoral dissertation. However, these data are available from the corresponding author for means of collaboration or by reasonable request.
The United States Department of Agriculture-Natural Resources Conservation Service (USDA-NRCS) control sample data were collected from the National Cooperative Soil Survey support data of New York City. Data are available upon request to corresponding author. Details can be found at: https://www.soilandwater.nyc/ uploads/7/7/6/5/7765286/reconnaissance_soil_survey_report.pdf.