Forensic Implications for the Diagenetic Alteration of the Stable Isotopic Composition of Pig Ribs Following Exposure to Aqueous Conditions.

CHAPTER ONE: INTRODUCTION

The application of stable isotope analysis to the forensic anthropological objective of

human identification serves as an opportunity for law enforcement to pursue new investigative

avenues in cases where all previous leads and forensic techniques have been exhausted without

results. The National Missing and Unidentified Persons System (NAMUS), a database of missing

and unidentified individuals in the United States, currently maintains the profiles of 13,616

unidentified individuals whose remains are stored in medical examiners offices across the

country. Of these profiles, 902 belong to unidentified individuals whose remains were discovered

in the state of Florida. Additionally, 782 of these were discovered many decades ago (between

1957 and 2010) yet remain unidentified. While a NAMUS profile is not created for every

unidentified individual in the United States, these database-specific statistics illustrate the critical

need for new forensic methods that can be used to produce investigative leads in cases of

unidentified individuals, particularly those that have gone decades with no new developments.

In these cases, stable isotope analysis may provide a relatively inexpensive and time

efficient catalyst for progress by providing isotopic data that can be used to narrow the

geographic region of origin of an unidentified individual (Bartelink et al., 2014; Bartelink et al.,

2016; Chesson et al., 2018). While stable isotope analysis is not a primary method of

identification, it is valuable in extreme circumstances, such as cases of skeletal mutilation and/or

severe fragmentation, that render traditional methods inapplicable (Holobinko, 2012; Chesson et

al., 2018), and in cases where previously utilized forensic methods have failed to produce an

identification. Furthermore, stable isotope analysis does not require the use of an antemortem

database for comparison, making it an invaluable tool in circumstances where antemortem data

are scarce or unavailable (Meier-Augenstein, 2018), including investigations of humanitarian

crises (Bartelink et al., 2020).

Although stable isotope analysis has been routinely utilized in bioarchaeological research

for over 40 years, the recent application of this method to the process of human identification in

a medicolegal context necessitates continued research to ensure reliability of isotopic data in

various contexts. Stable isotope analysis is not yet a routinely utilized methodology in forensic

anthropology, however, the proliferation of publications regarding the potential of this

application suggests that this technique will likely become a more frequently utilized method

(Bartelink et al., 2016). Therefore, ensuring that stable isotope values are authentic following the

exposure of bone to a variety of taphonomic contexts, including aqueous environments, is of key

importance for the application of this method to forensic investigations of human remains

(Ehleringer and Matheson, 2010).

Research specific to the reliability of stable isotope values following exposure to and

submersion within aqueous environments is particularly critical for forensic anthropologists

working in distinct geographic areas, such as Florida, where human remains are frequently

exposed to both marine and/or terrestrial water sources. While oxygen isotope analysis has been

routinely utilized in both bioarchaeological and paleontological studies, there have been

questions raised as to the authenticity of δ18O values following exposure to various depositional

environments, including aqueous environments, due to isotopic exchange between the skeletal

material and the water to which it is exposed (Zazzo et al., 2004; Hoke et al., 2018; Price et al.,

2019). Therefore, this research represents an important opportunity to determine the authenticity

of ẟ18O values following submersion within fresh, brackish, and distilled water. The results of

this research can be applied to cases of unidentified human remains recovered from distinct

aqueous environments to ensure the authenticity and reliability of isotopic data prior to their use

in a medicolegal context.

Purpose

The purpose of this research is to evaluate the application of stable isotope analysis to the

process of human identification by forensic anthropologists in circumstances where unidentified

human remains are recovered from an aqueous context in Florida. This research will evaluate the

authenticity of the δ18O isotopic value of pig bone carbonate, used as a proxy for human bone,

that has been exposed to brackish water, freshwater, and distilled water for a period of one, three,

and six months.

At a broad level, this study seeks to answer the following question:

1. Does the δ18O value of bone carbonate differ significantly from control values following

exposure to aqueous environments?

Specific research questions are as follows:

1. Does exposure to an aqueous environment cause a significant shift in the δ18O isotopic

composition of bone carbonate?

a. Does exposure to different aqueous environments (fresh water, brackish water,

and distilled water) cause a significant shift in the δ18O isotopic composition of

bone carbonate?

This research is important because stable isotope analysis can be used to narrow the

potential geographic region of origin of unidentified human skeletal remains, and focus

investigative leads, when primary methods of identification are inapplicable due to extreme

conditions (Holobinko, 2012; Chesson et al., 2018). While a potentially valuable tool for forensic

anthropologists working in a medicolegal context, stable isotope analysis must be rigorously

tested in a variety of circumstances, including various depositional environments, to justify the

admittance of isotopic evidence in a court of law (Ehleringer and Matheson, 2010).

Chapter Summary

This thesis focuses on the application of stable isotope analysis to forensic anthropology

and the authenticity of stable isotopic data following exposure to various aqueous environments.

In this chapter, I introduced the value of stable isotope analysis in medicolegal investigations and

the necessity of continued research to ensure that isotopic data are a reliable source of evidence.

Chapter Two will provide an in-depth overview of background literature related to stable isotope

analysis in both bioarchaeological and forensic contexts and diagenetic processes that may alter

the isotopic composition of bone. The third chapter will discuss the materials and methods used

in this thesis research, including laboratory protocol for carbonate and Fourier Transform

Infrared Spectroscopy (FTIR) analysis. The fourth chapter will detail the authenticity of data and

the results of the statistical analyses used to detect statistically significant differences between

sample groups. Chapter Five will include a discussion and interpretation of the results. The sixth

and final chapter will provide a conclusion, including a summary, study limitations, and future

research directions.

CHAPTER TWO: LITERATURE REVIEW/BACKGROUND

The first half of this chapter focuses on stable isotopes, particularly stable oxygen

isotopes, their circulation through natural systems, and geographically defined isotopic

signatures. Additionally, background will be provided regarding the application of stable isotope

analysis to both bioarchaeology and, more recently, forensic investigations. In particular, the

identification of human remains using stable isotope analysis will be discussed in detail,

including the circumstances in which this method may provide valuable evidence to forensic

investigators. The second half of this chapter focuses on the diagenetic alteration of bone

carbonate, including past research regarding diagenesis in aqueous environments, as diagenetic

processes may impact the reliability of isotopic data. Finally, information regarding the

hydrology of Florida, the location of this study, is provided to establish the characteristics of

aqueous environments in this region.

Introduction to Stable Isotopes

Isotopes are atoms of a single element that have a different number of neutrons in their

nuclei while containing the same number of protons and electrons (Ambrose, 1993; Katzenberg

and Walters-Rist, 2019). In other words, isotopes are elements of differing masses that fill the

same position on the periodic table (Ambrose, 1993; Fry, 2006). The term ‘isotope,’ based on the

Greek words isos, meaning ‘same,’ and topos, meaning ‘place’ was inspired by this definition

(Fry, 2006; Britton, 2017). While radioactive isotopes, such as δ14C, will decay over time, stable

isotopes will not decay or break down to form another element unless the material is altered by

diagenesis or contamination (Ambrose, 1993; Katzenberg and Walters-Rist, 2019), making them

ideal for geochemical studies of past populations. Although over 300 stable isotopes have been

identified, archaeologists and forensic anthropologists most often utilize “traditional isotopes,”

hydrogen, carbon, nitrogen, oxygen, and sulfur, in research studies (Fry, 2006). This study will

focus on the δ18O isotopic composition of domestic pig bone carbonate (bioapatite) which will be

used as a proxy for human bone.

Each element has both a heavy and light isotope variety; light isotopes are usually the

more abundant within natural systems (Fry, 2006). Heavy isotopes contain one extra neutron in

the nucleus which increases their mass when compared to their lighter counterpart (Sharp, 2017).

This difference in mass forms the basis of detectable differences in isotope properties and

behavior, known as fractionation. Fractionation is the separation of heavier and lighter isotopes

that occurs during the transition between phases in natural systems. This separation results in the

variable distribution of isotopes within the biosphere (Fry, 2006; Sharp, 2017). Biosphere pools

serve as stable reservoirs for isotopes as they circulate throughout the biosphere. Oxygen, the

most abundant light element on earth, is pooled largely in the ocean as H2O and in the

atmosphere as O2 (Fry, 2006; Meier-Augenstein, 2018).

In nature, oxygen takes the form of either 16O, 17O, or 18O, with an abundance of 99.76%,

0.04%, and 0.02%, respectively (Sharp, 2017). These isotopes are incorporated into plant

products via biochemical compounds such as water (H2O), Oxygen (O2), and carbon dioxide

(CO2) (Meier-Augenstein, 2018). Oxygen isotope values are determined by fractionation caused

by the hydrological cycle and vary based on geographical characteristics such as latitude,

longitude, elevation/altitude, and distance from the ocean or other large bodies of water (Chesson

et al., 2018; Katzenberg and Waters-Rist, 2019). For example, δ18O values are lower in high

altitude rain than in low altitude rain due to preferential removal of H218O in precipitation of

rising clouds (Pederzani and Britton, 2019). The variation in isotopic values caused by these

climactic and geographical characteristics are largely consistent across the globe and have been

mapped via precipitation studies and the Global Network of Isotopes in Precipitation (GNIP), an

international network for monitoring and reporting oxygen and hydrogen isotopes in

precipitation. These geospatial patterns of known and predicted stable isotope ratios, known as

isoscapes, can then be used to predict δ18O variation across regions (Bowen and Wilkinson,

2002).

These isotopes are incorporated into human tissues via drinking water (δ18Odw), the direct

consumption of edible plants, indirectly via consumption of animal products made from plant-

eating animals, and the inhalation of air (Chesson et al., 2018). While both food and drinking

water influence the stable oxygen isotopic compositions of human tissue, the impact of drinking

water on bone bioapatite much greater than that of food products (Chesson et al., 2018).

Additionally, metabolic, anatomical, and physiological factors of the species or individual,

including climate adaptations (e.g., water retention, heat loss mechanisms, and body size), may

produce oxygen isotopic variation in the fractionation between environmental drinking water and

the isotopic signature of bone carbonate (Lutz et al., 1984; Daux et al., 2008). Due to the

geographically defined signatures of oxygen isotopic values, they are most often utilized in

archaeological research studies of origin and paleomigration (Makarewicz and Sealy, 2015;

Katzenberg and Waters-Rist, 2019) and forensic investigations related to trafficking of humans

and/or materials across the globe (Chesson et al., 2018; Meier-Augenstein, 2018).

The isotopic composition of a sample is expressed as a δ (delta) value, representing the

relative difference in the ratio of the heavier isotope over the lighter between the sample and an

international reference standard (Sharp, 2017; Hoefs, 2018). The δ value, reported in per mil

(parts per thousand), is based on ion signals measured by an isotope ratio mass spectrometer

(IRMS) (Sharp, 2017; Hoefs, 2018). Low δ values indicate depletion in heavy isotopes while

higher δ values indicate enrichment by heavy isotopes when compared to the selected standard

(Fry, 2006; Sharp, 2017). In other words, a δ 18O value of +18.5‰ represents an 18O/16O ratio

that is 18.5 per mil (1.85 percent) higher than the selected standard (Sharp, 2017). The delta

value (δ) of a sample is calculated via the following equation:

δ in ‰ = (R(sample) / R(standard) – 1) x 100

Skeletal Biology

Although any human tissue can be utilized in stable isotopic studies, bone collagen, bone

carbonate (bioapatite), tooth dentin, tooth enamel, hair, and nails are the most frequently

analyzed (Table 1) (Chesson et al., 2018). Human bone is comprised of approximately 30%

organic and 70% inorganic components (Ambrose, 1993; Katzenberg and Waters-Rist, 2019).

Furthermore, approximately 85–95% of the organic component of bone is collagen while the

remaining 10-15% consists of other proteins and lipids (Katzenberg and Waters-Rist, 2019). The

inorganic (mineral) component of bone is bioapatite, a crystalline calcium phosphate mineral that

is similar to hydroxyapatite (Ca10(PO4)6OH2) (Metcalfe et al., 2009).

Due to variation in formation and remodeling processes of tissues, different tissues from

the same individual will exhibit different isotopic signatures (Pederzani and Britton, 2019).

Similarly, the mineral and organic components of skeletal tissues will differ in isotopic

composition due to fractionation caused by physiological and metabolic processes, such as

oxygen fluxes and body temperature (Kohn and Cerling, 2002; Podlesak et al., 2008). The three

most frequently analyzed components of human skeletal tissue (e.g., bone and teeth) are

bioapatite carbonate, bioapatite phosphate, and collagen (Pederzani and Britton, 2019). This

study will focus on the δ18O composition of bone carbonate via the analysis of the isotopic

composition of pig ribs, used as a proxy for humans, due to similarities in physiological

characteristics (Connor et al., 2018).

While this study will focus solely on the carbonate component of bone, it is important to

note that, when available, multiple tissues and/or tissue components may be isotopically

analyzed and used in conjunction with one another to form more robust interpretations regarding

the life history of an unidentified human individual. The differential formation and remodeling

processes of various tissues (i.e., hair, nails, bone, teeth) allow researchers to examine multiple

time periods throughout the life of an individual. Tooth enamel does not continuously remodel

once formed, meaning that the isotopic composition of enamel remains consistent throughout the

entire life course and after death. Therefore, the isotopic composition of tooth enamel provides

data regarding the childhood/adolescence of an individual (Humphrey, 2016). Conversely, hair

and nails grow incrementally, at a known rate, and do not remodel once fully formed (Fraser et

al., 2006). Similarly, bone continues to remodel throughout the entire lifetime of an individual, at

a rate of approximately 10-15% per year depending on the bone (Fahy et al., 2017), and age and

health of the individual, providing an average isotopic composition of at least the last several

years before death. Depending on the length of the sampled hair and nails and the tooth/bone

selected for analysis, the isotopic composition can model chronological changes in diet and

geographic location for a period of months to years before death.

Table 1. Summary of human tissues commonly utilized in stable isotopic analysis. Derived

from: (Fraser et al., 2006; Fahy et al., 2017; Mancuso and Ehleringer, 2018)

Tissue

Component

Interval

Stable Isotopes

Information

Nail

Keratin

Months before

death

(depending on

length)

13C, 15N,

34S, 2H, 18O

Geographic

history, Diet

Hair

Keratin

Months before

death

(depending on

length)

13C, 15N,

34S, 2H, 18O

Geographic

history, Diet

Bone

Hydroxyapatite

(bioapatite)

Years before

death

13C, 18O,

87Sr/86Sr

Geographic

history, Diet

Collagen

13C, 15N, 34S

Diet

Teeth

Enamel

(bioapatite)

Infancy to

adolescence

(depending on

tooth type)

13C, 18O,

87Sr/86Sr

Childhood

residence,

Childhood diet

Dentin

(Collagen)

13C, 15N

Childhood diet

Stable Isotope Analysis in Bioarchaeology

In bioarchaeology, osteological data are used to reconstruct aspects of identity and lived

experience based on the notion that characteristics of the human skeleton are representative of

the life of an individual (Larsen, 1997). This notion extends to stable isotope analysis, as isotopic

composition in human tissues directly reflects the food and drink consumed by an individual

during their life which can be used as a proxy to reconstruct the dietary patterns and geographic

origins of that individual (Bogaard and Outram, 2013). Stable isotope analysis was first applied

to archaeological research in the late 1970s when researchers measured the ratio of carbon

isotopes in human bone to estimate the dietary patterns of prehistoric individuals from North

America (Vogel and van der Merwe, 1977).

The continued development of stable isotopic methodology over the last forty years,

including the advancement of continuous flow mass spectrometry in the late 1990s, has

expanded the application of these analyses far beyond their previous scope (Makarewicz and

Sealy, 2015). Bioarchaeologists now routinely apply stable isotope analysis, including carbon,

nitrogen, oxygen, and strontium isotope analysis, to a variety of research studies including

studies of diet (Newsome et al., 2004), migration (Dupras and Schwarcz, 2001), zooarchaeology

(Zangrando et al, 2014), paleopathology (Reitsema, 2013; Scorrano, 2018), and breastfeeding

and weaning practices (Dupras and Tocheri, 2007).

One important impact of stable isotope analysis on archaeological research is its ability to

produce data related to individual life histories (Tykot, 2004; Britton, 2017; Singh and Kaur,

2017). Stable isotope analysis allows bioarchaeologists to examine the life and lived experiences

of an individual by interpreting and identifying dietary and migratory patterns derived from

isotopic compositions of various human skeletal tissues (Singh and Kaur, 2017). The application

of stable isotope analysis to bioarcheological research provides data that also can elucidate

certain individual characteristics and experiences, such as social status (Colleter et al., 2017),

gender (Sommerville et al., 2015), age cohorts (Pearson et al., 2015), infectious disease (Roberts

et al., 2013), and chronic illnesses (Scorrano et al., 2014).

Oxygen (δ18O) isotopes are most often utilized in archaeological studies to determine the

geographic origin of human and animal skeletal remains, based on predictable variation in

oxygen isotopic signatures based on geographic and climatological factors (Bowen and

Wilkinson, 2002). While oxygen isotope analysis is routinely utilized in geochemical and

paleodietary contexts, the first preliminary studies of oxygen isotope analysis as a method of

determining geographic origin in an archaeological context were not published until the 1990s

(Dupras and Schwarcz, 2001; White et al., 1998). These early, influential publications

demonstrated the value of oxygen isotope analysis to archaeological studies of mobility. Since

the δ18O signature of bone and teeth are representative of local drinking water (i.e., precipitation,

groundwater), the isotopic data obtained from the analysis of bones and teeth can be compared to

geospatial maps of oxygen isotopic signatures (comprised of baseline data) to better understand

the residency and movement of individuals in an archaeological population, including identifying

first generation immigrants within an assemblage of archaeological human remains (Lightfoot

and O’Connell, 2016). Furthermore, as teeth and bones represent different periods within the life

of an adult individual (i.e., childhood, adulthood) it is often possible to determine if an individual

migrated throughout their lifetime based on observed differences in the isotopic signatures of

sampled skeletal tissues (Lightfoot and O’Connell, 2016). Based on these methodological

principals, bioarchaeological research using oxygen isotopes to differentiate between locals and

foreigners has been frequently published (Turner et al., 2009; Theden-Ringl, 2011; Toyne et al.,

2014; Sengeløv, et al., 2020). Therefore, the application of oxygen isotope analysis to

bioarcheological research involving human skeletal remains serves as a predecessor to the use of

isotopic data to narrow geographic origin and identify human remains in a medicolegal context.

Stable Isotopes in Forensic Investigations

Forensic science is the application of scientific techniques to the investigation of crimes

and the prosecution of criminals via the generation of facts that may be used as evidence in a

court of law (Netzel et al., 2014). Most often forensic techniques used in criminal investigations

share the common objective of matching a sample, or samples, associated with the defendant or

victim to a sample collected from the crime scene (Berger, 2003; Ehleringer and Matheson,

2010). However, prior to admittance in a court of law, forensic techniques must meet rigorous

scientific and reliability standards (see below for details regarding evidence admissibility).

Stable isotope analysis is used in a variety of forensic contexts including investigations of

food production, illicit drug trafficking, firearms and explosives, wildlife forensics, and the

identification of unidentified human remains (Daeid et al., 2009; Ehleringer et al., 2015; Chesson

et al., 2018). Within these contexts, stable isotope analysis is often used to match or differentiate

between samples, particularly those that would otherwise be chemically identical (Gentile et al.,

2011). This utilization allows for three key applications of stable isotope analysis in a forensic

context (Ehleringer et al., 2007; Ehleringer and Matheson, 2010). First, this method can be used

to determine if two samples have the same origin. This application may be enacted when

comparing samples from the crime scene to that of the defendant or victim. Second, this method

can be used to compile databases of expected isotope ratios of specimens of a specific type and

origin to allow for future comparisons. Third, this method can be used to determine the most

likely geographic source location of an unknown specimen via a comparison to geospatial maps,

or isoscapes, of the distribution of predicted and known stable isotope ratios.

In investigations of food production, stable isotope analysis can be used to determine if

manufacturers are complying with federal regulations of manufacturing protocols and/or

providing authentic information regarding source materials (Ehleringer et al., 2007; Chesson et

al., 2014; Chesson et al., 2018). For example, Zhou et al. (2018) analyzed commercial honey

from multiple countries and determined that many of the samples were fraudulently adulterated

with inauthentic ingredients and/or labeled with an inaccurate geographic origin. Stable isotope

analysis can be used in investigations of illicit drugs, including marijuana, cocaine, and heroin, to

determine the geographic origin of seized samples and discriminate between distinct batches

and/or manufacturers (Ehleringer et al., 2007; Daeid et al., 2009; Chesson, et al., 2014).

Similarly, stable isotope analysis has been used to differentiate between types, batches, and

manufacturers of explosives, particularly those used by terrorist organizations (Ehleringer et al.,

2007; Daeid et al., 2009). In the context of wildlife forensics, stable isotope analysis can be used

to determine the geographic origin and dietary patterns of animals, making it possible to track

their movement across multiple geographic areas, which is useful in investigations of

unregulated animal trade, animal trafficking, and differentiating between captive-bred and wild

animals (Bowen et al., 2005; Chesson et al., 2018; Meier-Augenstein, 2019). In the past, stable

isotope analysis has played a key role in the forensic investigation of several high-profile cases.

For example, this method was used to identify the culture medium and geographic origin of a

microbial agent (Bacillus anthracisis) used as a bio-weapon against news agencies and multiple

United States senators in the 2001 anthrax attacks, known as “Amerithrax” (Kreuzer-Martin and

Jarman, 2007; Ehleringer and Matheson, 2010).

The applications of stable isotope analysis to the forensic contexts mentioned previously

have been researched, tested for reliability, and published in peer–reviewed journals. This has

successfully established the utility of isotope analysis as a valid and reliable forensic evidentiary

resource in specific circumstances. However, new applications, preparation methods, and

instrumentation must continue to be tested to ensure the continued reliability of new methods,

such as stable isotope analysis, as a source of evidence in criminal investigations (Ehleringer and

Matheson, 2010).

To be admitted to a court of law, evidence must meet Daubert standards, laid out by the

Supreme Court case Daubert vs. Merrill Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993)

(Rehnquist, W., dissenting), stating that evidence must be approved by a judge based on four

main criteria: 1) whether the method producing the evidence has been adequately tested, 2)

whether the method producing the evidence has been peer reviewed and published, 3) whether

the method has a potential and/or known rate of error, and 4) whether the method is generally

accepted within the relevant field/discipline (Christensen et al., 2009; Bethard and DiGangi,

2019). Additionally, the judge may evaluate the evidence based on the relevance of the data and

associated expert testimony to the circumstances of a specific case (Ehleringer and Matheson,

2010; Holobinko, 2012; Netzel et al., 2014). The Daubert standards create a high degree of

specificity for the admissibility of evidence to a court of law, meaning that forensic

anthropologists must rigorously validate methods, particularly new and/or novel methods

(Christensen, 2004), before they may be applied to the facts of a particular case in a court of law

(Christensen and Crowder, 2009; Bethard and DiGangi, 2019).

While stable isotopic research involving skeletal remains often involves the assumption

that the isotopic composition of the sampled material is authentic (i.e., not altered by taphonomic

or diagenetic alteration), there are various circumstances in which the sampled material may no

longer be authentic due to diagenetic processes. Diagenesis, the chemical and/or structural

alteration of organic and/or non-organic materials, may alter the isotopic composition of skeletal

materials when exposed to certain environments. Therefore, a judge and/or prosecutor may be

concerned about the reliability of stable isotope values following a period of exposure to a

depositional environment resulting in diagenesis. Continued research is necessary to determine

the stability of the isotopic value of a sample when exposed to certain circumstances, such as

aqueous environments, for a specific period (Ehleringer and Matheson, 2010). Therefore, the

application of stable isotope analysis to specific cases and circumstances may require additional

research and publication in a peer-reviewed journal before being admitted as evidence in

criminal proceedings (Ehleringer and Matheson, 2010).

Application to Identification of Human Remains

While there are multiple well-established methods of identifying human remains,

including DNA analysis, forensic anthropological examination (e.g., age, sex, ancestry, stature)

and comparative radiography, these methods may not be applicable to every case due to

circumstances such as skeletal mutilation, severe fragmentation of remains, lack of comparative

DNA, medical and/or dental records, and relocation of remains far from their geographic origin

(Holobinko, 2012; Chesson et al., 2018; Meier-Augenstein, 2018). In these cases, stable isotope

analysis can be particularly effective as, unlike many of the aforementioned identification

methods (e.g., DNA analysis, comparative radiography), it does not rely on subject specific

databases or direct comparisons to antemortem data sources (Meier-Augenstein, 2018).

Additionally, this method is helpful in cases of severe decomposition and/or skeletonization in

which fingerprinting is no longer an option (Rauch et al., 2007) and may provide valuable data in

cases of commingled skeletal remains (Trammell et al., 2018).

However, it is important to note that stable isotope analysis is not a primary method of

identification and will not result in an individual identification unless used in conjunction with

other forensic and/or investigative methods (Daeid et al., 2009; Holobinko, 2012). Similarly,

geospatial analyses using isotopic data may not provide a definitive location of origin but can

provide new investigative leads by narrowing search parameters and excluding certain

geographic regions (Bartelink et al., 2014; Chesson et al., 2018). Instead, the goals of stable

isotope analysis in a forensic anthropological context are twofold: 1) to determine if the

unidentified individual originated in the region from which the remains were recovered; and 2) to

determine potential regions through which the unidentified individual may have traveled at

various times (Chesson et al., 2018).

Additionally, stable isotope analysis provides information that can be used in the

identification of pre-teens, children, infants, and neonates, for whom distinguishing skeletal

morphological features, including secondary sex characteristics, have yet to develop (Meier-

Augenstein, 2018). Stable isotope analysis conducted on the tissues of infants and neonates may

also provide information related to the dietary intake and geographic location of the mother

during pregnancy (Lehn et al., 2015). This application was utilized in the case of a partially

skeletonized infant found outside a shopping mall in Germany. The stable isotopic composition

of hair and bone samples suggested that the mother may have moved from North America to

Germany part way through her pregnancy. In fact, once located by investigators, the mother was

discovered to have moved from New Mexico to Germany during the third trimester of pregnancy

(Lehn et al., 2015).

Although the potential application of stable isotope analysis to the identification of

human skeletal remains in a medicolegal context was first recognized in the late 1980s

(Katzenberg and Krouse, 1989), it was not until the early 2000s that stable isotope analysis first

led to the confirmed identity of the victim and the conviction of a killer (Rauch et al., 2007;

Meier-Augenstein and Fraser, 2008). Meier Augenstein and Fraser (2008) state that the body of

the murder victim was “mutilated” prior to deposition, and many skeletal elements, including the

skull, were never recovered. Rauch et al. (2007) describe the body of the murder victim as being

badly decomposed, and with dental characteristics suggesting that the individual was likely not

originally from the region in which his body was discovered. These cases represent key

examples of circumstances in which stable isotope analysis was a critical aspect of the

investigation, providing necessary forensic evidence to investigators.

Stable isotopic analysis has also been applied to several cases in which the isotopic data

successfully provided new investigative leads but did not result in a positive identification. This

was the outcome in one early application of stable isotope analysis that occurred when the

dismembered torso of a 5-year-old boy was discovered in the Thames River in London, England

(O’Reilly, 2007). While the identity of the child was never discovered, stable isotope analysis

was able to determine a non-local origin and narrow the geographic source range to a 50 by 100-

mile area in Nigeria (O’Reilly, 2007).

In subsequent years, this method has continued to be utilized in cases of unidentified

human remains including an embalmed head (Miller, 2015), the remains of a woman found in

Utah (Kennedy et al., 2011), and the remains of a man whose identification led to the discovery

of a Vietnamese drug smuggling operation in the United Kingdom (Meier-Augenstein, 2018).

Most recently, in June of 2020, δ18O isotopic analysis was used to narrow the geographic region

of origin of a “John Doe,” whose dismembered skeletal remains were found in Ohio in 2016, to a

small area in Kentucky. This determination, along with traditional investigative techniques, led

detectives to successfully identify the remains as those of Ryan Zimmerman, who had recently

moved to Columbus, Ohio from Corbin, Kentucky (Shriver, 2020).

In the United States, stable isotope analysis is most frequently utilized by the Defense

POW/MIA Accounting Agency (DPAA) where it has proven effective in distinguishing

deceased United States military personnel from local individuals in areas of past conflicts

(Bartelink et al., 2014), including the Vietnam War (Regan, 2006; Someda et al., 2016) and

World War II (Font et al., 2015). In this context, stable isotope analysis is often utilized solely to

determine if the individual was raised within the United States, as opposed to the local region of

conflict, rather than to determine the exact provenience of the individual within the United

States. This distinction is necessary and valuable as, in certain circumstances, a positive

identification may be attained if the stable isotopic data produced from the skeletal remains of

the unidentified individual are consistent with those of the United States. For example, stable

isotope analysis was used to positively identify a deceased Vietnam War pilot from a single

fragment of tooth enamel recovered from the crash site after the stable isotopic data concluded

that the individual was raised in the United States (Holland et al., 2012).

Similarly, this method has been used to predict the region of origin of unidentified border

crossers at the Mexico-US border (Juarez, 2008; Juarez et al., 2020; Ammer et al., 2020). In

these cases, stable isotope analysis is particularly valuable as unidentified border crossers are

often completely skeletonized upon recovery, are not frequently reported missing in the United

States, and do not have readily available antemortem reference data (Bartelink et al., 2020).

Additionally, unidentified Border Crossers are a heterogenous group comprised of individuals

traveling from a variety of regions in Central and South America, making the use of stable

isotope analysis to distinguish between potential geographic regions of origin a critical addition

to forensic anthropological assessments (Kramer et al., 2020; Bartelink et al., 2020).

Furthermore, in recent years, studies have been conducted to assess the efficacy of stable

isotope analysis in humanitarian identification efforts following large scale events such as

periods of sociopolitical violence, mass disasters, forced disappearances, and genocide (Zimmer-

Dauphinee et al., 2020; Castellanos Gutiérrez et al., 2020). In this application, using stable

isotope analysis to narrow the region of origin of decedents allows humanitarian investigators to

focus other investigative methods, such as the collection of comparative DNA samples,

interviewing family/community members, and using local news outlets to display case specific

information (e.g., photos of personal effects), to specific geographic locations (Zimmer-

Dauphine et al., 2020; Palma et al., 2020).

However, while some researchers are confident that the integration of stable isotope

analysis, specifically the analysis of oxygen and strontium, to the “forensic science toolkit” used

by forensic anthropologists in humanitarian efforts will support identification efforts (Palma et

al., 2020), others conclude that lack of baseline data in certain countries (i.e., Peru, Columbia,

and Mexico) limit the use of isoscapes to effectively distinguish between specific geographic

locations (Eck et al., 2019; Castellanos Gutiérrez et al., 2020; Zimmer-Dauphinee et al., 2020).

While the application of stable isotope analysis to human identification in cases of large-scale

humanitarian efforts shows promise, more research is necessary to acquire baseline data and

refine isoscapes (Zimmer-Dauphinee et al., 2020). While it is evident that the application of

stable isotope analysis to medicolegal investigations is slowly increasing in frequency, continued

research is necessary to ensure reliability of results and build trust between forensic scientists,

investigators, prosecutors, and the general public.

Carbonate Reliability

Diagenesis

Diagenesis refers to any chemical and/or structural transformation that occurs following

exposure of an organic material to its depositional environment (Ambrose, 1993; Hedges, 2002),

though the inorganic portions of complex materials, such as bone, can also affected by this

process (Wilson and Pollard, 2002). The chemical and structural composition of bone can be

altered by a multitude of natural processes including weathering, microbial activity, uptake of

external minerals, recrystallization, thermal alteration, and isotopic exchange (Pate and Hutton,

1988; Nielsen-Marsh and Hedges, 2000; Trueman and Martill, 2002; Hedges, 2002; Munro et al.,

2007). Previous research suggests that diagenetic processes are highly influenced by the

environmental conditions, including climatological conditions, in which the skeletal remains are

deposited (Hedges and Millard, 1995; Maurer et al., 2014; Sasso et al., 2014).

Research regarding the impact of diagenetic processes on the isotopic values of human

tissue is critical, as alteration of isotopic composition may limit the ability of researchers to

produce accurate interpretations based on resulting data (Wilson and Pollard, 2002; Munro et al.,

2007). Therefore, researchers must demonstrate the reliability of these values following exposure

to various depositional environments and anthropogenic post-mortem processes, such as burning

and/or cremation (Harbeck et al., 2011; Snoeck et al., 2016; Mamede et al., 2017), boiling

(DeNiro et al., 1985; Munro et al., 2007), and freezing (Gordon et al., 2018). These studies

aimed to determine the maximum extreme (i.e., temperature, moisture, etc.) to which a tissue can

be exposed before resulting isotopic values are no longer an authentic data source. Generally,

studies of diagenesis focus on understanding the relationship between various depositional

environments and the observed isotopic alteration (Nelson et al., 1986; Hedges et al., 1995), as

well as the evaluation of preparation methods to assess the severity of diagenesis (Zazzo et al.,

2004; Young Shin and Hedges, 2012).

Although it is widely believed that the risk of diagenetic alteration in forensic cases is

minimal due to the relatively short depositional periods of less than 70 years (Bartelink et al.,

2016; Meier-Augenstein, 2018), skeletal remains may be subject to diagenesis prior to recovery

and analysis. Certain depositional environments, such as rivers, lakes, water-logged graves, and

the ocean may result in the alteration of bone microstructure (Bell et al., 1995; Bell and Elkerton,

2008), which may compromise isotopic compositions (Meier-Augenstein, 2018). The potential

diagenetic alteration of skeletal remains may negatively impact forensic casework as unidentified

human remains are frequently recovered from a variety of aqueous contexts following purposeful

deposition and/or accidental death. Therefore, the risk of diagenetic alteration, even in cases of

relatively short depositional periods, must be addressed and adequately investigated to more

fully understand the chemical processes that may impact the integrity of stable isotope values in

forensic cases.

While literature has been published regarding the diagenesis of ẟ18O isotopes exposed to

various environments, including aqueous environments, much of this literature focuses on long

term exposure rather than short temporal intervals. Specifically, much of this literature is written

regarding the diagenesis of fossilized organic materials used in paleontological reconstruction,

often following aqueous deposition (Wang and Cerling, 1994; Trueman et al., 2003; Zazzo et al.,

2004; Tütken et al., 2008). Decades of debate regarding the reliability of the δ18O isotopic

composition of bone and enamel carbonate following long term exposure to depositional

environments has not yet reached a definitive conclusion.

While some literature concludes that both bone and enamel apatite are reliable sources of

isotopic data used for palaeoclimatological and paleodietary reconstruction (Koch et al., 1997;

Clementz, 2012; Zazzo et al., 2014), various studies have conversely concluded that both bone

and enamel apatite may be significantly altered by diagenetic processes in certain environments

(Schoeninger et al., 1970, Nelson et al., 1986; Tütken et al., 2008; Price et al., 2019). While it is

largely accepted that enamel is the least porous, most durable mineralized tissue in the human

body, making it less susceptible to diagenetic alteration than bone (Kendall et al., 2018), it may

still be significantly diagenetically altered in certain environments, including aqueous

environments (Price et al., 2019). When discussing diagenetic alteration caused by exposure to

and/or submersion within aqueous environments, the porosity of the skeletal tissue is critically

important as the structure of the tissue itself (porosity, collagen preservation, etc.), along with the

hydrology of the deposition environment, control the movement of water (and exogenous

materials within the water) through and within the skeletal tissue (Hedges and Millard, 1995),

which may facilitate carbonate substitutions.

One short term study tested the resistance of δ18O to isotopic alteration following the

submersion of modern animal bone and enamel within sterile and microbial water in a laboratory

setting (Zazzo et al., 2004). In this study, powdered bone and enamel were placed in nylon bags

and submerged in an aqueous microbial environment, consisting of water and microbe induced

soil, and a sterile aqueous environment, consisting of water and mercuric chloride to prohibit

microbial growth. This methodology was based on the expectation that δ18O from aqueous fluid

may be incorporated into the apatite structure of bone and enamel via isotopic exchange

following a period of one year, resulting in altered isotopic values. The authors observed

significant changes to the δ18O values of bone carbonate, especially in samples exposed to

microbial conditions, while enamel was noted to be less significantly altered in both sterile and

microbial contexts. Furthermore, the authors state that subsequent acetic acid treatment was

unable to fully eliminate the isotopic shift in bone (Zazzo et al., 2004). The study by Zazzo et al.

(2004) indicates that δ18O isotopic values of bone may be rapidly altered when exposed to

aqueous environments, particularly environments with high microbial activity. While δ18O has

largely been accepted as reliable in both bioarchaeological and paleontological studies, Zazzo et

al. (2004) represents the only laboratory study designed to test the early diagenetic processes

caused by short term submersion within aqueous environments. However, it is important to note

that the samples utilized in the study by Zazzo et al. (2004) were powdered, greatly increasing

the surface area and exposure of the sample to the aqueous environment, which is unlikely to

occur in either a forensic or bioarchaeological context.

Furthermore, recent archaeological studies have noted the diagenetic impact of

groundwater on bone and enamel carbonate in various environmental contexts (Maurer et al.,

2014; Price et al., 2019), including within marine water, which may differentially alter the

structural and isotopic composition of bone carbonate (Tütken et al., 2008). Moreover, previous

research suggests that the consistency of the interaction between bone and water produce

differential bone preservation and diagenetic processes, with consistent (static) exposure to water

contributing to better preservation while fluctuating periods of wet and dry (cyclical) exposure

may contribute to structural degradation and chemical alteration (Hedges and Millard, 1995;

Nielsen-Marsh and Hedges, 2000; Kendall et al., 2018). Therefore, continued research is

necessary to better understand the diagenetic alteration of bone carbonate following deposition in

different aqueous environments for various short-term intervals, as would be relevant to forensic

investigations.

Methodological Considerations

Although both the carbonate and phosphate fractions of bone and tooth enamel can

produce oxygen isotopic data, phosphate is generally better preserved and less likely to be

altered by diagenetic processes. The C–O bonds in the carbonate are weaker and more

susceptible to exchange with exogenous materials than the stronger P–O bonds in the phosphate

fraction of bone and tooth enamel (Demény, et al. 2019). Although the resilience of phosphate to

diagenetic alteration makes it the preferred tissue component for oxygen isotopic analysis, the

methodological requirements of phosphate analysis are much more extensive and expensive than

that of carbonate and necessitate the use of dangerous chemicals (Crowson et al., 1991; Stuart-

Williams and Schwarcz, 1995). Therefore, although carbonate may be more likely to be

diagenetically altered, it is the most frequently utilized skeletal tissue component for oxygen

isotopic analysis because it is more readily processed in laboratories and more affordable

(Demény et al., 2019). Due to the accessibility and frequent use of carbonate analytical methods,

it is critical to acknowledge the potential diagenetic alteration of carbonate samples in certain

environments and continue researching the diagenetic processes that may impact sample

authenticity. Furthermore, both phosphate and carbonate samples should be evaluated for

diagenetic alteration prior to use in bioarchaeological research or forensic investigations.

The most frequently utilized method for the detection and evaluation of diagenetic

alteration of skeletal tissues is Fourier Transform Infrared Spectroscopy (FTIR) analysis. FTIR

measures the absorption of infrared radiation at various wavelengths to evaluate the composition

of skeletal tissues (Yoder and Bartelink, 2010; Beasley et al., 2014). This analysis produces a

spectra of raw wavelength data that can be used to calculate the Crystallinity Index (CI) and

Carbonate Content Ratio (C/P) of each sample. These quantitative measures may indicate

recrystallization and/or mineral substitution when results fall outside of the acceptable range

(Yoder and Bartelink, 2010; Beasley et al., 2014). A detailed description of this methodology

will be provided in the next chapter after an introduction to the research context.

Hydrology of Florida

Water, which covers over 71% of the earth’s surface, is a highly variable substance that

can be broadly separated into three categories: marine (salt) water, fresh water, and brackish

water. Fresh and brackish water are differentiated based on their salinity, a measure of dissolved

solids (often salt) in water (Wetzel, 2001; Tundisi and Tundisi, 2011). Fresh water, found in

inland water sources such as lakes, rivers, and streams, has an average salinity of 0.5 parts per

thousand (ptt) while marine water has an average salinity of 33.0 ppt (Tundisi and Tundisi,

2011). Brackish water, often found in estuaries, has a salinity between that of marine and

freshwater (Tundisi and Tundisi, 2011; Rich and Maier, 2015), ranging from 0.5–33.0 ppt

depending on the season and the volume of fresh and marine water present in a specific body of

water, which generally increases as the river or estuary moves away from the fresh water source

in the direction of the ocean (Ohrel and Register, 2006). Due to the variability of salinity in

brackish water, there are several defined brackish water classifications based on salinity level

including oligohaline (0.5–5.0 ppt), mesohaline (5.0–18.0 ppt), and polyhaline (18.0–30.0 ppt)

(Ohrel and Register, 2006).

Freshwater wetlands, including ponds, lakes, marches, scrub-shrub wetlands, and

swamps, account for approximately 29% of the landscape of Florida (Haag and Lee, 2010). In

Central Florida, these wetland features are often small, isolated, and distributed across a large

area, separated by a variety of environmental features and human-made structures (Haag and

Lee, 2010). Freshwater wetlands are often acidic with low dissolved oxygen levels due to the

accumulation of decomposing of leaves and organic materials in the water. Additionally, these

wetlands are subject to biogeochemical cycles in which elements (e.g., carbon, nitrogen, etc.)

move between the atmosphere, the water, soil, and sediment. Furthermore, freshwater wetlands

usually maintain at least a few inches to several feet of standing water throughout the year,

though this varies between wetlands based on environment and climate (Haag and Lee, 2010).

Brackish water, present in both surface features and sub-surface aquifers, is characterized

by high concentrations of dissolved solids, including metals, minerals, and salt, in excess of

1,000 to 10,000 mg/L (Stanton et al., 2017; Anning et al., 2018). Solids often dissolved within

brackish water include calcium, magnesium, sodium, chloride, sulfate, and bicarbonate (Anning

et al. 2018). Brackish water is often located in transitional areas between terrestrial fresh water

sources and marine water, such as within estuaries, where these water sources meet and combine

resulting in a unique aqueous environment (Rich and Maier, 2015). Across the United States, the

chemical composition of brackish water varies widely based on geological, hydrological, and

geochemical processes. The United States Geological Survey describes brackish water in Florida

as belonging to “group 3,” characterized as a sodium-chloride (salt) dominant water type with a

mean concentration of dissolved solids measuring approximately 8,440 mg/L (Stanton et al.,

2017).

Previous studies have observed both macroscopic and microscopic structural changes to

bone recovered from high salinity marine water (Bell and Elkerton, 2007; Pokines and Higgs,

2015; Eriksen et al., 2020), suggesting that salinity may impact bone structure and preservation.

However, there is currently a gap in the literature regarding the potential taphonomic and

diagenetic alteration of bone in low salinity (i.e., fresh and oligohaline brackish) water. Fresh and

brackish water conditions may differentially impact diagenetic processes affecting bone due to

their distinct compositions. As each of these water types are easily accessible in Florida, this

study represents a viable opportunity to evaluate the impact of distinct submerged conditions.

Chapter Summary

Here, the background theory and applications of stable isotope analysis was established,

including the relatively new application of this method to medicolegal contexts. Additionally,

this chapter provided a review of the diagenetic processes that may alter the oxygen isotopic

signature of bone carbonate and potentially impact the reliability of resulting isotopic data. This

discussion provides context not only for the importance of this research, but for the methodology

detailed in the following chapter.

CHAPTER THREE: METHODS AND MATERIALS

The purpose of this chapter is to detail the methodology followed throughout this

research as well as the materials used to enact this methodology. In particular, this chapter will

include discussions of the collection of water, the collection and preparation of pig ribs,

laboratory protocols for the preparation and processing of bone carbonate, protocols for Fourier

Transform Infrared Spectroscopy (FTIR) analysis, and the statistical analysis of resulting data.

Water Collection

As discussed in the previous chapter, the composition of water can vary significantly

including the level of salinity. Therefore, multiple water compositions (distilled, fresh, and

brackish water) were selected for use in this study to determine if the unique characteristics of

water would differentially impact the isotopic composition of bone samples following exposure.

Fresh water and brackish water were chosen for this study due to their accessibility across the

entirety of Central Florida. Distilled water was selected because it is not saline.

Distilled water was collected from the Biological Sciences Building (Room 430) on the

University of Central Florida campus. The fresh water was collected from Lake Claire on the

University of Central Florida Campus, Orlando, Florida (Latitude 28.608552, Longitude —

81.202807) and the brackish water was collected from Tomoka River Basin in Tomoka State

Park, Florida (Latitude 29.3555, Longitude —81.0904), respectively (Fig. 1). Four-liter

Nalgene® containers were used to collect additional fresh water and brackish water to refill

beakers occasionally throughout the duration of the experiment to ensure conditions remained

submerged in water following evaporation.

Figure 1. Satellite images of water collection sites. (A) Satellite image of Lake Claire on the

University of Central Florida Campus. (B) Satellite image of the Tomoka River Basin in

Tomoka State Park.

The fresh water and brackish water were collected from the same depth, approximately

one meter, to ensure consistency across samples. Prior to collection, a photograph of each

collection site was taken (Fig. 2). Additionally, the temperature of the water was recorded using

a 305mm Fisherbrand® lab thermometer, and the pH of each water sample, including the

distilled water, was recorded using pH test strips. All climatological data (i.e., temperature,

humidity, etc.), provided by the National Weather Service, was recorded for each collection site

(Table 2). Five milliliter Nalgene® sample vials were used to collect three samples for direct

isotopic analysis from the fresh water and brackish water collection sites in June, October, and

January to construct a baseline for natural occurring isotopic variation at these locations. These

containers were sealed with parafilm wax and refrigerated immediately following collection.

While monthly sampling of water from each collection site (along with sampling of

precipitation) would be best practice for the establishment of baseline data, seasonal sampling

was deemed appropriate for the scope of a preliminary study. Additional isotopic data from the

Global Network of Isotopes in Precipitation (GNIP) and the Global Network of Isotopes in

Rivers (GNIR) (accessed via https://www.iaea.org/services/networks/gnip) will be used in

conjunction with the water samples collected directly from Tomoka State Park and Lake Claire

to create a more robust comparative baseline for oxygen isotope variation in Central Florida.

Furthermore, the salinity of each water type was measured using a YSI Pro DSS

Salinometer®. The distilled water was measured at 0.00 ppt, the fresh water from Lake Claire

was measured at 0.08 ppt, and the brackish water was measured at 1.43 ppt. Based on this

salinity level, the brackish water utilized in this study is classified as oligohaline water (Ohrel

and Register, 2006). Moreover, while it is recognized that microbial activity may impact bone

preservation, the level of microbial activity within each water type was not directly measured.

However, with the exception of the distilled water, the fresh and brackish water utilized in this

study were not purified following collection, so it is expected that microbes were present.

Additionally, the beakers containing the water and bone samples were left uncovered in a fume

hood, which may have allowed for the incidental introduction of microbes throughout the

duration of this study.

Figure 2. Photographs of water collection sites. (A) Fresh water collection site at Lake

Claire on the University of Central Florida Campus. (B) Brackish water collection site at

the Tomoka River Basin in Tomoka State Park.

Table 2. Climatological data for each collection site on the day of water collection.

Collection

Month

June

October

January

Collection

Site

Tomoka

River

Lake

Claire

Tomoka

River

Lake

Claire

Tomoka

River

Lake

Claire

Weather

Partly

cloudy

Fair,

cloudy

Light rain

Fair,

cloudy

Mostly

cloudy

Few

clouds

Temperature

80°F

(27°C)

77°F

(25°C)

84°F

(29°C)

87°F

(31°C)

68°F

(20°C)

76°F

(24°C)

Humidity

69%

82%

80%

67%

47%

54%

Barometer

30.07 in

(1018.0

mb)

30.06 in

(1018.1

mb)

29.92 in

(1013.0

mb)

29.88 in

(1011.9

mb)

30.21 in

(1023.0

mb)

30.08 in

(1018.8

mb)

Sus Scrofa (domestic pig) rib selection and preparation

Sus scrofa (domestic pig) bone was chosen as a proxy for human bone in this study

because pigs have long served as a valuable analogue for human remains in forensic research

including studies of decomposition (Weitzel, 2005; Pakosh and Rogers, 2009; Matuszewski et

al., 2019) and taphonomy (Janjua and Rogers, 2008; Huculuk and Rogers, 2009; Roberts and

Dabbs, 2015). Additionally, pig bone has been frequently analyzed in both archaeological

(Hamilton et al., 2009; Mulville et al., 2009; Guiry et al., 2012) and modern (Warinner and

Tuross, 2009; Frémondéau et al., 2012) isotopic studies, including controlled feeding studies

(Tuross et al., 2008; Webb et al., 2016; 2017a; 2017b). However, while bone collagen and

enamel carbonate are frequently analyzed in isotopic studies using pig remains (Frémondéau et

al., 2012), there are few published studies that analyze pig bone carbonate.

For the last thirty years, pigs have been the most frequently used proxy for human

remains in forensic science because they are easy to acquire, cost effective, ethical, and allow for

easy control of individual variables (Matuszewski et al., 2019). Furthermore, pigs and humans

share many physiological and anatomical characteristics including omnivorous diet, tissue

density, skeletal structure, internal anatomy, and fat-to-muscle ratio (Connor et al., 2018). Pig

bone is microscopically representative of human bone due to similarities in lamellar structure,

mineral density (Aerssens et al., 1998), and remodeling processes (Pearce et al., 2007).

Therefore, for the purposes of this study, pig bone was determined to be the most appropriate

proxy for human bone.

The pig ribs to be used in this study were sourced from a local butcher and meat

wholesaler in Wimamua, Florida. According to the distributer, the pig selected for this study was

classified as a “hog” due to their weight of over 120 pounds. This is a solely weight specific

classification and does not refer to a difference in speciation. Therefore, the term “pig” is used

throughout this thesis in reference to the species rather than the weight. The pig was originally

believed to have been raised in Central Florida, however, it was later determined that the pig was

sourced from dealer in Tennessee that purchases and relocates pigs to and from a variety of

locations throughout the continental United States. While the exact origin of the pig remains

uncertain, the δ18O signature of the bone carbonate suggests that the pig was not from the South

Eastern United States, but may have been born and raised in the North Eastern, South Western,

or Western United States (Kendall and Coplen, 2001). The pig was sacrificed at approximately

six months old on June 26th, 2020, the day before sample acquisition. Additionally, the sex of the

pig is unknown. All ribs used within this study were sourced from the same pig as it is assumed

that ribs from the same individual will have similar isotopic values (Fig. 3).

Figure 3. Pig selected for use in this study.

The ribs were not frozen before or after acquisition to ensure that the original isotopic

values were not compromised by severe temperature changes. Prior to submersion, all flesh,

muscle, and periosteum were removed from bone via manual maceration using a selection of

steel kitchen knives, steel wire brushes, and sterile steel scalpels. The ribs were not boiled or

exposed to water during the maceration process. Next, seven right ribs were manually sectioned

into fifty individual samples (approximately eight or nine samples per rib) measuring

approximately 1.5 inches from costal to sternal end, resulting in five samples per category, for a

total of 50 samples (Table 3, Fig. 4). The ribs were sectioned using a handheld Dremel® with

diamond wheel saw blades at a low rotation speed. Photographs were taken prior to maceration,

after maceration, and following sectioning into individual samples. Additionally, all samples

were weighed prior to submersion in water.

Table 3. Temporal and environmental distribution of samples processed.

Sample Type

Immediate

1-Month

3-Month

6-Month

Total

Control

Freshwater

Brackish Water

Distilled Water

Total

Figure 4. Sectioned rib samples prior to submersion.

Exposure to Aqueous Environments

Plywood shelves coated in polyurethane wood sealer were constructed to safely store the

submerged samples underneath the fume hood in the Laboratory for Bioarchaeological Sciences

(Room 150 A) on the University of Central Florida campus (Fig. 5). Prior to submersion, the

sectioned rib samples were randomly assigned a sample number, corresponding to a water type

and time interval (e.g., one-month, fresh water), to ensure that none of the sample groups

included all samples of the same size, cortical bone thickness, and/or location within the rib (e.g.,

sternal end, vertebral end). This randomization was meant to ensure that any observed isotopic

variation following submersion could be attributed to the submersion, rather than characteristics

of the samples themselves. Thus, different sections of the same ribs ended up in different

conditions. The samples were then placed in 250 milliliter Pyrex beakers containing 150

milliliters of either distilled water, freshwater, or brackish water. The samples were left to soak

in the water filled beakers for a period of one, three, or six months, depending on the sample.

Control samples were not exposed to water and were processed shortly after sectioning to

provide a basis for comparison of isotopic values. Once the submersion period had ended, the

samples were removed from the water, photographed, weighed, and prepared for analysis

following the laboratory protocol for carbonate and Fourier transform infrared spectroscopy

(FTIR) analysis.

Figure 5. Plywood shelf holding submerged samples in water-filled beakers.

Laboratory protocol

All sample preparation took place in the Laboratory for Bioarchaeological Sciences at the

University of Central Florida following the laboratory protocol for carbonate analysis based on

previously published literature (Lee-Thorp and van Der Merwe, 1987; Lee-Thorp and van Der

Merwe, 1991; Koch et al., 1997; Garvie-Lok et al., 2004). This method aims to remove both

organic materials and exogenous carbonates from bone. First, the samples were ground into a

fine powder of less than 180 microns but greater than 63 microns separated via graded mesh

sieves. In this study, samples were powdered using a handheld Dremel® with a silicone carbide

grinding stone attachment (Fig. 6). To remove adhered materials, such as any remaining

periosteum, and/or exogenous contaminants, the outer layer of cortical bone of each sample was

removed via Dremel® prior to collecting powdered sample for analysis. The desired weight of

each sample is approximately 20-30 milligrams. Additionally, 10% of samples (n=5) were

duplicated, meaning that two vials of bone carbonate powder were collected per sample and

processed in different batches. Duplicated samples are used to demonstrate consistency between

batches and/or identify inconsistencies in processing methods that may skew the resulting

isotopic data. Similarly, the five control samples (unsubmerged) were processed in different

batches to provide an additional measure of inter-batch reliability.

Figure 6. Powdering sample using a handheld Dremel®.

Samples are treated in 0.04 ml of 2% reagent-grade bleach per mg of sample to break

down organic material (primarily collagen). Samples were agitated every 24 hours using a vortex

mixer. After 72 hours, the samples were thoroughly rinsed with distilled water a minimum of

five times. Next, samples were soaked in 0.04 ml of 0.1M acetic acid per mg of sample,

uncapped, covered with plastic wrap, and left at room temperature for four hours to remove

accumulated diagenetic carbonates. Samples were agitated every hour using a vortex mixer.

After rinsing, the vials were frozen to avoid possible isotopic exchange between excess water

and sample. Samples were then placed in a freeze dryer for approximately 24 hours, until

completely dry. After removal from freeze dryer, the weight of the treated sample was calculated

and recorded. The yield was calculated to determine carbonate preservation. The following

formula was used to calculate yield:

(Treated sample ÷ Untreated sample) x 100.

Conversion Equations

Oxygen isotopic ratios are compared to standards known as Peedee Belemnite (VPDB)

and Standard Mean Ocean Water (VSMOW). To accurately compare oxygen isotopic data sets to

one another, the isotopic standards must be the same. There are multiple linear regression

equations available in the literature to convert isotopic values from VPDB to VSMOW, and vis

versa (Coplen et al., 1983; Daux et al., 2008; Chenery et al., 2012). However, certain formulae

require addition conversions depending on the sampled tissue component (e.g., carbonate,

phosphate). The tissue component utilized in this thesis is bone carbonate. Therefore, the

following regression formulae was selected to convert between VPDB and VSMOW in this

study (Coplen et al., 1983):

δ18OVSMOW = (1.03091 x δ18OVPDB) + 30.91

Following conversion from VPDB to VSMOW, an additional conversion is necessary to

allow for comparison between the VSMOW standard and modern drinking water. The following

regression formulae was selected to convert between VSMOW and drinking water in this study

as it does not require additional conversions based on the selected tissue component (Chenery et

al., 2012):

δ18ODW = (1.59 x δ18OVSMOW) – 48.634

IRMS

The prepared samples were packaged and sent to the Light Stable Isotope Mass

Spectrometry Laboratory in the Department of Geological Sciences at the University of Florida

in Gainesville, Florida. A Finnigan-MAT DeltaPlus isotope ratio mass spectrometer with a Kiel

III carbonate preparation device was used to analyze the bone carbonate samples and obtain

oxygen isotope (δ18O) ratios from the prepared bone carbonate. A Picarro L2120-I Isotopic

Liquid Water and Water Vapor Analyzer was used to analyze the water samples from each

collection site and obtain oxygen isotope (δ18O) ratios.

IRMS mass spectrometer data precision was quantified for carbonate analysis by

comparing international standard reference materials (SRM) with known isotopic values to the

measured IRMS values (Table 4). During bone carbonate analysis, two SRM, NBS-18 and NBS-

19, were used over two analytical runs. NBS-18 was measured twice, and NBS-19 was measured

eight times during each analytic run.

Table 4. Standard reference material (SRM) used for IRMS calibration in this study,

formatted according to guidelines provided by Szpak et al., 2017. SRM values as described

by Brand et al. 2014.

Standard

IRMS Analysis

Material

Type

ẟ18O (‰, VPDB)

NBS-18

Apatite

Carbonatite

International

–2.2‰

NBS-19

Apatite

Limestone

International

–23.01‰

FTIR Protocol

Bone carbonate was evaluated for diagenetic alteration via Fourier Transform Infrared

Spectroscopy (FTIR) (Fig. 7). The infrared spectra produced by FTIR is used to evaluate two

characteristics of the mineral component of bone –the crystalline structure and the carbonate

content (Wright and Schwarcz, 1996; Yoder and Bartelink, 2010; Beasley et al., 2014). The

crystalline structure is evaluated via the Crystallinity Index (CI), also known as the Infrared

Splitting Factor (IRSF), while the carbonate content is evaluated via the ratio of absorbances

between the CO3 and PO4 peaks (C/P). The CI is calculated by adding the values recorded by the

FTIR at the split phosphate peak (565cm-1 and 605cm-1) and dividing the resulting value by the

value of the valley between peaks (595cm-1) (Shemesh, 1990; Wright and Schwarcz, 1996;

Yoder and Bartelink, 2010). While the expected range of CI is variable depending on the tissue,

age, treatment, and preservation state of the sample, 2.1–4.0 was selected as the expected range

for this study as this range is specific to natural (not chemically treated) modern bone (France et

al., 2020). A CI of a higher or lower value suggests potential diagenetic alteration. This equation

used to calculate CI is as follows:

Crystallinity Index = (A605 + A565)/ A595

The C/P is calculated by dividing the value of the absorbances of the CO3 peak (1415cm-

1) and PO4 peak (1035cm-1). The phosphate peak at 1035cm-1 is selected for this calculation

because it is the main phosphate absorbance peak and is not impacted by the splitting of the

phosphate peak at 565cm-1 and 605cm-1 (Wright and Schwarcz, 1996; Yoder and Bartelink,

2010; Beasley et al., 2014). Similar to CI, the expected range of C/P is variable based on the

characteristics of the sample. The range of 0.20–0.43 was selected as the expected range in this

study as it is specific to natural (not chemically treated) modern bone (France et al. 2020). Any

value above or below this value indicating potential diagenetic alteration. This equation to

calculate C/P is as follows:

Carbonate Content= A1415/A1035

Furthermore, the presence of non-apatite minerals within each sample will be noted as

they may be indicative of diagenetic alteration. In particular, peaks in the infrared spectra at

750cm-1 and 1096cm-1 indicate the presence of calcite (CaCO3) and francolite (a variety of

fluorapatite), respectively (Shemesh, 1990; Wright and Schwarcz, 1996).

FTIR samples were prepared by grinding bone samples to a fine powder (between 45 and

63 microns), measured through a sieve, using a handheld Dremel® with silicone carbide grinding

attachment. Approximately 10-15mg of powder were collected in a 2 mm vial and stored until

analysis. Samples were not chemically treated prior to analysis. The Spectra Two FTIR

Spectrometer, located on the University of Central Florida campus, was used to analyze the FTIR

of the bone samples for this research following the established protocols.

Prior to the analysis of each sample, the FTIR-ATR attachment and diamond plate were

cleaned with ethanol. Additionally, the “background check” function was run between each

sample to ensure no erroneous results caused by contaminants on the diamond plate. The

absorbance spectra were plotted from 1800cm-1 to 450cm-1. Each sample was scanned 8 times to

produce the resulting infrared spectra. Approximately 3–5mg of powdered sample was placed on

the diamond plate for each analysis. Five duplicate samples were analyzed to ensure reliability

and reproducibility of the resulting infrared spectra.

Figure 7. Example FTIR Spectra of sample EKM-50 with labelled peaks.

Analytical Strategy

Statistical Analyses

Prior to statistical analysis, the two sets of data (i.e., baseline water data, pig rib data)

were organized using Microsoft Excel. Next, descriptive statistics, including the mean, standard

deviation, standard variance, and range, were calculated for both data sets. While the bone

carbonate data includes both δ18O and δ13C, only δ18O will be statistically analyzed for the

purposes of this research. However, δ13C data will be used for the sole purpose of creating a

biplot to illustrate the range of variation among samples following exposure and will not be

analyzed statistically.

For the purposes of this research, the alpha level was set at 0.05 for all statistical

analyses. Each variable was tested for homogeneity of variance and normality prior to

conducting additional statistical analyses because homogeneity and normality, along with

linearity and independence, are assumptions of parametric tests (i.e., t-tests, ANOVA, etc.). If

these tests show that the data do not meet the assumptions of a parametric test, statistical

analyses must be conducted using a non-parametric alternative (i.e., Kruskal-Wallis, Mann-

Whitney, etc.) instead. Therefore, each variable was tested for normality using the Shapiro-Wilks

test in the stats package (stats package; R core team), the D’Agostino Omnibus test in the fbasics

package, (fbasics package; Wuertz et al., 2017), and the Anderson Darling test in the nortest

package (nortest package; Gross and Ligges, 2015). Additionally, each variable was tested for

homogeneity of variance using the Bartlett and Levene test in the car package (car package; Fox

et al., 2020).

These tests of homogeneity of variance and normality determined the data to be non-

parametric. Therefore, R statistical software was used to perform Kruskal-Wallis tests using

quantitative δ18O isotopic data to compare variables and determine if they are statistically

significantly different. The Kruskal-Wallis test was used to compare the difference between two

or more groups (i.e., data from each temporal category) (Lewis, 2013). This test is appropriate

for comparing quantitative data with small sample sizes. Additionally, both scatter plots and box

and whisker plots were created to visually assess and present the data.

Hypotheses

To evaluate the impact of diagenetic processes on the δ18O isotopic values of bone

carbonate exposed to distinct aqueous environments for different intervals, I have established the

following testable hypotheses:

1. H0: There will be no statistically significant difference between δ18O isotopic values

obtained from bone exposed to different aqueous environments (distilled water, fresh

water, and brackish water) and the control values.

2. H0: There will be no statistically significant difference between δ18O isotopic values

obtained from bone exposed to an aqueous environment for various intervals (one, three,

and six months) and the control values.

3. H0: There will not be a statistically significant difference between δ18O isotopic values

obtained from bone exposed to different aqueous environments (distilled water, fresh

water, and brackish water) for various intervals (one, three, and six months) and the

control values.

Chapter Summary

This chapter provided a detailed overview of the methodology used in this thesis. The

first half of this chapter discussed the procurement of necessary materials including water

selection and collection and the collection and preparation of the pig ribs used as a proxy for

human remains. The second half of this chapter discussed the preparation and pretreatment of

samples for carbonate analysis and FTIR analysis, as well as the analytical strategy. The

following chapter will discuss the results of all statistical analyses.

CHAPTER FOUR: RESULTS

The results of the stable oxygen isotope analyses of pig bone carbonate are presented in

this chapter, along with data regarding the preservation and accuracy of duplicate samples.

Oxygen isotopic baseline data for the state of Florida are presented. Additionally, FTIR data,

including the CI and C/P, are detailed. The pig bone carbonate data were analyzed using

descriptive statistics and Kruskal-Wallis tests were run to determine if there were statistically

significant differences between sample groups.

IRMS Calibration

The precision of the SRM NSB-19, averaged over two analytical sessions, was 0.043‰

for δ18Osc. This averaged error margin of less than one per mil (‰) suggests that the IRMS

measurements over both analytical sessions were precise and accurate.

Sample Duplication

Ten percent (n=5) of samples were duplicated to evaluate consistency of bone carbonate

preparation and pretreatment between sample batches. Each of these samples were also

duplicated for FTIR analysis to evaluate the consistency of the preparation and analysis of FTIR

samples. Unfortunately, there was an error during the pretreatment of the three-month batch, Ba-

20-13, as evidenced by the erroneous isotopic data of the duplicate sample (EKM-15) in that

batch. Based on these data, the decision was made to remove all samples from batch Ba-20-13,

comprised of the three-month sample group, from the analytical tests, results, and interpretations.

Therefore, all analytical analyses, results, and interpretations of δ18Osc are based solely on the

control, one, month, and six-month sample groups. However, the FTIR samples in this study

were not subject to pretreatment so the three-month sample group was not removed from FTIR

analyses, results, and interpretations.

Duplicated bone carbonate samples produced an average difference of 0.67‰ for δ18Osc

including EKM-15 and 0.13‰ following removal of this sample. Duplicated FTIR samples

produced an average difference of 0.028 for CI and 0.011 for C/P calculations (Table 5). Apart

from batch Ba-20-13, the precision demonstrated by the duplicated bone carbonate and FTIR

samples indicates consistency in preparation and pretreatment processes.

Table 5. List of duplicated samples including differences in δ18Osc (‰), CI, and C/P.

Samples Duplicated

Differences in ẟ18Osc

(‰, VPDB)

Differences in

FTIR (CI)

Differences in

FTIR (C/P)

EKM-04

0.05

0.002

0.007

EKM-15

2.82

0.068

0.012

EKM-22

0.30

0.047

0.024

EKM-35

0.16

0.014

0.006

EKM-48

0.02

0.008

0.004

Average Accuracy

0.67

0.028

0.011

Average Accuracy

(after EKM-15 removal)

0.13

N/A

Pooled Standard

Deviation

0.89

0.027

0.011

Pooled Standard

Deviation

(after EKM-15 removal)

0.12

N/A

Baseline Data

Water samples were collected from the Tomoka River Basin in Tomoka State Park

(Volusia County, Florida) and Lake Claire on the University of Central Florida Campus (Orange

County, Florida) in June 2020, October 2020, and January 2021 for the purpose of constructing

ẟ18O isotopic baseline data for the Central Florida region. The isotopic data produced from these

samples provide an isotopic baseline for the aqueous environment to which the pig rib samples

were directly exposed (table 6). These samples were collected multiple times throughout the

year, providing a range, encompassing multiple seasons, of ẟ18O isotopic values of –1.92 to –

2.94‰ for Volusia and Orange County in Central Florida.

Additionally, seasonal variation in the average δ18O isotopic values of each water type

were observed (table 7). For fresh water, the average δ18O varied seasonally from 2.41‰ in June,

0.72‰ in October, and 0.95‰ in January. For brackish water, the average δ18O varied seasonally

from 0.77‰ in June, –1.77‰ in October, and 0.83‰ in January. Distilled water, measured only

once in early February 2021, produced an average δ18O of –1.39‰.

Table 6. List of water samples from each collection site including the type, source,

collection date, and δ18O (‰, VSMOW) isotopic data.

Sample #

Water Type

Water Source

Collection

Date

ẟ18O

(‰, VSMOW)

TSP- 01

Brackish

Tomoka River Basin

06/17/2020

1.26

TSP-02

Brackish

Tomoka River Basin

06/17/2020

0.19

TSP-03

Brackish

Tomoka River Basin

06/17/2020

0.86

TSP-04

Brackish

Tomoka River Basin

10/11/2020

–1.61

TSP-05

Brackish

Tomoka River Basin

10/11/2020

–1.93

TSP-06

Brackish

Tomoka River Basin

01/30/2021

0.81

TSP-07

Brackish

Tomoka River Basin

01/30/2021

0.78

TSP-08

Brackish

Tomoka River Basin

01/30/2021

0.89

LC-01

Fresh

Lake Claire

06/18/2020

2.13

LC-02

Fresh

Lake Claire

06/18/2020

2.15

LC-03

Fresh

Lake Claire

06/18/2020

2.94

LC-04

Fresh

Lake Claire

10/11/2020

0.85

LC-05

Fresh

Lake Claire

10/11/2020

0.59

LC-06

Fresh

Lake Claire

01/30/2021

0.90

LC-07

Fresh

Lake Claire

01/30/2021

1.01

LC-08

Fresh

Lake Claire

01/30/2021

0.93

D-01

Distilled

UCF (Laboratory for

Bioarchaeological Sciences)

02/01/2021

–1.35

D-02

Distilled

UCF (Laboratory for

Bioarchaeological Sciences)

02/01/2021

–1.44

Table 7. Seasonal variation in δ18O (‰, VSMOW) for each water type.

Collection

Month

June

October

January/February

Water Type

Fresh

Brackish

Fresh

Brackish

Fresh

Brackish

Distilled

Seasonal

Average ẟ18O

(‰, VSMOW)

2.41

0.77

0.72

-1.77

0.95

0.83

–1.39

The data from the water samples collected for this study were supplemented by

previously published δ18O data from a variety of sources in the state of Florida, but not specific

to Central Florida, including rivers, lakes, and tap water (Fig. 8). Brooks et al. (2014) provided

δ18O data for Floridian lakes, ranging from –1.90 to 4.20‰. Riverine δ18O data from Kendall and

Coplen (2001) provides a range of –4.50 to 2.34‰. Coplen et al. (2013) and Bowen et al. (2008)

provide the δ18O range of Floridian tap water as –4.08 to –0.39‰ and -2.80 to –1.10‰,

respectively.

Figure 8. Range plot depicting ranges of ẟ18O (‰, VSMOW) provided by previously

published literature.

Overall Pig Rib Samples

Overall, 50 samples were analyzed in this study, including control samples that were not

exposed to water (n=5), samples exposed to fresh water (n=15), samples exposed to brackish

water (n=15), and samples exposed to distilled water (n=15). However, after the removal of

batch Ba-20-13, the fresh, brackish, and distilled water sample groups were left with only ten

samples each, for a total of 35 samples (including the control group). The samples were

originally analyzed in reference to the VPDB standard but were converted to VSMOW and

drinking water (DW) (Table 8). However, all statistical analyses were conducted using the

VPDB δ 18O (‰) data. Table 8 also presents δ13C data even though these data were not included

in statistical analyses.

Table 8. Sample list including interval of aqueous exposure, pretreatment batch number,

δ18O (‰, VPDB, VSMOW, DW), and δ13C results. (* indicates that sample was removed

from statistical analyses due to sample processing error).

Sample #

Water

Type

Interval

Batch #

δ18O

(‰,

VPDB)

δ18O (‰,

VSMOW)

δ18O

(‰,

DW)

δ13C (‰,

VPDB)

EKM-01

Control

n/a

Ba-20-12

–7.22

23.47

–11.32

–4.48

EKM-02*

Control

n/a

Ba-20-13

–57.44

–28.30

–93.63

–36.90

EKM-03*

Control

n/a

Ba-20-13

–29.61

0.39

–48.02

–17.56

EKM-04-01

Control

n/a

Ba-20-12

–6.93

23.77

–10.84

–4.94

EKM-04-02

Control

n/a

Ba-20-14b

–6.88

23.82

–10.76

–5.00

EKM-05

Control

n/a

Ba-20-14b

–7.06

23.63

–11.07

–4.79

EKM-06

Brackish

1 month

Ba-20-12

–6.80

23.89

–10.64

–4.68

EKM-07

Brackish

1 month

Ba-20-12

–6.79

23.91

–10.62

–4.66

EKM-08

Brackish

1 month

Ba-20-12

–7.10

23.59

–11.12

–4.55

EKM-09

Brackish

1 month

Ba-20-12

–6.95

23.74

–10.88

–4.49

EKM-10

Brackish

1 month

Ba-20-12

–6.80

23.90

–10.63

–4.61

EKM-11*

Brackish

3 month

Ba-20-13

–9.71

20.90

–15.40

–6.02

EKM-12*

Brackish

3 month

Ba-20-13

–10.51

20.08

–16.71

–6.37

EKM-13*

Brackish

3 month

Ba-20-13

–9.93

20.67

–15.76

–6.54

EKM-14*

Brackish

3 month

Ba-20-13

–9.70

20.91

–15.39

–6.14

EKM-15-01*

Brackish

3 month

Ba-20-13

–9.81

20.80

–15.57

–6.12

EKM-15-02

Brackish

3 month

Ba-20-14b

–6.99

23.71

–10.94

–4.50

EKM-16

Brackish

6 month

Ba-20-14a

–6.79

23.91

–10.62

–4.52

EKM-17

Brackish

6 month

Ba-20-14a

–6.80

23.90

–10.64

–4.35

EKM-18

Brackish

6 month

Ba-20-14a

–6.74

23.96

–10.54

–4.33

EKM-19

Brackish

6 month

Ba-20-14a

–6.52

24.18

–10.18

–4.55

EKM-20

Brackish

6 month

Ba-20-14a

–6.58

24.13

–10.27

–4.41

EKM-21

Fresh

1 month

Ba-20-12

–6.69

24.01

–10.45

–4.57

EKM-22-01

Fresh

1 month

Ba-20-12

–6.76

23.94

–10.57

–4.69

EKM-22-02

Fresh

1 month

Ba-20-14b

–7.06

23.64

–11.05

–4.58

EKM-23

Fresh

1 month

Ba-20-12

–7.00

23.70

–10.96

–4.38

EKM24

Fresh

1 month

Ba-20-12

–6.89

23.81

–10.77

–4.48

EKM-25

Fresh

1 month

Ba-20-12

–6.73

23.97

–10.52

–4.55

EKM-26*

Fresh

3 month

Ba-20-13

–16.48

13.92

–26.51

–9.47

EKM-27*

Fresh

3 month

Ba-20-13

–7.90

22.77

–12.43

–4.91

Sample #

Water

Type

Interval

Batch #

δ18O

(‰,

VPDB)

δ18O (‰,

VSMOW)

δ18O

(‰,

DW)

δ13C (‰,

VPDB)

EKM-28*

Fresh

3 month

Ba-20-13

–9.02

21.61

–14.27

–5.27

EKM-29*

Fresh

3 month

Ba-20-13

–10.80

19.77

–17.20

–6.32

EKM-30*

Fresh

3 month

Ba-20-13

–9.35

21.27

–14.81

–5.31

EKM-31

Fresh

6 month

Ba-20-14a

–6.29

24.42

–9.80

–4.24

EKM-32

Fresh

6 month

Ba-20-14a

–6.75

23.95

–10.56

–4.26

EKM-33

Fresh

6 month

Ba-20-14a

–6.33

24.38

–9.86

–4.27

EKM-34

Fresh

6 month

Ba-20-14a

–6.39

24.32

–9.96

–4.26

EKM-35-01

Fresh

6 month

Ba-20-14a

–6.49

24.22

–10.12

–4.13

EKM-35-02

Fresh

6 month

Ba-20-14b

–6.65

24.06

–10.38

–4.12

EKM-36

Distilled

1 month

Ba-20-12

–7.03

23.66

–11.01

–4.32

EKM-37

Distilled

1 month

Ba-20-12

–7.01

23.68

–10.99

–4.57

EKM-38

Distilled

1 month

Ba-20-12

–7.21

23.47

–11.31

–4.57

EKM-39

Distilled

1 month

Ba-20-12

–6.94

23.76

–10.86

–4.53

EKM-40

Distilled

1 month

Ba-20-12

–7.02

23.67

–10.99

–4.57

EKM-41*

Distilled

3 month

Ba-20-13

–10.28

20.31

–16.34

–6.79

EKM-42*

Distilled

3 month

Ba-20-13

–9.76

20.85

–15.48

–5.96

EKM-43*

Distilled

3 month

Ba-20-13

–9.12

21.51

–14.43

–5.65

EKM-44*

Distilled

3 month

Ba-20-13

–14.67

15.78

–23.54

–8.48

EKM-45*

Distilled

3 month

Ba-20-13

–9.15

21.47

–14.49

–5.44

EKM-46

Distilled

6 month

Ba-20-14a

–6.67

24.04

–10.42

–4.35

EKM-47

Distilled

6 month

Ba-20-14a

–6.65

24.05

–10.39

–4.39

EKM-48-01

Distilled

6 month

Ba-20-14a

–6.57

24.14

–10.26

–4.31

EKM-48-02

Distilled

6 month

Ba-20-14b

–6.59

24.12

–10.29

–4.28

EKM-49

Distilled

6 month

Ba-20-14a

–6.21

24.51

–9.67

–4.24

EKM-50

Distilled

6 month

Ba-20-14a

–6.53

24.18

–10.19

–4.46

Descriptive Statistics

The mean δ18O (in VPDB) of all bone samples (excluding batch Ba-20-13) is –6.77‰,

with a standard deviation of 0.26‰ and a range from –6.21‰ to –7.22‰. Descriptive statistics

for each of the sample groups, not separated by exposure interval, are presented in Table 9.

Table 9. Descriptive statistics of δ18O (‰, VPDB) for each sample group.

Water Type

Control

(n=5)

Fresh

(n=10)

Brackish

(n=10)

Distilled

(n=10)

Mean

–7.07

–6.63

–6.79

Std. Dev.

0.15

0.24

0.16

0.31

Std. Error

0.08

0.05

0.10

Variance

0.02

0.06

0.03

0.09

Range

0.29

0.70

0.57

1.00

Descriptive statistics for each of the water types, separated by exposure interval, are

presented in Table 10. While δ13C was not considered as part of the analysis, the data are used to

create a visualization depicting the variation of data across all sample groups, as is typical in

isotopic analysis of carbonate. Therefore, a scatterplot was created to illustrate the relationship

between δ18O and δ13C for samples exposed to each water type for both time intervals (Fig. 9).

Table 10. Descriptive statistics of δ18O (‰, VPDB) for carbonate samples exposed to each

water type and time interval.

Water Type

Control

Fresh

Brackish

Distilled

Interval

n/a

1 mo.

6 mo.

1 mo.

6 mo.

1 mo.

6 mo.

Mean

–7.07

–6.81

–6.45

–6.89

–6.69

–7.04

–6.53

Std. Dev.

0.15

0.13

0.19

0.13

0.10

0.19

Std. Error

0.08

0.06

0.08

0.06

0.05

0.08

Variance

0.02

0.03

0.02

0.01

0.03

Range

0.29

0.31

0.46

0.30

0.27

0.28

0.46

Figure 9. Scatterplot illustrating the relationship between ẟ18O (‰, VPDB) and ẟ13C (‰,

VPDB) for all samples. Shapes and colors differentiate between samples of each water and

interval group.

Water Type Comparisons

A Kruskal-Wallis test was conducted to determine if there were statistically significant

differences (p-value = 0.05) between the control groups and samples exposed to water for one

month (Table 11), between the control group and samples exposed to water for six months

(Table 12), and between samples exposed to water for one and six months (Table 13).

Furthermore, box and whisker plots were constructed to visualize the differences between the

control and exposure groups for each water type (Fig. 10, 11, 12).

No statistically significant differences were produced between the control group and

samples exposed to fresh, brackish, or distilled water for one month. Conversely, statistically

significant differences were produced between the control group and samples exposed to fresh

(p-value = 0.01), brackish (p-value = 0.01), and distilled water (p-value = 0.01) for six months as

the oxygen isotopic composition of the bone slowly became more positive. Next, the samples

exposed to the same water type for one month and six months were compared for each water

type. Fresh and brackish water did not produce a statistically significant difference while distilled

water did produce a statistically significant difference (p-value = 0.02).

Table 11. Results of Kruskal Wallis test between the δ8O (‰, VPDB) of

the control group and samples exposed to each water type for one month.

Comparison

Control/ One Month

Water Type

Fresh

Brackish

Distilled

H- Statistic

–1.36

–1.03

–0.28

P-Value

0.09

0.15

0.39

Table 12. Results of Kruskal Wallis test between the δ18O (‰, VPDB) of

the control group and samples exposed to each water type for six months.

Comparison

Control/ Six Months

Water Type

Fresh

Brackish

Distilled

H- Statistic

–2.91

–2.66

–2.46

P-Value

0.01

Table 13. Results of Kruskal Wallis test between the δ18O (‰, VPDB) of samples

exposed to each water type for one month and six months.

Comparison

One Month/Six Months

Water Type

Fresh

Brackish

Distilled

H- Statistic

–1.79

–1.88

–2.52

P-Value

0.07

0.06

0.02

Figure 10. Box and whisker plot illustrating the difference in δ18O (‰, VPDB) between the

control group, samples exposed to fresh water for one month, and those exposed for six

months.

Figure 11. Box and whisker plot illustrating the difference in δ18O (‰, VPDB) between the

control group, samples exposed to brackish water for one month, and those exposed for six

months.

Figure 12. Box and whisker plot illustrating the difference in δ18O (‰, VPDB) between the

control group, samples exposed to distilled water for one month, and those exposed for six

months.

Interval Comparison

A Kruskal-Wallis test was conducted to determine if there were statistically significant

differences (p-value = 0.05) between samples exposed to different water types for the same time

interval. This was determined by comparing samples exposed to fresh water for one month to

those exposed to brackish water for one month. The samples exposed to these water types for six

months were also compared (table 14). Additionally, samples exposed to fresh water for one

month were compared to those exposed to distilled water for one month. The samples exposed to

these water types for six months were also compared (table 15). Finally, comparisons were made

between samples exposed to brackish water for one month and those exposed to distilled water

for one month, along with the samples exposed to these water types for six months (table 16).

Furthermore, box and whisker plots were created to illustrate differences between samples

exposed to each water type for the same interval (Fig. 13 and 14). While the samples exposed to

fresh water for one month were statistically significantly different from the samples exposed to

distilled water for one month (p-value = 0.02), none of the other comparisons produced

statistically significant differences.

Table 14. Results of Kruskal-Wallis test between samples

exposed to fresh and brackish water for one and six months.

Comparison

Fresh/Brackish

Interval

One Month

Six Month

H-Statistic

–1.06

–2.05

P-Value

0.14

0.06

Table 15. Results of Kruskal-Wallis test between samples

exposed to fresh and distilled water for one and six months.

Comparison

Fresh/Distilled

Interval

One Month

Six Month

H-Statistic

–2.55

–0.71

P-Value

0.02

0.24

Table 16. Results of Kruskal-Wallis test between samples

exposed to brackish and distilled water for one and six months.

Comparison

Brackish/Distilled

Interval

One Month

Six Month

H-Statistic

1.49

-3.4

P-Value

0.14

0.17

Figure 13. Box and whisker plot illustrating the difference in ẟ18O (‰, VPDB) between the

control group and samples exposed to each water type for one month.

Figure 14. Box and whisker plot illustrating the difference in ẟ18O (‰, VPDB) between the

control group and samples exposed to each water type for one month.

FTIR Results

Both the crystallinity index (CI) and carbonate content (C/P) were calculated for all fifty

samples and five duplicates (Table 17), including the samples from batch Ba-20-13 which were

removed from previous analytical tests due to a sample processing error during the carbonate

pretreatment. Since, FTIR samples are not subject to pretreatment (See Metcalfe et al., 2009 and

Munro et al., 2008 for examples of FTIR analysis on untreated samples) all fifty samples were

analyzed for CI and C/P. Additionally, none of the spectra were baseline corrected. Descriptive

statistics for the CI and C/P of the control samples, the samples exposed to fresh water, the

samples exposed to brackish water, and the samples exposed to distilled water are presented in

Tables 18–21. Additionally, a scatterplot was created to illustrate the relationship between the

δ18O (‰, VPDB) and the CI of samples, excluding samples from batch Ba-20-13 (Fig. 15).

Table 17. Sample list including results of FTIR analyses.

Sample #

Crystallinity

Index (CI)

Carbonate Content

(C/P)

EKM-01

2.352

0.344

EKM-02

2.321

0.359

EKM-03

2.296

0.366

EKM-04-01

2.343

0.366

EKM-04-02

2.346

0.374

EKM-05

2.358

0.364

EKM-06

2.400

0.334

EKM-07

2.330

0.369

EKM-08

2.368

0.344

EKM-09

2.388

0.336

EKM-10

2.360

0.327

EKM-11

2.368

0.351

EKM-12

2.384

0.346

EKM-13

2.387

0.334

EKM-14

2.364

0.367

EKM-15-01

2.440

0.343

EKM-15-02

2.372

0.356

EKM-16

2.398

0.360

EKM-17

2.370

0.364

EKM-18

2.376

0.380

EKM-19

2.347

0.365

EKM-20

2.353

0.377

EKM-21

2.324

0.366

EKM-22-01

2.333

0.334

EKM-22-02

2.381

0.358

EKM-23

2.391

0.352

EKM-24

2.380

0.361

Sample #

Crystallinity

Index (CI)

Carbonate Content

(C/P)

EKM-25

2.360

0.362

EKM-26

2.357

0.361

EKM-27

2.413

0.327

EKM-28

2.387

0.329

EKM-29

2.360

0.359

EKM-30

2.354

0.361

EKM-31

2.350

0.365

EKM-32

2.356

0.359

EKM-33

2.370

0.358

EKM-34

2.348

0.365

EKM-35-01

2.355

0.357

EKM-35-02

2.370

0.363

EKM-36

2.398

0.348

EKM-37

2.372

0.360

EKM-38

2.374

0.339

EKM-39

2.380

0.340

EKM-40

2.396

0.350

EKM-41

2.364

0.354

EKM-42

2.349

0.364

EKM-43

2.388

0.332

EKM-44

2.359

0.357

EKM-45

2.399

0.346

EKM-46

2.316

0.362

EKM-47

2.334

0.357

EKM-48-01

2.361

0.359

EKM-48-02

2.369

0.363

EKM-49

2.355

0.351

EKM-50

2.376

0.330

The CI of all samples fell within the expected range of 2.1–4.0 for well-preserved modern

bone carbonate. Similarly, the C/P of all samples fell within the expected range of 0.20–0.43 for

well-preserved modern bone carbonate. The CI and C/P of all samples, including the control

samples, are homogenous with very little variation according to aqueous exposure and interval.

Furthermore, the correlation coefficient (R2) for the relationship between δ18O (‰, VPDB) and

CI is 0.12, indicating that these two variables have a weak relationship and are not influenced by

one another (Fig. 10).

Table 18. Descriptive statistics for the CI and C/P of the control samples.

Interval

Control

Analysis

C/P

Mean

2.334

0.360

Std. Dev.

0.026

0.009

Range

0.062

0.022

Table 19. Descriptive statistics for the CI and C/P of the samples exposed to fresh water for

one, three, and six months.

Interval

One Month

Three Months

Six Months

Analysis

C/P

Mean

2.358

0.355

2.374

0.347

2.356

0.361

Std. Dev.

0.029

0.013

0.025

0.018

0.009

0.004

Range

0.067

0.032

0.059

0.035

0.022

0.009

Table 20. Descriptive statistics for the CI and C/P of the samples exposed to brackish water

for one, three, and six months.

Interval

One Month

Three Months

Six Months

Analysis

C/P

Mean

2.369

0.342

2.389

0.348

2.369

0.369

Std. Dev.

0.027

0.016

0.031

0.012

0.020

0.008

Range

0.071

0.041

0.076

0.033

0.050

0.020

Table 21. Descriptive statistics for the CI and C/P of the samples exposed to distilled water

for one, three, and six months.

Interval

One Month

Three Months

Six Months

Analysis

C/P

Mean

2.384

0.347

2.372

0.351

2.348

0.352

Std. Dev.

0.012

0.009

0.021

0.012

0.024

0.013

Range

0.026

0.021

0.051

0.034

0.060

0.032

Figure 15. Scatterplot depicting the relationship between ẟ18O (‰, VPDB) and CI.

Chapter Summary

While batch Ba-20-13, comprised of the samples exposed to aqueous environments for

three months, was removed from analyses, samples from each other batch and context were

statistically tested. Kruskal-Wallis tests were run, and statistically significant differences were

found in δ18O values between the control samples and samples exposed to fresh, brackish, and

distilled water for six months. Furthermore, FTIR samples were analyzed and CI and C/P were

calculated for each sample. All fifty samples were found to be within the expected range for

well-preserved modern bone. An interpretation of the isotopic variation seen in the oxygen data

and the homogeny seen in the FTIR data is presented in Chapter 5.

CHAPTER FIVE: DISCUSSION

This chapter will address the proposed research questions and hypotheses through the

examination of the oxygen isotopic variation present in the pig bone carbonate samples

following submersion within fresh, brackish, and distilled water for different intervals. Salinity,

microbial activity, and duration of the exposure will each be discussed in terms of their potential

contribution to the observed isotopic variation. Furthermore, the detection of diagenetic

alteration using the crystallinity index and carbonate content provided by FTIR analysis will be

considered. This chapter will conclude with a brief discussion regarding the implications of this

research for the use of stable isotope analysis in a medicolegal context.

Hypotheses Outcomes

1. H0: There will be no statistically significant difference between δ18O isotopic values

obtained from bone exposed to different aqueous environments (fresh, brackish, and

distilled water) and the control values.

This hypothesis was rejected as there were statistically significant differences between

the ẟ18O isotopic values obtained from bone exposed to each aqueous environment (fresh,

brackish, and distilled water) and the control values.

2. H0: There will be no statistically significant difference between ẟ18O isotopic values

obtained from bone exposed to an aqueous environment for various intervals (one, three,

and six months) and the control values.

This hypothesis was rejected as there were no statistically significant differences noted

between the ẟ18O isotopic value of bone exposed to aqueous environments for one month and the

control values. However, statistically significant differences were noted between the control

values and the ẟ18O isotopic values obtained from bone exposed to each aqueous environment

(fresh, brackish, and distilled) for six months.

3. H0: There will not be a statistically significant difference between ẟ18O isotopic values

obtained from bone exposed to different aqueous environments (distilled water, fresh

water, and brackish water) for various intervals (one, three, and six months) and the

control values.

This hypothesis was rejected because statistically significant differences in ẟ18O isotopic

values were noted between the samples exposed to fresh and distilled water for one month.

Excluding the control values, however, none of the other sample groups were statistically

significantly different from one another when compared based on water type and interval.

The isotopic variation observed following submersion in each water type is thought to

have been caused by isotopic exchange between the oxygen isotopes within the bone samples

and those within the water. Since the pig utilized in this study was non-local, the ẟ18O isotopic

composition of the control samples were much more negative (–11.32‰ to –10.84‰, converted

from VPDB to drinking water) than that of the local water analyzed in this study (–1.93‰ to

2.94‰) in which it was submerged. The difference between the original ẟ18O isotopic

composition of the bone samples and the local water allowed for isotopic exchange to occur and

slowly alter the ẟ18O composition of bone towards more positive values, over a period of six

months, as the bone shifted towards equilibrium with isotopic composition of the water to which

it was exposed. Had the pig been locally raised, the difference between the original oxygen

isotopic composition of the bone and the local water would have been much less notable,

possible at or near equilibrium, which may have obscured the observation of any isotopic

exchange following submersion. While the variation in this study is attributed largely to isotopic

exchange, there are a variety of other factors, including salinity, microbial activity, and

submersion interval, that may have also played a role in the observed variation.

Salinity

The fresh, brackish, and distilled water used in this study each maintain a different level

of salinity. The distilled water had a salinity level of 0.00 ppt, the fresh water had a salinity level

of 0.08 ppt, and the brackish water had a salinity level of 1.43 ppt. A salinity level of 1.43 ppt

places the brackish water used in this study within the range of oligohaline brackish water, the

least saline of the brackish water classifications. Despite quantifiable differences in salinity level

between the water types, salinity does not seem to be an influential factor in this study as

samples reacted to aqueous exposure in largely the same manner, regardless of salinity level. As

previously stated, there were no statistically significant differences between the control samples

and samples exposed to fresh, brackish, or distilled water for one month while each of these

water types produced statistically significant differences after samples were exposed for six

months.

Furthermore, with the exception of the samples exposed to fresh and distilled water for

one month, none of the samples exposed to distinct water types were statistically significantly

different from one another at either one or six month intervals. Consequently, at these time

durations salinity does not appear to be a differentiating factor in the rate or severity of

diagenetic alteration, up to oligohaline salinity levels, but higher salinity levels (e.g., mesohaline,

polyhaline, marine) may be a factor. Previous research suggests that long term exposure to

marine environments may impact the structural integrity of bone (Arnaud et al., 1978; Bell and

Elkerton, 2007; Pokines and Higgs, 2015; Eriksen et al., 2020), but these changes may be

attributed to a combination of other marine environmental factors (i.e., microbial attack by

cyanobacteria, scavenging by marine fauna, adhering algae/seaweed), rather than salinity.

Moreover, samples from this study were not subject to the natural variation in salinity

levels that occur on a seasonal basis in brackish water. While the samples in this study were

exposed to static salinity levels, skeletal remains deposited within a natural brackish water

feature would be subjected to seasonal variation in salinity levels that may span multiple salinity

classifications (i.e., oligohaline, mesohaline, polyhaline). Therefore, conclusions regarding the

impact of salinity in this study should not be extended to all salinity levels and/or exposure

intervals.

Microbial Activity

First and foremost, it is important to note that microbial activity was not quantified for

each water type during this study. Distilled water was originally assumed to be sterile while both

fresh and brackish water were assumed to maintain a higher level of microbial activity, relative

to the distilled water, as they were not intentionally sterilized after collection from terrestrial and

riverine aqueous environments where microbes (e.g., fungi, bacteria, protozoa) would certainly

be present. However, within ten days, mold began to form in a small number of the beakers filled

with each water type, including distilled water. Within 23 days, mold had begun to form in all

fifty beakers with no apparent difference in quantity according to water type. Once removed

from the water, mold accumulation was observed on the surface of all fifty samples. The color of

mold ranged from white, gray, and beige, to light green, with no apparent association between

the color of mold and the water type in which it developed. The rapid development of mold

suggests that the fresh, brackish, and distilled water used in this study each maintain a certain

level of microbial activity. Once developed, the mold was present in the beakers, floating in the

water and, in some cases, surrounding the sample, for the remainder of the study (Fig. 16).

Figure 16. Mold development in brackish (EKM-20), fresh (EKM-32)

and distilled (EKM-50) water at six months.

The introduction of microbes to the distilled water may be the result of the open vessel

method used in this study. Therefore, while microbial activity was not directly quantified, it may

have played a role in the diagenetic alteration producing the statistically significant differences

noted in this study, as previous research demonstrates that microbial activity is a critical

component of diagenetic processes of both the organic (collagen) and inorganic (carbonate)

portions of bone in terrestrial and aqueous depositional environments (Arnaud et al., 1978; Zazzo

et al., 2004; Jans et al., 2004; Bell and Elkerton, 2007). Microbial attack has been known to

cause both microscopic and macroscopic structural changes to bone by deteriorating both the

organic and mineral components (Child, 1995). More specifically, microbial attack may result in

two types of microscopic tunneling— Wedl tunneling, caused by aquatic microorganisms (e.g.,

cyanobacteria), and non-Wedl tunneling, caused by terrestrial bacteria (Eriksen et al., 2020),

depending on the depositional environment. While no microscopic analysis was conducted, there

were no macroscopic structural changes observed on the samples from this study following

submersion.

Interval

The results of this study indicate that exposure interval is a key factor in diagenetic

alteration and subsequent shifts in the isotopic composition of bone carbonate. Although no

statistically significant differences were noted between the control values and those of samples

exposed to aqueous environments for one month, small (non-significant) shifts in isotopic

composition were observed at this interval. Conversely, statistically significant differences were

noted between the control values and those of samples exposed to fresh, brackish, and distilled

water for six months. This suggests that while statistically significant shifts in the δ18O isotopic

composition of bone carbonate may take several months to develop, the early stages of

diagenetic alteration may begin within days or weeks of exposure to aqueous environments.

These findings are consistent with those of Zazzo et al. (2004), who concluded that exposure to

aqueous environments causes “rapid and significant” changes to both the ẟ18O and ẟ13C

composition of bone and enamel carbonate during the early stages of diagenesis.

The samples utilized in this study were manually macerated prior to aqueous exposure.

Therefore, the exposure intervals and subsequent diagenetic alteration and shifts in isotopic

composition presented in this thesis are specific to skeletonized remains. While six months of

aqueous exposure was long enough to create a statistically significant difference in the oxygen

isotopic composition of skeletonized remains, this same interval may not produce significant

differences if remains are fresh (fleshed) and have yet to decompose prior to submersion. The

diagenetic alteration of bones is not thought to begin until after soft tissues have decomposed,

leaving skeletal remains exposed to their depositional environment (Guareschi et al., 2021).

Although decomposition in an aqueous context often proceeds more slowly than in a dry,

terrestrial environment (Caruso, 2016), the complete skeletonization of human remains in water

is variable based on a multitude of factors (i.e., climate, clothing, current, etc.) and may take

anywhere from three weeks to 18 months or more (Haglund an Sorg, 1997). Therefore, when

evaluating the potential risk of diagenetic alteration and subsequent shifts in the isotopic

composition of bones recovered from aqueous environments, accurate estimations of both Post-

Mortem Interval (PMI) and Post-Mortem Submersion Interval (PMSI) are necessary, where PMI

is defined as an estimation of time since death and PMSI is defined as an estimation of the time

from submersion in an aqueous environment to recovery (Guareschi et al., 2021).

FTIR

FTIR analysis was used to quantify the crystallinity index (CI) and carbonate content

(C/P) of each sample following aqueous exposure to determine if the samples underwent

detectable structural changes. The CI and C/P values of each sample fell within the expected

range for modern, well preserved, untreated, bone carbonate. Furthermore, neither the CI nor C/P

values deviate notably from values of the control samples, regardless of the water type to which

the samples were exposed or the interval of exposure. While it is expected that diagenetic

alteration resulting in significant variations in the isotopic composition of bone carbonate will be

correlated with changes in the crystallinity index (Wright and Schwarcz, 1996; Nielsen-Marsh

and Hedges, 2000), previous studies have shown that isotopic shifts in both bone carbonate and

phosphate oxygen composition may occur with no indication of recrystallization provided by

FTIR analysis (Pucéat et al., 2004; Munro et al., 2007; Munro et al., 2008). Moreover, in a

similar study, Zazzo et al. (2004) concluded that early diagenetic alteration of modern bone

samples caused by aqueous exposure resulted in statistically significant shifts in isotopic

composition prior to any detectable change in either crystallinity or the content of organic matter.

Therefore, this research contributes to the body of evidence suggesting that the early stages of

diagenetic alteration may occur before the emergence of structural changes, including

recrystallization, even if the diagenetic processes resulted in significant shifts in the oxygen

isotopic composition of skeletal tissues.

Local Baseline Data

It is important to note that the results of this research are specific to the water

compositions of fresh, brackish, and distilled water sourced from Central Florida and should not

be directly applied to the water compositions of other regions and environments, particularly if

they have a much different climate. Similar studies should be conducted in other regions of the

United States to determine if varied water composition and/or climatological factors will result in

differential isotopic exchange and/or diagenetic alteration. Moreover, the water collected from

Tomoka State Park (Volusia County) and Lake Claire (Orange County), as well as the distilled

water collected from the UCF campus (Orange County), provide a ẟ18O range of –1.93‰ to

2.94‰, which is much smaller than the range provided by the GNIP and GNIR databases, which

span almost 10‰ (–4.50‰ to 4.20‰), including water from both rivers and lakes throughout

Florida. This difference may be attributed to the fact that the ranges provided by GNIP and

GNIR, as well as the previously published studies that produced the data included in these

databases (Kendall and Coplen, 2001; Bowen et al., 2008; Coplen et al., 2013; Brooks et al.,

2014), include data produced from water that was sampled from multiple sources (i.e., lakes,

rivers, tap water) throughout the entire state of Florida in multiple seasons. While the available

baseline data, including the data produced from the samples collected during this research, do

include various water sources, seasons, and collection locations, they do not provide a

comprehensive range of ẟ18O data for the state of Florida as there is currently no long-term data

available for this region that can account for oxygen isotopic variation between and within years

or decades. The construction of long-term baseline data for regions within the United States is

critical as more robust baseline data, incorporating long term isotopic variation, will provide a

more accurate estimation of the geographic origin of unidentified individuals in a medicolegal

context.

Implications for Forensic Anthropology

Although this research focuses on the stable isotopic analysis of bone carbonate

following submersion in an aqueous environment, it is important to discuss a few general

observations and suggestions for the application of stable isotope analysis to forensic

anthropology. In the United States, stable isotope analysis is currently infrequently applied to

routine forensic casework, even though previously published cases, from both the United States

and Europe, demonstrate that stable isotopic data can be an extremely valuable form of evidence

when attempting to identify skeletal remains (Rauch et al., 2007; Meier-Augenstein and Fraser,

2008; Kennedy et al., 2011). While routine forensic anthropological methodology, including the

estimation of sex, age, stature, and ancestry, is often sufficient to identify skeletal remains when

there is a clear match to a local missing persons report, there are many circumstances, such as

incomplete or fragmented remains, remains recovered far from their geographic origin, and

remains whose biological profile do not match a local missing persons report, in which stable

isotope analysis should be considered as an immediate addition to routine methodology. While

this method has very clear value for cold cases (i.e., cases that have gone without a lead for a

year or more), forensic investigators should not hesitate to utilize this method much more

frequently and much earlier in their investigations, when certain criteria are met. Figure 17

includes a decision tree illustrating multiple circumstances in which a forensic case might benefit

from the timely utilization of stable isotope analysis.

Figure 17. Decision tree illustrating the timing and circumstances in which stable isotope

analysis should be utilized in forensic cases.

Moreover, the decision to incorporate stable isotopic analysis into a medicolegal

investigation must often be made by medical examiners and/or coroners (depending on the

jurisdiction), rather than forensic anthropologists, as forensic anthropologists are often employed

by medical examiners and/or coroners on a consultation basis (Crowder et al., 2016). Therefore,

if stable isotope analysis is to be more frequently incorporated into medicolegal investigations,

forensic anthropologists must make an effort to educate both the law enforcement personnel,

medical examiners, and coroners with whom they work. Furthermore, most unidentified persons

databases, including the Violent Criminal Apprehension Program (ViCAP) maintained by the

Federal Bureau of Investigation or the Florida Unidentified Decedents Database, do not provide

a section for forensic investigators to upload, share, or access stable isotopic data related to

specific cases. This is a critical oversight as these data must be easily distributed and accessible

to maximize their potential as an evidentiary resource. Additionally, providing forensic

investigators with a section to incorporate stable isotopic data into the case file for a specific

unidentified individual within these databases may encourage more medical examiners, coroners,

and law enforcement personnel to learn about and advocate for the incorporation of this method

into their own investigations.

Research Implications

While multiple statistically significant differences were evident in the δ18O isotopic data,

it is important to note that these differences occur in a maximum range of only 1‰. In fact, most

samples, across all water types and exposure intervals, fall within a range of 0.70‰ or less. This

range is much smaller than that of many regional ranges depicted in previously published ẟ18O

isoscapes of the United States, which often span between 1‰ to 3‰ (Kendall and Coplen, 2001;

Bowen et al., 2008). Moreover, the baseline ẟ18O isotopic data for Central Florida, constructed

using samples from each collection site across multiple seasons, span a range of 4.87‰ (from –

1.93‰ to 2.94‰), much wider than the range produced by diagenetic alteration in the

submerged bone samples. While the ẟ18O isotopic values of samples exposed to fresh, brackish,

and distilled water for six months were statistically significantly different from the control group,

they may not be practically significantly different when utilized in a medicolegal context as such

small shifts in the ẟ18O isotopic composition of bone are unlikely to completely eliminate the

geographic region of origin of an unidentified individual from investigative consideration.

Moreover, no statistically significant differences in ẟ18O composition were noted between

the control group and the samples submerged in fresh, brackish, or distilled water for only one

month, suggesting that skeletonized remains submerged in an aqueous context for one month or

less will produce authentic ẟ18O data, indistinguishable from local ranges, that can be reliably

used by forensic anthropologists in a medicolegal context. Therefore, the results of this study

suggest that skeletonized remains submerged in low saline aqueous environments for six months

or less may still produce ẟ18O isotopic data that are very close to the original stable isotope

compositions and can be cautiously used to narrow the geographic region of origin of

unidentified individuals and provide reliable investigative leads to forensic anthropologists in a

medicolegal context. However, forensic anthropologists should use ẟ18O isotopic data derived

from skeletal remains recovered from aqueous contexts with caution as diagenetic alteration and

subsequent shifts in isotopic composition begin very shortly following submersion. Therefore,

the accurate assessment of PMI and PMSI play a key role in understanding the potential risk of

diagenetic alteration on a case-by-case basis.

While this research was specific to water collected from open water features (i.e., a river

and a lake), the results of this research may extend to a variety of other circumstances in which

skeletal remains may encounter an aqueous environment. Skeletal remains buried in a terrestrial

environment may become completely or partially submerged in groundwater during seasonal

and/or annual climatological events (i.e., hurricane season in Florida) where flooding frequently

occurs. For example, it is known that cemetery remains within a coffin environment may be

subjected to cycles of repeated wet and dry conditions as the groundwater table rises and falls

due to a variety of climatological and/or environmental circumstances (Schultz, 2012). This

repeated exposure to aqueous conditions often causes a variety of taphonomic alterations to the

skeletal remains, including cortical flaking (e.g., flaking of the cortical bone surface which

exposes trabecular bone) on the shafts of long bones, warping of cranial bones, and the growth of

mold and algae which may cause bone staining (Berryman et al., 1991; Schultz, 2012; Pokines et

al., 2016). Previous research suggests that both bone and enamel carbonate may be altered by

exposure to groundwater in various environments (Maurer et al., 2014; Price et al., 2019) and

that cyclical, as well as static, aqueous exposure are sufficient to cause isotopic exchange and

diagenetic alteration in skeletal tissues (Hedges and Millard, 1995; Kendall et al., 2018).

Therefore, while this research is specific to the continuous submersion of bone, it contributes to a

growing body of literature regarding the diagenetic and taphonomic impact of water on skeletal

remains.

Another factor worth considering is the potential impact of various maceration processes,

including boiling bone to remove flesh and cartilage, on the isotopic composition of bone. While

previous studies suggest that boiling bone up to approximately 300°C does not impact the

isotopic composition of bone phosphate (Munro et al., 2007), suggesting that previously

macerated bone may still be reliably used to produce authentic stable isotopic data in a

medicolegal context. However, the use of chemical cleaning agents to aid in defleshing during

the maceration process may have an unknown impact on the stable isotopic composition of bone,

so continued research is necessary to better understand the relationship between various

maceration procedures and potential diagenetic processes. Until more is known about impact of

various maceration procedures on the stable isotopic composition of human skeletal tissues,

forensic investigators should avoid the use of cleaning agents prior to sampling.

Chapter Summary

This chapter provided context regarding factors that may impact the diagenetic alteration

of bone in an aqueous environment — salinity, microbial activity, and submersion interval.

Within the context of these factors, the results of this specific study were related to previously

published literature and important considerations. Additionally, the lack of detection from the

FTIR analysis of CI and C/P was discussed in the context of previous literature, which suggests

that CI and C/P are not always affected by early diagenetic processes. Furthermore, this study

demonstrated that, while statistically significant differences in the oxygen isotopic composition

were noted, these differences represent only a small range of variation and may not negatively

impact the use of oxygen isotopic data to narrow the geographic origin of an unknown individual

in a medicolegal context.

CHAPTER SIX: CONCLUSION

The goal of this research was to evaluate the application of oxygen stable isotope analysis

to the process of human identification in a medicolegal context, specifically, in circumstances

where unidentified human remains are recovered from a Floridian aqueous environment.

Although statistically significant differences were observed between the samples submerged in

fresh, brackish, and distilled water for six months, these differences may not be practically

significant when applied to the case of an unidentified human. While forensic anthropologists

should report and utilize the results of oxygen isotope analysis cautiously when it is known that

the skeletal remains of an unidentified individual were recovered from an aqueous context, the

data may provide reliable information regarding geographic origin if the aqueous context

maintained a low salinity level and the remains were submerged for six months or less.

Limitations

The main limitations of this study are the relatively short submersion intervals, the lack

of quantification of microbial activity, and the fact that this study is a laboratory experiment

removing samples from potentially impactful variables that would be present in forensic cases.

While this study was designed to test the early diagenetic alteration of bone following short

submersion intervals, data regarding longer term submersion (e.g., 9 months, 12 months) would

be beneficial for determining the maximum interval in which bone can be submerged in water

prior to the development of statistically and practically significant differences that would impede

estimations of geographic origin. Although microbial activity is known to be an important factor

in the diagenetic alteration of bone, it was not quantified or fully addressed in this study. While

microbes were certainly present in each type of water, based on the presence of mold, differences

in microbial activity between fresh, brackish, and distilled water are not well understood in this

study. Moreover, this study represents a laboratory experiment in which samples were removed

from many of the variables that would be present and potentially impactful in a forensic case

regarding skeletal remains recovered from an aqueous context. For example, samples in this

study were not subject to seasonal variation in the oxygen isotopic composition of water or

salinity, climatological factors, tide, animal scavenging, or bioerosion caused by sand or other

abrasive materials present in an aqueous environment.

Future Directions

To better answer questions regarding the use of oxygen isotopic data in forensic

investigations following the submersion of skeletal remains in aqueous environments, more

research is necessary. As previously mentioned, continued research including longer term

submersion intervals would be beneficial. Additionally, research regarding the submersion of

fresh (fleshed or clothed) remains would contribute greatly to the overall understanding of the

diagenetic alteration of skeletal remains following both decomposition and submersion in an

aqueous environment. Furthermore, while this research included various types of low salinity

water (i.e., distilled, fresh, and oligohaline brackish water), other types of water, including

mesohaline brackish water, polyhaline brackish water, and marine water should be studied to

more fully understand the role of salinity in the diagenetic processes affecting skeletal remains.

While the effects of marine water on bone have been noted in previous publications (Guareschi

et al., 2021), these observations are often based on skeletal remains recovered from historical

shipwrecks rather than experimental research where variables can be more fully accounted for

and controlled. Studies regarding submersion in non-naturally occurring water types, such as

chlorinated water, may be beneficial as well.

Additionally, the effects of different water compositions and submersion intervals on

tooth enamel should be more thoroughly studied, as well. While enamel is expected to be more

resistant to diagenetic alteration than bone, previous research suggests that enamel carbonate

may still be altered following exposure to aqueous contexts (Zazzo et al., 2004; Price et al.,

2019). Although pig remains are often used as proxies for human remains in forensic and

isotopic research due to physiological and anatomical similarities, research regarding the

authenticity of isotopic data in a medicolegal context would benefit from the inclusion of data

derived from human skeletal remains under known conditions. Furthermore, the construction of

more robust baseline data, constructed from long-term research programs incorporating water

from various sources in Central Florida and elsewhere (i.e., precipitation, rivers, lakes, etc.),

would benefit forensic anthropologists seeking to identify the geographic origin of unidentified

individuals.

Although this study did include multiple limitations, it represents a necessary

contribution to the broader literature regarding the use of oxygen stable isotope analysis by

forensic anthropologists in a medicolegal context. The use of stable isotope analysis to narrow

the geographic region of origin of unidentified individuals is an immensely important

investigative avenue that is available to forensic anthropologists in cases where established

methods are unable to produce an identification. Continued research is necessary to establish the

authenticity and reliability of stable isotopic analysis in other contexts from which skeletal

remains are frequently recovered in forensic investigations.

Forensic Implications for the Diagenetic Alteration of the Stable Isotopic Composition of Pig Ribs Following Exposure to Aqueous Conditions.

Students also viewed