








































Forensic Anthropology Vol. 2, No. 2: 96–101
DOI: 10.5744/fa.2019.1003

© 2019 University of Florida Press

Introduction

In November 1943 a three- day battle ensued between U.S. 
and Japanese forces over control of a strategically located air-
field on a three- square- mile Pacific island. Over 1,100 U.S. 
and over 6,000 Japanese forces and Korean conscript labor-
ers were casualties in the 76- hour battle for Betio Island, Tar-
awa Atoll, Republic of Kiribati (at the time, the Gilbert 
Islands). The logistics of taking control of the island and uti-
lizing its airfield in combination with the reality of large 
casualty numbers and a hot, tropical environment led to hasty 
burials by personnel with limited to no training in proper 
grave documentation (Alexander 1995; Wukovits 2006). 
After the battle an additional 23 U.S. service members lost 
their lives in aircraft crashes or accidents on or around the 
atoll and were subsequently buried on Betio Island.

Postwar cemetery consolidation efforts led to the 
repeated disinterment, relocation, and reinterment of U.S. 
World War II casualties buried overseas, including those 
buried on Betio Island. Poor initial burial documentation, 
multiple relocations, and anthropogenic changes to the island 
resulted in the disjoining, commingling, and loss of identity 
of many Tarawa casualties (Steere & Boardman 1957). By the 
1949– 1951 interment of the Tarawa casualties in their final 
resting place, the National Memorial Cemetery of the Pacific 
(NMCP) in Hawai’i, 92 of the approximately 1,100 lost were 
assigned X- numbers (e.g., X- 001) and buried as unknowns. In 
addition to the 92 unknowns, 219 sets of incomplete remains 
were interred as a group, while an unknown number of indi-
viduals were incompletely or never removed from their 
original burial locations on Betio Island. To date, over 400 
casualties from the Battle of Tarawa remain unaccounted for.

In 2016 the Defense POW/MIA Accounting Agency 
(DPAA) began disinterment of Battle of Tarawa unknowns 
from the NMCP in a large- scale project aimed at resolution 
of commingling and identification of U.S. casualties (for 
details see Taylor et al. 2019). In addition to the NMCP dis-
interments, the Tarawa Project includes approximately 6,000 
elements from 83 field accessions representing osseous 
remains recovered from Betio Island. This assemblage also 
includes remains exposed by storm damage in the 1960s and 
by infrastructure development in the 1970s. Archaeological 

CASE REPORT

The Tarawa Project Part II
Resolved Commingling and Identification of Casualties from  
the Battle of Tarawa

Audrey L. Scotta ●  Heli Maijanenb ● Rebecca J. Taylora* ● Anthony J. Koehla,c ● Willa R. Traska

ABSTRACT: The Battle of Tarawa during World War II resulted in over 7,000 casualties on an approximately three- square- mile 
Pacific island that necessitated hasty burials, often with poor associated documentation. The loss of original burial information combined 
with postwar cemetery consolidation efforts resulted in the commingling, disassociation, and loss of identity of casualty remains. This 
report discusses two cases that are the product of ongoing efforts to recover remains still buried on the Tarawa Atoll and to identify Tarawa 
unknowns disinterred from the National Memorial Cemetery of the Pacific. Both cases involve the reassociation of disinterred elements to 
elements recently excavated from their original wartime burial locations on Betio Island. These cases illustrate the steps taken to reasso-
ciate elements, the multiple lines of evidence necessary to do so, and the eventual outcome of Tarawa Project identification efforts.

Dental and chest radiograph comparisons of the disinterred remains were used to create and narrow short lists of unaccounted- for ser-
vice members. These comparisons were combined with historical documentation and DNA analysis of remains recovered from the Tarawa 
Atoll to suggest association of additional elements. Articulation, refitting of fragments, pair- matching, and osteometric sorting supported the 
association of the disparate elements to single individuals. Ultimately, the dental and chest radiograph comparisons, biological profile, DNA 
testing, and historical data led to the positive identification of both individuals. The complicated nature of the Tarawa commingled assem-
blage emphasizes the importance of a multidisciplinary approach to the resolution and identification of commingled human remains.

KEYWORDS: forensic anthropology, commingled remains, Tarawa, World War II

aDefense POW/MIA Accounting Agency, Joint Base Pearl Harbor–
Hickam, Hawai’i, USA

bOulun Seudun Ammattikorkeakoulu, Archaeology, Oulu, Finland
cSNA International, Alexandria, VA, USA
*Correspondence to: Rebecca J. Taylor, Defense POW/MIA 

Accounting Agency, 590 Moffet Street, Joint Base Pearl Harbor–Hickam, 
JBPHH Hawai’i 96853, USA

E-mail: rjwbones@gmail.com

Received 5 June 2018; Revised 13 August 2018;  
Accepted 14 August 2018



Scott et al. 97

recoveries of skeletal remains have been ongoing by DPAA, 
its predecessor organizations, and nongovernmental orga-
nizations since the mid- 1990s. These recoveries primarily 
involve partial skeletons in two contexts: assemblages of 
remains disturbed but redeposited out of respect by those 
who accidentally discovered them, and elements never recov-
ered from their original burial locations during postwar 
cemetery consolidation efforts. These latter category con-
texts usually involved small elements that may be easily 
missed by those undertaking hurried disinterments with lit-
tle osteological training: vertebrae, ribs, hands, patellae, and 
feet (sometimes still in boots). However, intact individual 
burials have also been recovered.

We will present two cases of skeletal remains disinterred 
from the NMCP with which elements recovered from ongo-
ing field excavations were associated. The Tarawa Project 
employs a multidisciplinary approach to facilitate both the 
association of skeletal and dental elements to a single indi-
vidual, and as a means to identify missing U.S. service mem-
bers following standard protocols for analyzing commingled 
human remains consistent with the prevailing literature 
(Adams & Byrd 2014; Jin et al. 2014). All remains in the Tar-
awa Project undergo a rigorous commingled human remains 
analysis (for details on methods see Taylor et al. 2019).

Case 1

In November 2016 a mostly complete skeleton was disin-
terred from the NMCP as part of the Tarawa Project. The 
skull was fractured and incomplete; most of the thoracic ele-
ments were missing, including the sacrum; the lower left 
arm was absent; and left foot elements were also absent. The 
majority of both hands were absent; only two right carpals 
were recovered. Review of the historical X- number file 
revealed a dental chart of the remains with two names indi-
cated as possible associations. Both named individuals were 
killed in the January 1944 crash of a B- 24 attempting take 
off from Betio Island and subsequently buried on the island 
in the “Main Marine Cemetery” (Cemetery 33).

During analysis, attempts were made to associate the 
remains with either of the two individuals named in the 
X- file dental chart. The right maxilla and right portion of 
the mandible were missing from the remains, thus prevent-
ing a full comparison to antemortem dental charts. DNA 
analysis of remains disinterred from the NMCP has been 
limited; during postwar identification and consolidation 
efforts the remains were treated with formaldehyde powder, 
which inhibits DNA amplification and thus limits tradi-
tional sequencing using the Sanger method. All NMCP Tar-
awa remains must therefore be tested using next- generation 
sequencing (NGS), which only provides a mitochondrial 
DNA (mtDNA) profile. Next- generation testing of the right 

tibia yielded an mtDNA sequence; however, neither candidate 
had an mtDNA family reference sample (FRS) on file. Addi-
tionally, antemortem chest radiographs needed for compari-
son were not available for either candidate. At that time 
no other lines of evidence were available and identification 
efforts were hampered.

In July 2017 the DPAA Laboratory received remains as 
a unilateral turnover from History Flight, Inc, a nongovern-
mental organization in partnership with the DPAA. During 
excavations, skeletal elements had been uncovered in situ 
with an identification tag bearing the name of one of the two 
individuals named in the X- number dental chart. An imme-
diate attempt was made to associate the remains recovered 
from Betio Island and the X- number remains disinterred 
from the NMCP using the methods normally applied to the 
commingled assemblages.

The elements recovered from the Betio Island field 
excavations were first inventoried and found to be the same 
elements missing from the NMCP assemblage (Fig. 1). The 
left radius was sampled for DNA analysis but did not yield a 
profile. The field- recovered left radius, ulna, and foot ele-
ments were compared to the NMCP right antimeres and 
visually pair- matched (Byrd &Adams 2003). The left radius 
and ulna exhibited sound anatomical articulation with the 
disinterred left humerus (Fig. 2). In addition to visual pair- 
matching and articulation, the left radius and ulna were sta-
tistically compared to elements from the right arm using 
linear regression (Adams & Byrd 2014; Byrd 2008; Byrd & 
LeGarde 2014). The results of the analysis could not provide 
statistical evidence for excluding either element, thus sup-
porting the association of the left forearm with the assem-
blage. The left foot similarly articulated with the NMCP left 
tibia and fibula, and visually pair- matched the NMCP right 
foot. The majority of the spinal column was recovered during 
field excavations, including the sacrum, which articulated 
with the innominates from the NMCP disinterment (Fig. 3). 
The four vertebrae that were disinterred with the NMCP 
remains displayed sound anatomical articulations among 
themselves and with field- recovered vertebrae. The right and 
left hands were recovered during field excavations, except 
the right capitate and trapezium, which were included in the 
NMCP assemblage; all hand elements articulated. Both 
patellae were recovered and articulated with the NMCP fem-
ora, and the sternal body articulated with the disinterred 
manubrium. Three left ribs and four right ribs were present 
in the disinterred NMCP assemblage, were morphologically 
consistent with the nine left and eight right ribs recovered 
from Betio Island, and articulated with the recovered verte-
brae. All possible articulations between the NMCP remains 
and the elements recovered from field excavations were 
tested with positive results.

In addition to articulations among elements, several 
fractured and broken elements had been separated during the 



FIG. 1—Homunculus depicting skeletal remains associated with Case 1. 
Elements depicted in gray were disinterred from the NMCP. Elements 
depicted in black were recovered from archaeological excavations on 
Betio Island, Tarawa Atoll, Republic of Kiribati. Elements depicted in 
white are absent.

FIG. 4—Case 1, left scapula, posterior view. The lighter portions were 
excavated from Betio Island; the darker portions were disinterred from 
the NMCP. (Photograph by W. Lavin)

FIG. 3—Case 1, sacro- iliac articulation, superior view. The sacrum 
was recovered from Betio Island excavations; the os coxae were 
disinterred from the NMCP. (Photograph by W. Lavin)

FIG. 2—Case 1, left elbow, posterior view. The humerus (right) was 
disinterred from the NMCP; the radius and ulna (left) were found in situ 
with an identification tag in the individual’s original burial location. 
(Photograph by A. Scott)

98 Tarawa Project II

postwar cemetery consolidation efforts, yet refit perfectly at 
the margins. The right scapula, left scapula (Fig. 4), and right 
fourth rib (Fig. 5) all exhibited postmortem breakage with 
refitting portions recovered from the two proveniences. The 
mandible exhibited a midline perimortem fracture consistent 

with the perimortem fragmentation seen in the cranium. The 
right maxillary alveolus was separated from the body of the 
maxilla and exhibited a perimortem fracture at the location 
of the first molar. The right maxillary alveolus anterior to this 
location was in articulation with the left maxilla. The right 



FIG. 5—Case 1, right fourth rib. The lighter sternal end was recovered 
during field excavations; the darker portion was part of the NMCP 
disinterment assemblage. (Photograph by W. Lavin)

FIG. 6—Case 1, mandible, anterior view. Note two distinct taphonomic 
signatures on each of the mandible halves (the right portion from Betio 
Island and the left portion from the NMCP) and the refitting fracture 
margin. (Photograph by A. Scott)

Scott et al. 99

portions of maxilla and mandible were recovered with the 
remains excavated on Betio Island and perfectly refit with 
the left portions disinterred from the NMCP (Fig. 6). The 
refitting of breakage and fracture margins provided strong 
support to the association of the field- recovered elements 
with the NMCP remains.

With the refitting of the maxilla and mandible portions 
into complete elements, a full dental analysis became pos-
sible. Comparison of the dental remains to the antemortem 
records of the individual named on the identification tag 
yielded a possible match. At the conclusion of dental analy-
sis, the right mandible portion recovered during field exca-
vation was sampled for DNA analysis. Unlike the remains 
interred in the NMCP, the remains excavated from their 
original wartime burial locations regularly yield mtDNA 
sequences through traditional Sanger method analysis and 
can be used for nuclear DNA testing. Although no mtDNA 

FRS data were available for the suspected candidate, the 
Armed Forces DNA Identification Laboratory (AFDIL) had 
nuclear FRS data on file for comparison. The mandible sample 
was therefore tested for ySTR and autosomal DNA. Accord-
ing to AFDIL analytic results, the resulting sequences are 
consistent with the FRS of the service member named on the 
identification tag.

Although DNA analysis did not contribute to the asso-
ciation of elements in this case, it did contribute to identifi-
cation. Resolution of the fragmentation was achieved through 
methods normally utilized in commingled assemblages; pair- 
matching, articulation, statistical osteometric sorting, and 
refitting fracture and breakage margins. The re- associated 
dental remains provided the basis for a possible name asso-
ciation when compared to antemortem dental records. The 
skeletal elements recovered from the recent field excavations 
provided the DNA sample necessary for comparison to a FRS 
and led to the positive identification.

Case 2

Similar to the first case, Case 2 was disinterred from the 
NMCP in the autumn of 2016. The cranium was fragmented 
and incomplete, several vertebrae were missing, all elements 
of the left arm were absent, and the remains lacked any ele-
ments from the hands or feet (Fig. 7). Initial anthropological 
analysis suggested a minimum of two individuals were rep-
resented, as the articulation of the right ulna with the right 
radius and humerus was questionable. The occipital, right 
ulna, and right tibia were sampled for NGS analysis and 
yielded mtDNA to confirm the suspected commingling and 
aid association of elements.

Comparison of the partial dentition present in the NMCP 
remains to all Battle of Tarawa casualties produced a proba-
ble match to a single candidate. Although several cervical 
vertebrae were absent from the NMCP remains, a chest 
radiograph comparison (CXR) to the dental candidate was 
possible. This comparison found multiple points of concor-
dance between the radiograph of the postmortem clavicles 
and vertebrae and the antemortem radiograph of the candi-
date (Fig. 8; Stephan et al. 2011).

The FRS associated with the service member from the 
CXR and dental association matched with the mtDNA 
sequence shared by several skeletal elements recovered from 
Betio Island. In 2013, several large assemblages of fragmented 
and commingled remains excavated by History Flight, Inc. 
were unilaterally turned over to the DPAA Laboratory in 
multiple batches. A left ulna included in one of these assem-
blages yielded an mtDNA sequence matching the FRS of the 
single dental and CXR candidate. The ulna was recovered 
with an articulating left radius. Additionally, four metacar-
pals, a left patella, a first right metatarsal, and a first left 



FIG. 7—Homunculus depicting skeletal remains associated with Case 2. 
Elements depicted in light gray were disinterred from the NMCP. 
Elements depicted in dark gray and black were recovered during two 
separate archaeological excavations on Betio Island and subsequently 
unilaterally turned over to the DPAA Laboratory. Elements depicted in 
white are absent.

FIG. 8—Case 2, comparison of antemortem chest radiographs of the 
probable dental match (top) with postmortem radiographs of the left 
and right clavicles (bottom) white arrows indicate locations of 
morphological concordance. The images provided are for illustrative 
purposes only and do not depict the original size and resolution of the 
images used in the CXR, as these are best viewed onscreen using the 
electronic file. The difficulty of CXR is illustrated here by the poor 
quality of the available antemortem radiograph.

100 Tarawa Project II

metatarsal all yielded the same mtDNA sequence. These ele-
ments originated from two separate field excavations from 
the same general provenience of a trench burial in the 
Cemetery 33 area. The NMCP unknown was originally 
recovered from Cemetery 33 in the 1940s.

Anthropological analysis confirmed the association of 
the mtDNA- sequenced elements recovered from the field in 
2013 with the partial skeleton disinterred from the NMCP in 
2016. The left radius that was recovered in association with 
the left ulna visually and statistically pair- matched the NMCP 
right radius. The right ulna originally disinterred with the 
remains eventually yielded an mtDNA sequence not match-
ing that of the other elements, confirming the suspicion of 
commingling within the NMCP casket. The two first meta-
tarsals were recovered in situ within boots in articulation 
with the entirety of the elements of their respective feet. The 
tarsals of each foot articulated with the distal tibiae of the 
NMCP remains. With the reassociation of the field recoveries 
to the NMCP remains, the DNA analysis required for the dis-
interred remains was reduced to only the cranial fragments, 
which could not be articulated with the postcranial elements 
due to missing portions. The occipital sample yielded an 
mtDNA sequence matching the sampled field elements.

Despite the original association having been lost 
through repeated disinterments and reinterments, DNA and 
anthropological analysis led to the reassociation of elements 
representing a single individual. The dental evidence, chest 
radiograph comparison, and mtDNA data taken in totality 
provided undeniable evidence on which to base an identifica-
tion of the reassociated remains.

Conclusion

Since 2016, the Tarawa Project has assisted in the identifica-
tion of 41 previously unaccounted- for service members, 



Scott et al. 101

including 26 unknowns disinterred from the NMCP. Eight 
of those cases required the reassociation of skeletal elements 
recovered from field excavations with remains disin-
terred from the NMCP. While DNA is often relied upon for 
association and identification of commingled remains, its 
use is not always reliable for cases of complex open com-
mingled populations, in which a definite list of missing indi-
viduals is not available (Puerto et al. 2014). The ongoing 
anthropogenic land alterations, lack of original deposition 
documentation, and occurrence of primary, secondary, and 
tertiary burials further limit the use of DNA for resolution 
and identification. Many of the Tarawa cases rely on the abil-
ity to associate elements recovered from Betio Island to 
NMCP disinterred remains through a multifaceted anthropo-
logical approach that includes articulation of elements, refit-
ting of fragments, osteometric sorting, and chest radiograph 
comparisons. DNA analysis and dental comparison provide 
the basis either for development of preliminary short lists of 
possible candidates or for identity verification. The two cases 
discussed here illustrate the complex, multidisciplinary pro-
cess necessary for resolving commingling and achieving 
identifications in the Tarawa Project. Due to the nature of the 
project assemblage, we anticipate more identifications will 
require the association of elements excavated from their 
original burial locations to NMCP disinterments. The Tar-
awa Project, and other ongoing commingled human remains 
projects, is one facet of the DPAA efforts to achieve the full-
est accounting possible of lost U.S. service members.

Acknowledgments

We would like to acknowledge all of the service members 
who have made the ultimate sacrifice; DPAA’s strategic part-
ner, History Flight, Inc., for their recovery efforts on Betio 
Island; and our project team and colleagues for feedback on 
this paper. Special thanks to Ms. Wenona Lavin for element 
photography.

The opinions presented here are the private views of the 
authors and do not necessarily reflect those of DPAA, the 
Department of Defense, or the United States government.

References

Adams BJ, Byrd JE. Commingled Human Remains: Methods in 
Recovery, Analysis, and Identification. San Diego: Academic 
Press; 2014.

Alexander JH. Utmost Savagery: The Three Days of Tarawa. 
Annapolis, MD: Naval Institute Press; 1995.

Byrd JE. Models and methods for osteometric sorting. In: Adams 
BJ, Byrd JE, eds. Recovery, Analysis, and Identification of 
Commingled Human Remains. Totowa, NJ: Humana Press; 
2008:199– 220.

Byrd JE, Adams BJ. Osteometric sorting of commingled human 
remains. Journal of Forensic Sciences 2003;48:717– 724.

Byrd JE, LeGarde C. Osteometric sorting. In: Adams BJ, Byrd JE, 
eds. Commingled Human Remains: Methods in Recovery, 
Analysis, and Identification. San Diego: Academic Press; 
2014:167– 192.

Jin J, Burch AL, LeGarde C, Okrutny E. The Korea 208: A large- 
scale commingling case of American remains from the Korean 
War. In: Adams BJ, Byrd JE, eds. Commingled Human 
Remains: Methods in Recovery, Analysis, and Identification. 
San Diego: Academic Press; 2014:407– 424.

Puerto MS, Egana S, Doretti M, Vullo CM. A multidisciplinary 
approach to commingled remains analysis:  Anthropology, 
genetics, and background information. In: Adams BJ, Byrd 
JE, eds. Commingled Human Remains: Methods in Recovery, 
Analysis, and Identification. San Diego: Academic Press; 2014: 
307– 335.

Steere E, Boardman TM. Final Disposition of World War II Dead, 
1945– 51. Quartermaster Corps, QMC Historical Studies, 
Series II, No. 4, Washington DC: U.S. Army; 1957.

Stephan CN, Winburn AP, Christensen AF, Tyrrell AJ. Skeletal 
identification by radiographic comparison: Blind tests of a 
morphoscopic method using antemortem chest radiographs. 
Journal of Forensic Sciences 2011;56:320– 322.

Taylor RJ, Scott AL, Koehl AJ, Trask WR, Maijanen H. The Tarawa 
Project Part I: A multidisciplinary approach to resolve com-
mingled human remains from the Battle of Tarawa. Forensic 
Anthropology 2019;2(2):87–95.

Wukovits J. One Square Mile of Hell: The Battle for Tarawa. New 
York: New American Library; 2006.


