EXCAVATION AND STUDY OF COMMINGLED HUMAN SKELETAL REMAINS

EXCAVATION AND STUDY OF COMMINGLED HUMAN SKELETAL REMAINS The Cyprus Institute Science and Technology in Archaeology and Culture Research Center (STARC) Guide No. 2 Authors: Efthymia Nikita, Anna Karligkioti & Hannah Lee Reviewers: Kathryn Marklein, University of Louisville ∞ Ioanna Moutafi, University of Cambridge ∞ Christina Papageorgopoulou, Democritus University of Thrace ∞ Niki Papakonstantinou, Aristotle University of Thessaloniki ∞ Paraskevi Tritsaroli, University of Groningen Version 1.0 Nicosia, 2019 2 EXCAVATION AND STUDY OF COMMINGLED HUMAN SKELETAL REMAINS The Cyprus Institute Science and Technology in Archaeology and Culture Research Center (STARC) Guide No. 2 Authors: Efthymia Nikita, Anna Karligkioti & Hannah Lee Reviewers: Kathryn Marklein, University of Louisville ∞ Ioanna Moutafi, University of Cambridge ∞ Christina Papageorgopoulou, Democritus University of Thrace ∞ Niki Papakonstantinou, Aristotle University of Thessaloniki ∞ Paraskevi Tritsaroli, University of Groningen Version 1.0 Nicosia, 2019 The compilation of the manuscript was made possible through generous funding from the European Union Horizon 2020 (Promised, Grant Agreement No 811068) and the Research and Innovation Foundation (People in Motion, EXCELLENCE/1216/0023) This work is distributed under a creative commons licence (CC BY-NC 2.0). Nicosia 2019 (Version 1.1, layout edited 2021) ISBN 978-9963-2858-5-3 CONTENTS 1 Preface 3 Introduction 5 Excavation 5 Laboratory analysis 6 Bone/tooth inventory 6 Sorting procedures 18 Estimation of the number of individuals 20 Sex assessment 20 Age-at-death estimation 21 Pathological lesions 21 Activity markers 21 Nonmetric traits 21 Morphoscopic traits 21 Metrics 21 Stature estimation 21 Post-mortem bone alteration 22 Additional Resources 22 References 27 Recording Sheets PREFACE This document is the second in a series of guides aimed at promoting best practice in different aspects of archaeological science, produced principally by members of the Science and Technology in Archaeology and Culture Research Center (STARC) of The Cyprus Institute. The current document was largely developed in the context of two projects: People in Motion and Promised. The implementation of People in Motion involved the laboratory study of a large commingled skeletal assemblage from Byzantine Amathus, Cyprus, which came to light in the context of excavations led by the Cypriot Department of Antiquities. Osteological work on this assemblage was co-funded by the European Regional Development Fund and the Republic of Cyprus through the Research and Innovation Foundation (Project: EXCELLENCE/1216/0023). In addition, Promised aims at promoting archaeological sciences in the Eastern Mediterranean, with funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 811068. Commingled assemblages pose special challenges in their study, nonetheless such a study can reveal key information on the osteobiography of those comprising the assemblage and the funerary practices. In addition, since commingling is both a natural and cultural process, it should be viewed not strictly as an impediment to study (though admittedly methodology has to be adapted and ‘traditional’ bioarchaeological conclusions are often limited), but as a kind of ‘life history’ of a skeletal assemblage. In line with the above, the aim of this guide is to cover various aspects of the study of a commingled skeletal assemblage. It should be seen as a supplement to the ‘Basic guidelines for the excavation and study of human skeletal remains; STARC Guide no. 1 ’, which outlines the key general methods for human skeletal excavation and analysis. As the first guide, it focuses on the excavation and study of bioarchaeological assemblages, rather than forensic anthropological material, though many of the practices described are shared between these disciplines. Readers interested in the scientific investigation of multiple burials from forensic contexts are advised to consult the volume by Cox et al. (2008). It cannot be overemphasized that each commingled skeletal assemblage will pose different challenges and any approach to field recovery/excavation and laboratory procedures will have to be adapted to these. Therefore, the current guide is meant to serve only as a general outline, and the described field and lab-based methods should be modified depending on individual circumstances, such as the degree of commingling, sample size, preservation of the material, research questions and other parameters. A number of excellent edited volumes have been published in the past years, compiling diverse case studies on the retrieval and examination of commingled skeletal remains in archaeological and forensic contexts (Adams and Byrd 2008, 2014; Osterholtz et al. 2014a; Osterholtz 2016). A lot of the information presented here has been drawn from these resources, as well as from other publications and the authors’ personal experience. References are given throughout the document but our aim is by no means to provide an exhaustive account of the literature. This document is an open resource and it is anticipated to be updated at regular intervals. We would greatly appreciate any feedback and recommendations for future improvement. Efthymia Nikita Anna Karligkioti Hannah Lee * For suggestions about how to improve this guide, please contact Efthymia Nikita: e.nikita@cyi.ac.cy 1 INTRODUCTION The term “commingling” refers to the intermixing of the remains from more than one individual. This phenomenon is encountered both in archaeological and forensic contexts and poses particular challenges in the retrieval and study of such assemblages. The commingling of elements can occur at different stages during and after the deposition of the bodies and as a result of different factors. The factors that may cause commingling include funerary practices that involve the manipulation of the deceased at various postmortem stages, intentional interference with the bodies to eliminate incriminating evidence in forensic cases, scavenger activity, underground water and other taphonomic processes (Osterholtz et al. 2014b; Ubelaker 2014). Even without the intervention of any extrinsic factor, commingling will occur as a result of the decomposition processes when multiple individuals are buried together. As the soft tissues disintegrate, skeletal elements will tumble off the pile of bodies, and small elements (e.g. phalanges) will end up at the bottom of the body pile. The way the bodies were deposited, how many individuals were placed in the same grave, the condition of the remains prior to deposition (e.g. fleshed or skeletonised), the amount of free space between the bodies, and other parameters will also affect the degree of commingling (Duday 1985, 2005, 2009; Roksandic 2002). Commingled assemblages can be classified in different categories. A common division is between small-scale and largescale commingling. Small-scale commingling is characterised by a small number of disarticulated elements and/or an overall small number of individuals in the assemblage. Note that even in cases of large sample sizes, the degree of commingling may be classified as small-scale if the skeletal elements are still mostly articulated upon recovery. Large-scale commingling is characterised by large numbers of skeletons with intermixed elements and often very poor preservation (Adams 2014; Byrd and Adams 2003; Mundorff et al. 2014). An alternative classification is given by Osterholtz et al. (2014b) and is based principally on the duration of use of the multiple burial site: 1. assemblages created through long term usage, and 2. assemblages created through episodic usage. The main characteristics of each type of assemblage are given in Table 1 (adaptation of Figure 1 in Osterholtz et al. 2014b). Osterholtz et al. (2014b: 5) also identify a third type of commingling, that is, lab commingling “that can occur at any stage of analysis or curation”. This type will not be discussed here but the guidelines provided in the following sections to address the other two types of commingling largely apply to cases of lab commingling as well. The high variability in the characteristics of commingled skeletal assemblages suggests that any strategy for retrieval and study has to be case-specific. The current document aims at covering various approaches, applicable in different types of commingled assemblages, but at the same time it is to be approached as a guide that practitioners will need to adjust to the specific needs of their study. In this direction, it should be read in conjunction with the ‘Basic guidelines for the excavation and study of human skeletal remains; STARC Guide no. 1 ’, which outlines the general principles of human skeletal excavation and analysis. Table 1. Types of commingled assemblages | Adapted from Osterholtz et al. 2014b Long term usage Episodic usage Formation process Prolonged tomb use with multiple reopenings to deposit new bodies A single opening of the tomb to deposit multiple bodies Characteristics More pronounced commingling and fragmentation of the remains Less extensive commingling and better preservation of the remains Element representation Variable depending on the nature of the assemblage: • Primary long term (bodies decompose at the burial site): representation of smaller elements consistent with the number of individuals • Secondary long term (bodies decomposed elsewhere and then were collected and deposited at the burial site): under-representation of smaller elements Generally consistent with the demography of the assemblage Demography Reflecting the mortuary program Reflecting the factor that led to the death and burial of multiple individuals (e.g., plague → mostly very young and very old individuals; warfare → mostly young males) 3 4 EXCAVATION The stratigraphy of mortuary contexts that include the remains of multiple individuals is usually complex, especially in cases where the same location had been used for an extensive period of time. This section provides basic guidelines to the excavation of commingled skeletal remains, which should be adapted on a site by site basis pending the character of each archaeological assemblage and the available resources. For the steps involved in the excavation of commingled remains, see the guidelines provided in ‘Basic guidelines for the excavation and study of human skeletal remains; STARC Guide no. 1’. What should be stressed here is that grid construction is especially useful when excavating disarticulated commingled remains, where there is a great degree of fragmentation and dispersal of skeletal elements, as it allows for more accurate mapping of individual loose bones and subsequently the examination of patterns in the dispersal of the remains and their association. As stated in the first volume of this series, stratigraphic excavation is preferred. When excavating commingled remains, it should be remembered that remains found within the same stratigraphic level may be associated, but this is less likely for remains from different stratigraphic deposits. Therefore, when excavating disarticulated remains, one should first look for possible matches within the same deposit (Tuller and Hofmeister 2014). Recognition of partial articulation in the field is necessary as this information may later provide insights to past burial customs. Burial documentation by means of sketching, photography, and note-taking should follow the guidelines given in STARC Guide no. 1. Total stations or different GPS systems are increasingly used to record burial sites and scattered remains, or even the exact location of individual skeletal elements within the commingled assemblage, since they provide the opportunity to accurately record positional data (see examples in Christensen et al. 2014; Dupras et al. 2012). Therefore, these methods may allow the analysis of the spatial distribution of the remains and reveal patterns of post-depositional processing, as well as assist the reassociation of loose elements based on their relative distance within the assemblage (Naji et al. 2014; Tuller and Hofmeister 2014). An interesting combination of GIScaptured point data on individual skeletal elements and osteobiographic data (age, pathology) was performed by Calleja (2016), who used this approach in visualising the distribution of pathological bones at Bronze Age Tell Abraq. Considerations when excavating complete skeletons in multiple burials Attention should be given when digging complete skeletons as it is common that the first elements to be found at a higher elevation than the rest are the skull and the pubic symphysis. These anatomical areas are of great importance in age-at-death and sex estimation, thus care should be taken not to damage them. LABORATORY ANALYSIS Each assemblage reaching the lab will have specific properties depending on the depositional environment, the type of commingling (see Table 1) and the extent to which articulated elements and elements belonging to the same skeleton more generally have been identified during excavation. Figure 1 summarises the general procedure that may be followed when studying a commingled assemblage, but this should be adapted according to the nature of each assemblage under examination. Note that many of these steps are the same as for the general study of human skeletal remains and the reader is advised to consult ‘Basic guidelines for the excavation and study of human skeletal remains; STARC Guide no. 1’. In this section, we will focus exclusively on the methods that are specifically designed for commingled remains. 5 Figure 1. General procedure in the study of commingled assemblages BONE/TOOTH INVENTORY The bone and tooth inventory in cases of commingled remains will include each skeletal/dental element or bone/ tooth fragment as a separate entry. However, upon sorting the remains, elements belonging to the same individual must be noted as such in the database. As stated in STARC Guide no. 1, bone fragments too small to identify should be divided in broad categories (e.g. cortical bone/trabecular bone, cranial bone/post-cranial bone, axial skeleton/appendicular skeleton), sorted by size class based on maximum dimension, counted and weighted (Outram 2001). Refitting/conjoining studies can provide important insights to the original position and subsequent relocation and manipulation of the bodies inside the grave (Emberling et al. 2002; Moutafi 2016; Papathanasiou 2009). However, refitting is particularly time consuming, especially in large and highly fragmented assemblages (Knüsel and Robb 2016). When two or more bone fragments are conjoined, they should be input as a single entry in the database with an accompanying note. SORTING PROCEDURES During sorting, the bones belonging to each individual are identified (individuation process). Depending on the nature of the commingled assemblage (state of bone preservation, sample size, degree of commingling), this step may take place before the inventory and, subsequently, the remains should be inventoried per skeleton rather than per element. The first step of the sorting process involves the conjoining of fragmentary remains to the greatest extent possible. Bones should then be sorted by element type, side, and size using the most appropriate among the available techniques: visual pairmatching, articulation, process of elimination, osteometric comparison, and taphonomy. Elements that were articulated at the time of recovery should be maintained as a unit. Components of the biological profile (e.g., age-at-death, sex, and stature) may also be useful in the sorting process. Sorting 6 procedures should be used in conjunction with each other, as well as with contextual scene information (Adams 2014). After all macroscopic techniques have been used, chemical analysis and DNA profile data may be employed. Visual pair-matching Visual pair-matching refers to the association of left–right elements based on morphological similarities (Adams and Byrd 2006). Overall bone size and robusticity are the primary factors examined, while nonmetric traits (e.g. third trochanter) or entheseal changes can offer additional help in identifying paired elements. If the elements under study preserve age or sex markers (e.g. unfused epiphyses, pubic symphysis etc.), these are also important to take into consideration in pairing. Articulation The size and shape of adjoining bones is correlated as they form a functional joint (Buikstra et al. 1984). However, the strength of association varies depending on the elements considered (Adams and Byrd 2006), thus not all joints will be equally useful in sorting (Puerto et al. 2014 and Figure 2). HIGH • • • • • • Cranium – mandible Vertebrae 5th lumbar vertebra – sacrum Humerus – ulna Innominate – sacrum Tibia – talus • • • • • Ulna – radius Metatarsals 2-5 Metacarpals 2-5 Tarsals Tarsals – metatarsals MODERATE • • • • Cranium – atlas Tibia – fibula Femur – tibia Innominate – femur • • • • Patella – femur Navicular (scaphoid) – radius Carpals* Carpals – metacarpals LOW • • • • • Metatarsal 1 – other metatarsals Metacarpal 1 – other metacarpals Ribs – thoracic vertebrae Manubrium – clavicle Humerus – scapula *The articular surface of the pisiform is too small to articulate with confidence Figure 2. Articulations with associated confidence placed in each fit | Drawn from Table 3 in Adams and Byrd 2006 Reliability of articulations (from Puerto et al. 2014) • High reliability (≥ 90% of correct classification): vertebrae, sacrum-pelvis. • Moderate reliability (60%–90% of correct classification): temporomandibular joint, atlas–occipital condyles, humerus– ulna, radius–ulna, tibia–fibula, tibia–talus. • Low reliability (≤ 60% of correct classification): clavicle–scapula, clavicle–manubrium, sternum–ribs, ribs–vertebrae, humerus–scapula, radius–carpals, pelvis–femur, femur– tibia. Process of elimination Pair-matching The process of elimination is mostly applicable in cases of small-scale commingling. It is advisable to use this method after articulation and pair-matching have been employed (Adams and Byrd 2006). Method 1 Size (osteometric sorting) Osteometric sorting tests the null hypothesis that the two bones under examination are similar enough in size and shape to have originated from the same individual (Byrd 2008). It may be used for pairing bilateral elements, matching articulated elements, or identifying bones belonging to the same individual based on their relative size (Byrd and LeGarde 2014). Nonetheless, this method is only applicable to well-preserved skeletal elements and it will be of limited, if any, use to highly fragmented material. Nikita and Lahr (2011) proposed the easiest so far method in pairing bilateral elements based on bone dimensions. The authors provide an Excel macro where different measurements are input and the program considers all possible pairings. If the pairwise differences are below the acceptable user-defined threshold, the right and the left element under examination are given as a potential pair. This method is a way to “prescreen” possible pairs and minimises the time required to visually match the bones. Among the advantages of this method is that it is simple in its implementation, the user can employ any measurement he/she wishes, and additionally to bone measurements, the calculations can incorporate the degree of expression of entheseal changes or osteoarthritis. 7 Method 2 Most other methods use a reference sample of known paired bones to obtain information on the similarity between homologous measurements. Such methods measure the difference between two potentially paired bones, and then compare this difference to the reference sample to determine whether it is equal to or greater than that expected. In this direction, Byrd and LeGarde (2014) proposed to sum the differences in the values between left and right pair measurements in order to produce the so-called D-value. Subsequently, the D-value is compared to the summed differences of the reference sample for the same measurements. The reference sample serves as a representative source of the left–right differences seen in a population, and in the case of Byrd and LeGarde (2014), it included material from several documented skeletal collections housed in North American institutions. The D-value has zero subtracted, and is then divided by the standard deviation of the reference D-values to produce a t-statistic. Basically, the D-value is being compared against zero to determine whether the difference in size between the elements under examination is significantly different compared to the difference seen in the reference sample. The t-statistic is then compared to a two-tailed t-distribution to produce a p-value where the degrees of freedom are equal to the reference sample size minus 1. Any p-value that is less than or equal to 0.10 is considered significant, thus the elements are too different in size to have originated from the same individual. A p-value greater than 0.10 does not confirm the elements originated from a single individual, but indicates that the elements may belong to a single individual. In order to apply the Byrd and LeGarde (2014) method, you need a reference sample from which to estimate the standard deviation of the D-values to produce a t-statistic. If this is not available, you can use the values provided by the authors, so long as you adopt the same measurements as them in your comparisons (Table 2). See the original publication for a more detailed description of the measurements. Table 2. Measurements and reference population data for comparison of paired elements | Adapted from Table 8-9 in Byrd and LeGarde 2014 Skeletal Element Measurements N Standard deviation Humerus Maximum length 113 5.28 Minimum diameter of diaphysis (in any direction perpendicular to shaft) 73 0.72 Maximum length 100 3.56 52 1.34 93 3.60 45 1.62 Epicondylar breadth Capitulum-trochlea breadth Radius Midshaft sagittal diameter Midshaft transverse diameter Maximum shaft diameter at the radial tuberosity Maximum shaft diameter distal to the radial tuberosity Minimum shaft diameter distal to the radial tuberosity Ulna Maximum length Transverse diameter at point of maximum expression of interosseous crest Dorso-volar diameter perpendicular to transverse diameter at the same position along the diaphysis Transverse diameter at point of maximum expression of interosseous crest Dorso-volar diameter perpendicular to transverse diameter at the same position along the diaphysis Minimum diameter of diaphysis along the portion of the bone that includes the interosseous crest 8 Skeletal Element Measurements N Standard deviation Femur Maximum length 67 3.99 123 1.75 87 3.68 44 2.62 71 2.99 Epicondylar breadth Maximum head diameter Anterior-posterior subtrochlear diameter Transverse subtrochlear diameter Anterior-posterior subtrochlear diameter Transverse subtrochlear diameter Tibia Maximum length Maximum breadth of proximal epiphysis Maximum breadth of distal epiphysis Maximum diameter at nutrient foramen Transverse diameter at nutrient foramen Minimum anterior-posterior diameter of shaft Fibula Maximum length Maximum midshaft diameter A step-by-step example: Measurement Left-side element Right-side element Maximum length 492 480 Epicondylar breadth 85 77 Maximum head diameter 50 51 Anterior-posterior subtrochlear diameter 25 28 Transverse subtrochlear diameter 26 31 D 2.757 Reference sample standard deviation (from Table 2) 3.99 t (calculated as |D-0|/3.99) 0.691 p (from t-distribution, d.f. = 66, 2 tails) 0.492* *You can easily calculate this value in Excel using the command =TDIST(t, d.f.,2), whereby t is the t value you have already estimated (in our case 0.691) and d.f., the degrees of freedom, are equal to the sample size minus 1 (in our case 67-1 = 66). In our example, we have =TDIST(0.691,66,2) = 0.492 Vickers et al. (2015) conducted a validation study of the method proposed in Byrd and LeGarde (2014). Although they found that the number of potential pairs requiring visual matching was considerably reduced, they highlighted three serious shortcomings of this method: (i) violation of the normality assumption for use of a t-score approach, (ii) lack of accountancy of bilateral asymmetry, and (iii) high rate of false rejections. Lynch et al. (2018) more recently proposed two variants of this model and found the two new models (Models B and C in Table 3) to outperform the original one (Model A in Table 3). Table 3 presents the calculations involved in each model. 9 Method 3 An alternative method for osteometric pairmatching was proposed by Thomas et al. (2013). This method compares a calculated M-statistic, which expresses the metric similarity of bilateral elements, to a reference sample. The M-statistic is calculated using the following equation, where L and R represent left and right bilateral measurements, respectively: Table 3. Different models for osteometric pairing (from Figure 1 in Lynch et al. 2018) Model A Model B Model C Key: a and b indicate the left and right-side measurements, respectively; i is the index of the number of measurements; N is the reference sample size; ref and com refer to the reference and comparison samples, respectively of M from the reference database (given in Table 4). If the value of M is greater than that for the chosen percentile, the null hypothesis can be rejected, suggesting that the bones under examination are unlikely to have originated from the same individual. Rather than using the sum of multiple measurements as seen in the statistical models of Byrd and LeGarde (2014), this method examines each measure independently. Therefore, it is useful for evaluating highly fragmented commingled assemblages; however, it is also considered statistically weak. Pairs of skeletal elements are considered likely homologs if the M-statistic approaches zero, and this is statistically evaluated by means of confidence intervals calculated from reference samples. The reference samples come from several skeletal collections and databases, largely the same as the material used in the Byrd and LeGarde (2014) method. The null hypothesis that two homologs are from the same individual can be tested by calculating the value of M for each measurement for the bones in question and comparing it to the 90th and 95th percentiles or the maximum value Table 4. Maximum values and the 90th and 95th percentiles for the M statistic (drawn from Table 2 in Thomas et al. 2013) Skeletal Element Measurement* N 90th 95th Max M Clavicle Max Length 104 0.049 0.056 0.081 A-P Diameter Midshaft 93 0.182 0.154 0.222 M-L Diameter Midshaft 92 0.180 0.200 0.353 Height 102 0.031 0.039 0.077 Breadth 115 0.032 0.040 0.064 Max Height Glenoid Fossa 67 0.050 0.061 0.092 Max Breadth Glenoid Fossa 68 0.054 0.058 0.082 Max Length 152 0.021 0.026 0.071 Epicondylar Breadth 135 0.034 0.047 0.079 Capitulum-Trochlea Breadth 57 0.039 0.052 0.067 Head Diameter 128 0.040 0.043 0.089 A-P Head Breadth 46 0.034 0.038 0.043 Max Diameter Midshaft 118 0.091 0.099 0.160 Min Diameter Midshaft 138 0.074 0.101 0.162 Min Diameter Diaphysis 75 0.066 0.079 0.101 Scapula Humerus 10 Skeletal Element Measurement* N 90th 95th Max M Radius Length 134 0.019 0.029 0.064 A-P Diameter Midshaft 104 0.089 0.098 0.154 M-L Diameter Midshaft 104 0.097 0.117 0.143 Max Diameter at Radial Tuberosity 57 0.057 0.079 0.096 Max Diameter of Diaphysis 56 0.075 0.106 0.167 72 0.068 0.073 0.091 Length 129 0.022 0.026 0.041 Dorso-Volar Diameter 116 0.129 0.186 0.533 Transverse Diameter 132 0.116 0.160 0.375 Physiological Length 85 0.022 0.027 0.039 Min Diameter Osseous Crest 48 0.071 0.087 0.102 Min Diameter 50 0.062 0.081 0.095 Height 133 0.019 0.023 0.051 Iliac Breadth 132 0.025 0.030 0.077 Max Thickness at Sciatic Notch 68 0.092 0.118 0.158 Max Diameter of Acetabulum 46 0.031 0.035 0.046 Max Length 109 0.014 0.015 0.020 Epicondylar Length 99 0.014 0.015 0.021 Epicondylar Breadth 108 0.022 0.025 0.063 Head Diameter 122 0.026 0.037 0.048 A-P Subtrochlear Diameter 140 0.071 0.095 0.129 Transverse Subtrochlear Diameter 126 0.066 0.094 0.163 A-P Diameter Midshaft 79 0.053 0.067 0.083 S-I Neck Diameter 53 0.056 0.063 0.082 Length 136 0.014 0.015 0.033 Max Breadth of the Prox Epiphysis 104 0.026 0.031 0.041 Max Breadth of the Dist Epiphysis 101 0.042 0.051 0.078 Max Diameter at Nutrient Foramen 138 0.073 0.095 0.127 Transverse Diameter at Nutrient 122 0.083 0.097 0.424 47 0.060 0.073 0.090 49 0.074 0.086 0.094 Distal to Radial Tuberosity Min Diameter of Diaphysis Distal to Radial Tuberosity Ulna Os Coxa Femur Tibia Foramen Max A-P Diameter Distal to Popliteal Line Min A-P Diameter Distal to Popliteal Line 11 Skeletal Element Measurement* N 90th 95th Max M Fibula Length 107 0.013 0.016 0.041 Max Diameter Midshaft 75 0.092 0.118 0.133 Min Diameter of Diaphysis 58 0.108 0.128 0.149 Length 73 0.029 0.033 0.056 Middle Breadth 66 0.045 0.050 0.085 Calcaneus *For a definition of the measurements, see Thomas et al. (2013) Key: Max = maximum; Min = minimum Articulating Bone Portions Models for comparing adjoining bones are based on the difference in size of the articulating surfaces of these bones (Buikstra et al. 1984). For example, to test if an os coxa and a femur belong to the same individual, we subtract the maximum femoral head diameter from the maximum acetabulum diameter. The model takes the general form (Byrd and LeGarde 2014): where measurement i of bone c is subtracted from measurement j of bone d. The D value obtained from the skeletons under study is compared against the mean D value calculated from reference data in order to test the null hypothesis that the size difference between the two adjoining bones is small enough for them to come from the same individual. Byrd and LeGarde (2014) have used a broad American sample as reference. In order to obtain a p-value, the difference of D from the reference data mean is divided by the reference data standard deviation and evaluated against the two-tailed t-distribution. As was the case for pairing skeletal elements, to apply the Byrd and LeGarde (2014) method, a reference sample is necessary to estimate the standard deviation and the mean value of the D-values in order to produce a t-statistic. If this is not available, the values provided by the authors may be used, so long as the same measurements as theirs are adopted in the comparisons (Table 5). See the original publication for a more detailed description of the measurements. Table 5. Joints and reference population data for comparison of articulating elements (from Table 8-10 in Byrd and LeGarde 2014) Joint D N Mean Stand dev Shoulder Humerus head A-P breath – Glenoid fossa max breath 159 6.61 2.35 Elbow A Humerus capitulum-trochlea breadth – Radius head max diam 156 20.99 2.38 Elbow B Humerus capitulum-trochlea breadth – Breadth at distal end of ulnar semi-lunar notch 166 20.49 2.39 Hip Max diam of acetabulum – Femur max head diam 176 9.66 1.67 Knee Femur epicondylar breadth – Tibia max breadth of proximal epiphysis 270 5.20 2.20 Ankle Tibia max breadth of distal epiphysis –Talus min breadth of articular surface 147 17.91 2.85 12 A step-by-step example: Measurement Value Femur epicondylar breadth 88 Tibia max breadth of proximal epiphysis 86.7 D 1.3 Reference sample mean (N = 270) 5.2 Reference sample standard deviation (N = 270) 2.2 t (calculated as |D-5.2|/2.2) 1.77 p (for t-distribution with d.f.= 269, 2 tails) 0.0779 Other authors have developed regression equations for matching articulating elements (for example, see Anastopoulou et al. 2018 and 2019 for regression equations for matching articulating os coxae, femora, tibiae, tali and calcanei). Other Bone Portions Byrd and Adams (2003) propose the following approach for comparing the size of different bones: The measurements obtained on a bone are summed and the natural logarithm of this sum is used in regression models. To derive the t-value from the case specimens, the following model should be used (Byrd and LeGarde 2014): and Sχ is the reference sample standard deviation of the independent variable. At large sample sizes (N>200), a simpler model can be derived by using the difference between the case specimen dependent variable value and the predicted value divided by the standard error of the estimate: The value resulting from this calculation is then compared to the t distribution. For step-by-step examples and selected reference data, see Byrd and LeGarde (2014). Taphonomy Taphonomic patterns (see section ‘Post-mortem bone alteration’ in ‘Basic guidelines for the excavation and study of human skeletal remains; STARC Guide no. 1 ’) can be used for skeletal reassociation. Skeletal remains in different locations within a grave may be exposed to different taphonomic agents and the resulting bone alterations may be used to sort individuals exposed to different taphonomic processes during primary inhumation, who became commingled at a later stage. However, taphonomy must be used with caution in the sorting process as taphonomic differences can also be observed on the remains of a single individual, especially if they occurred post-disarticulation. Age where ŷ is the predicted value from the regression model, yi the dependent variable value of the case specimen, S.E. is the regression model standard error, N is the sample size used in the calculation of the regression model, xi is the independent variable value of the case specimen, x is the reference sample mean for the independent variable, The size and maturity of the skeletal remains can be particularly helpful in discriminating the remains of adult from nonadult remains. In this direction, a method has been proposed by Schaefer (2014) in order to identify commingling among juvenile remains or between juvenile and adult remains based on the stage of epiphyseal union (Figures 3-5). Figures 3-4: The tree diagrams in Figures 3 and 4 demonstrate the general sequence in epiphyseal union (adapted from Schaefer 2014). In each figure the central “tree trunk” represents the modal sequence pattern, while variations to this pattern are shown as “tree branches.” To the left of the trunk are epiphyses that occasionally commence/complete union before the “trunk”. Figures 5: To the left of the antenna diagram are the epiphyses that have completed union before the reference epiphysis, which is given in the middle of the diagram. The ratios in the right-hand boxes express the number of individuals with fused named epiphysis, while the reference epiphysis was still open to the number of individuals where both epiphyses were in the process of uniting. 13 “Beginning” union Begins union prior to reference epiphysis in a minority of cases Modal sequence pattern Begins union subsequent to reference epiphysis in a minority of cases Acetabulum Prox ulna/Dist humerus Coracoid process Acrom Dist tibia Med humerus Prox femur Isch tub Less troch Med hum Less troch Acrom Isch tub Less troch Isch tub Great troch Acrom Isch tub Acrom Dist femur Prox radius Prox femur Med humerus Med clavicle Prox Prox femur radius Dist tibia Prox radius Less trochant Dist Med Prox Great trochant tibia humerus femur Less Isch tuberos trochant Dist fibula Dist femur Prox radius Less Great Dist trochant trochant tibia Prox tibia Med clavicle Prox humerus Iliac crest Acromion Dist radius Prox humerus Iliac crest Prox fibula Med clavicle Prox humerus Iliac crest Dist radius Prox humerus Med clavicle Dist radius Prox Acromion Dist fibula femur Prox humerus Med clavicle Iliac crest Med clavicle Med clavicle Prox femur Dist femur Acrom process Isch Prox tibia tuberos Less Distal Great Isch Prox femur tuberos trochant trochant femur Prox fibula Acromion Iliac crest Dist radius Prox Prox Acromion Dist fibula femur Iliac Acromion Dist ulna humerus crest Med clavicle Dist ulna Prox Iliac Dist radius humerus crest Prox fibula Acro- Dist mion femur Figure 3. Tree diagram demonstrating the overall sequence in which epiphyses begin union (redrawn from Schaefer 2014) 14 “Complete” union Modal sequence pattern Completes union prior to reference epiphysis in a minority of cases Completes union subsequent to reference epiphysis in a minority of cases Dist humerus Prox ulna Med humerus Prox radius Coracoid process Acetabulum Less trochant Dist tibia Med humerus Acetabulum Prox radius Acetabulum Coracoid process Coracoid process Prox Med radius humerus Dist tibia Acrom Isch Dist tuber radius Dist fibula Prox femur Dist fibula Great trochant Prox fibula Dist fibula Less trochant Prox Prox femur radius Less Great trochant trochant Dist ulna Prox fibula Prox tibia Acromion Dist ulna Dist femur Prox fibula Prox tibia Isch tuber Dist ulna Dist femur Prox fibula Dist fibula Dist femur Dist radius Dist ulna Dist fibula Acromion Prox tibia Prox fibula Dist femur Isch tuber Dist ulna Isch tuber Dist radius Prox humerus Isch tuberos Iliac crest Prox radius Prox femur Great trochant Dist fibula Prox femur Acromion Dist fibula Prox tibia Acromion Great Dist fibula trochant Prox fibula Prox Acromion tibia Dist femur Prox fibula Prox tibia Acromion Dist fibula Dist Acromion femur Dist radius Prox humerus Iliac crest Less Prox femur trochant Dist ulna Prox Acromion tibia Isch tuber Prox humerus Figure 4. Tree diagram demonstrating the overall sequence in which epiphyses complete union (redrawn from Schaefer 2014) 15 always completes before reference epiphysis begins some�mes completes before reference epiphysis begins distal humerus proximal ulna prox ulna 2:1, acetabulum 1:1 prox ulna 2:4 coracoid process coracoid process 2:3, prox ulna 4:1, med humerus 1:1, prox radius 1:4 coracoid process, dist humerus med humerus dist humerus proximal radius coracoid process 1:2, prox ulna 2:1, med humerus 1:3, prox radius 1:5 acetabulum med humerus 2:3, prox radius 2:4, acetabulum 1:2 distal �bia prox ulna, dist humerus dist humerus lesser trochanter dist humerus proximal femur coracoid process, prox ulna, dist humerus prox radius 1:3 prox fibula 2:6, acetabulum 8:1, dist fibula 1:6, prox radius 7:1, prox femur 5:4, great trochanter 2:4, dist �bia 8:1, prox �bia 1:8, less trochanter 4:2 prox radius 2:2 greater trochanter distal fibula coracoid process, prox ulna, med humerus, dist humerus acromion process coracoid process, prox ulna, med humerus, dist humerus, acetabulum proximal �bia dist fibula 1:5, prox radius 1:8, prox femur 5:4, great trochanter 2:2, prox �bia 1:10, less trochanter 4:1 dist fibula 2:8, prox radius 8:1, prox femur 4:6, great trochanter 2:6, prox �bia 1:15, less trochanter 3:3, acromion process 2:10 proximal fibula coracoid process, prox ulna, med humerus, dist humerus, acetabulum, dist �bia dist fibula 9:1, prox fibula 2:1, prox radius 24:1, prox femur 16:1, great trochanter 14:1, prox �bia 4:2, acromion process 6:1 distal femur coracoid process, prox ulna, med humerus, dist humerus, acetabulum, dist �bia, less trochanter distal ulna distal radius prox ulna, med humerus, dist humerus, coracoid process prox ulna, med humerus, dist humerus, coracoid process, acetabulum ischial tuberosity proximal humerus iliac crest prox ulna, med humerus, dist humerus, coracoid process, acetabulum, prox radius, less trochanter medial clavicle dist fibula 7:4, prox fibula 1:5, prox radius 19:1, prox femur 12:2, great trochanter 10:2, prox �bia 3:6, acromion process 4:2 prox radius 2:3, acetabulum 1:2 dist fibula 3:6, prox fibula 1:7, prox radius 13:1, prox femur 7:3, great trochanter 5:3, acromion process 2:8, dist �bia 11:1, less trochanter 8:2 dist fibula 3:6, prox fibula 1:9, prox radius 10:1, prox femur 8:5, great trochanter 4:3, acromion process 1:10, dist �bia 12:1, less trochanter 7:2, prox �bia 1:12 all remaining epiphyses Figure 5. Antenna diagram (redrawn from Schaefer 2014) 16 Chemical analysis Recent research using X-ray fluorescence (XRF) spectrometry and laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) have shown promise in determining whether a set of remains belongs to a single or multiple individuals by analysing the elemental concentration in human bones (Castro et al. 2010; Gonzales-Rodriguez and Fowler 2013; Perrone et al. 2014; Stevens 2016). Note that the former technique is non-destructive, while the latter requires part of the bone to be destroyed during analysis. Inter-skeletal differences in bone mineral composition may be due to in vivo uptake from food and water, metabolic functions, chemical exposure or other means of absorption (Perrone et al. 2014). A confounding factor is that trace elements are stored differently throughout the human skeleton (Pemmer et al. 2013; Price 1989; Wittmers et al. 1988), thus, intra-individual variability in elemental composition may exceed inter-individual variability (Budd et al. 2000; Finnegan 1988; Grupe 1988). Trace elemental signatures also vary throughout an individual’s lifespan due to age-related metabolic and physiological functions (Darrah et al. 2009). Another confounding factor is postmortem contamination (diagenesis), which may alter the elemental concentrations in buried human bones. However, surface contamination should not be an issue as x-rays penetrate the bone surface by several millimetres during chemical analysis (Shackley 2011). Table 6. Skeletal elements per tier based on success rates in extracting DNA (drawn from Hines et al. 2014) Tier 1 Tier 2 Tier 3 Tier 4 High success rates Moderate success rates Successful less than half the time Even less successful Teeth Os coxa Ribs Clavicle Talus & other tarsals Fibula Cranium Ulna Scapula Humerus Radius Petrous portion of temporal Femur Hines et al. (2014) present a practical protocol for the effective sampling for partial/commingled remains, depending on the body parts that may be encountered in any given case. Assessment of the success rates for the various elements allowed categorization of sample success rates into four tiers (Table 6). Table 7 summarises the procedure Hines et al. (2014) recommend for sampling skeletal remains for DNA analysis. Figure 7 shows the locations for sampling per element. In black you see the first priority areas and in grey the second priority areas. Note that teeth should be extracted whole. Anterior maxillary or mandibular molars or premolars are preferred. The teeth sampled should not have caries or post-mortem damage, if possible. Metacarpals Vertebrae Tibia Metatarsals Table 7. Procedure for sampling skeletal remains for DNA analysis per Hines et al. (2014) BEFORE SAMPLING 1. Take photo of the element in its original condition (include photo scale and label) 2. Clean tools: 1. Water rinse to remove adherent material • 2. Rinse with a solution of 10% commercial bleach or wipe with bleach • 3. Rinse with ethanol DNA profile data DNA analysis is increasingly employed in commingling cases (e.g. Holland et al. 2003; Just et al. 2009; Mundorff et al. 2014; Parsons et al. 2007; Primorac 2004; Verdugo et al. 2017). In osteoarchaeological analysis, ancient DNA data mostly aim at addressing issues of kinship, migration patterns and genetic diseases, while in forensic contexts the aim of DNA analysis is the identification of the unknown individuals. Given its cost and destructive nature, DNA analysis should be best used in conjunction with the context of the remains and the results of the macroscopic skeletal analysis (Puerto et al. 2014). Mandibular body DURING SAMPLING 1. Use protective equipment (gloves, mask, safety glasses) to avoid contamination 2. Aim for sampling 12 to 25 grams (4 grams is minimal but acceptable) 3. Avoid sampling areas where the bone is discolored 4. Use particulate/fume extraction facilities if multiple samples are being taken, particularly in enlosed spaces 5. Place each sample in its own container, with a unique specimen number. It is important that the specimen is completely dried before packaging, and breathable (eg. paper) packaging, should be used where practical AFTER SAMPLING 1. Take photo of the sampled element with the extracted location clearly shown, the extraced sample code, and a photo scale 2. Do not expose the sample to conditions of elevated heat or humidity 17 For an alternative approach in collecting femur, rib, and tooth samples for DNA analysis in forensic settings, see Westen et al. (2008). ESTIMATION OF THE NUMBER OF INDIVIDUALS Estimates of the number of individuals present in a commingled assemblage fall under two broad categories: Minimum Number of Individuals (MNI) estimators and Most Likely Number of Individuals (MLNI) estimators. Minimum Number of Individuals (MNI) The MNI expresses the least number of individuals required to account for the skeletal elements present in the assemblage that has been recovered. The most common way to estimate the MNI is by sorting the bones by side and element and then taking the most frequent element as the estimate. In other words, the MNI is equal to the most repeated element after sorting by element and side: Max (L, R) (White 1953). In cases of fragmentary remains, make sure that there is an overlap of anatomical features on the fragmented remains (e.g. greater trochanter) in order to avoid counting the same individuals more than once. A variant of the MNI is the Grand Minimum Total (GMT) and is calculated as L + R – P, where P signifies the number of bone pairs (Chaplin 1971). This technique assumes that unpaired bones originate from different individuals (Adams and Konigsberg 2004). It requires the accurate identification of all pairs between the bilateral elements of the assemblage, while incorrect matches will bias the results. Figure 7. Locations for sampling for ancient DNA (adapted from Hines et al. 2014) 18 Example Assume that we have an assemblage where femora are the most numerous elements recovered. We have 145 left, 130 right, and 95 pairs of femora in our sample. Then: With the MLNI, it is possible to calculate confidence intervals. An approximate confidence interval can be calculated using the following equation: MNI = Max (L, R) = 145 GMT=L+R–P=180 In some contexts, MNI may be inferred from the biological profile of the elements. For example, the presence of some elements clearly suggestive of a male individual and others suggestive of a female, indicate that at least two individuals were buried together. Similarly, the observation of considerable differences in bone size, especially between bilateral or adjoining elements, also supports the presence of multiple individuals. Even though MNI and GMT can provide useful information on the smallest number of individuals that comprise the assemblage, they underestimate the true sample size whenever the recovery rates are less than 100%, which is often the case in archaeological assemblages (Konigsberg and Adams 2014; Nikita and Lahr 2011). Lincoln Index (LI) and the Most Likely Number of Individuals (MLNI) For example, an approximate 95% CI can be calculated as . However, because the number of individuals, N, follows a discrete distribution, it is not statistically accurate to give customary confidence intervals, such as 95% intervals. Adams and Konigsberg (2004) propose using instead the highest density region (HDR). For more information on the HDR and its calculation, see Adams and Konigsberg (2004) and the website http:// konig.la.utk.edu/MLNI.html. As with GMT, this method relies on the ability to make accurate pair matches between elements. Therefore, it is important for the elements to be well-preserved (Konigsberg and Adams 2014). The accuracy of MLNI estimators is expected to improve when multiple skeletal elements are taken into account simultaneously instead of using only the most abundant bone. For this reason, the following equations have been proposed (Nikita 2014): LI and MLNI estimators assess the initial number of individuals that comprised the assemblage under study based on the fact that the probability of identifying P pairs between R right and L left bones from N initial individuals follows the hypergeometric distribution (Adams and Konigsberg 2004). When using the LI or MLNI, it is important that probabilities of sampling the left and right sides within individuals are independent. An estimate of the original assemblage represented by the skeletal elements is determined by: The LI is a good approximation of the MLNI. A modification to this formula to account for sample bias was proposed by Seber (1973). Adams and Konigsberg (2004) have shown that Seber's formula represents the maximum likelihood estimate and refer to it as the Most Likely Number of Individuals (MLNI). It is calculated as: In the above equations n is the number of the various types of bones, L and R is the number of left and right elements respectively, subscripts 1, 2, …, n denote each skeletal element under study (e.g., 1 = femora, 2 = tibiae, etc.), and P is the sum of all pairs. where the symbols L represent the floor function that removes any decimal points. 19 Example Consider the following assemblage of human skeletal elements: Femora: R = 9, L = 9, P = 6 Tibiae: R = 7, L = 7, P = 4 Humeri: R = 7, L = 9, P = 2 Lincoln’s Index gives the following estimates: Nfemora = (9×9)/6 = 13.5, thus 14 individuals Ntibiae = (7×7)/4 = 12.25, thus 12 individuals Nhumeri = (7×9)/2 = 31.5, thus 32 individuals The discrepancy of the obtained values from different elements is due to the differential preservation of these elements and in our case it is attributed to the small number of pairs identified between right and left humeri. The average MLNI value is 19. In contrast, the values obtained from equations employing multiple elements simultaneously are: N = (9x9+7x7+7x9)/(6+4+2) = 193/12 = 16.1 N = (9+7+7+9+7+9)2/(4×3×(6+4+2)) = 2304/144 = 16 N = ((9+9)2 + (7+7)2 + (7+9)2)/((4×(6+4+2)) = 776/48 = 16.2 N = ((9+1)(9+1)+(7+1)(7+1)+(7+1)(9+1))/((6+4+2)+3)-1 = 244/15 - 1 = 15.3 It is seen that the first three equations employing multiple elements yield the same result, N = 16, and this may be adopted as the most likely number of individuals for the assemblage under consideration. MNI versus MLNI The MLNI estimates the original number of individuals that comprised the skeletal assemblage, whereas the MNI expresses only a minimum estimate. Furthermore, it is possible to provide confidence intervals with the MLNI, but not with the MNI. Thus, the MLNI should be preferred over the MNI, but in highly fragmented remains, estimation of the MNI may be the only viable option. SEX ASSESSMENT In commingled assemblages, sex assessment has to be performed on an element by element basis, except for cases of small-scale commingling where most elements have been sorted per skeleton. The methods that may be adopted are given in STARC Guide no. 1. As most morphological traits for sexing focus on the pelvis and the skull, most disassociated bones will remain unsexed. For this reason, we would recommend the additional adoption of metric methods for sex estimation from the postcranial skeleton, always bearing in mind the population-specificity of these methods and the potential impact of secular change. A compilation of worldwide studies on metric sex estimation using different skeletal elements can be found in Nikita (2017). AGE-AT-DEATH ESTIMATION In commingled assemblages, age estimation also has to be performed on an element by element basis, except for cases of small-scale commingling where most elements have been sorted per skeleton. Age-at-death is very difficult to estimate from isolated elements and most elements can only be 20 assigned to the general “adult” or “nonadult” categories based on their size. All fully fused or ‘adult-sized’ bone fragments and all permanent teeth with closed root apices and some degree of dental wear will be classified as “adult”. Deciduous teeth, still forming permanent teeth, bones with unfused epiphyses (except for late-fusing elements) and all bones that are clearly too small to be adult will be classified as “nonadult”. In cases where the skeletal or dental elements preserve sufficient information (e.g. the pubic symphysis or auricular surface on the os coxa), traditional methods for estimating skeletal age-at-death should be used (see STARC Guide no. 1). The age-at-death distribution of an assemblage comprised of commingled individuals should be established using the most abundant skeletal element that can be sided and aged (Siebke et al. 2019), except in cases of small-scale commingling, where the skeletal elements have been largely sorted per individual skeleton. PATHOLOGICAL LESIONS In commingled assemblages the fact that pathological lesions are identified on individual skeletal elements and their overall distribution on the skeleton cannot be examined, limits the potential of accurate diagnosis. In addition, it is difficult to estimate the actual prevalence of a pathological condition in the assemblage. Nonetheless, as highlighted by Brickley and Buckberry (2015), there is still great value in the palaeopathological analysis of partial and poorly preserved skeletons as important information can be obtained for conditions that affect the entire skeleton (e.g. metabolic bone diseases) or conditions which can be diagnosed in individual elements (e.g. osteoarthritis, fractures). See STARC Guide no. 1 for a brief description of different pathological conditions that may be identified on skeletal elements. ACTIVITY MARKERS See STARC Guide no. 1 for information on long bone crosssectional geometric properties, entheseal changes, dental wear and osteoarthritis. When estimating cross-sectional geometric properties on isolated long bones, it may be impossible to standardize biomechanical properties using body mass as a proxy for body size (Ruff 2008). In such cases, powers of bone length may be used: for second moments of area the recommended power is (bone length)5.33, whereas for the total area it is (bone length)3 (Ruff et al. 1993). NONMETRIC TRAITS with commingled remains except for cases of small-scale commingling, where a full re-association of all elements per skeleton has been accomplished. Regarding mathematical methods, whereas in articulated skeletons it is advisable to use the long bones of the lower limbs when estimating stature by means of regression equations, in disarticulated remains, every available element may/should be used. STARC Guide no. 1 gives representative equations for European and American populations, while Nikita (2017) provides a compilation of population-specific studies which use not only long bones but also other skeletal elements. Finally, as commingled remains are often also fragmented, stature estimation may be based on equations proposed for fragmented remains (e.g. Bidmos 2008). When estimating stature using regression equations, it is imperative to check the standard error of estimate as equations using short bones and fractured elements usually have higher error rates. This fact, coupled with the population-specificity of relevant equations and the effect of secular change, may render stature estimation impractical in many cases. POST-MORTEM BONE ALTERATION See STARC Guide no. 1 for post-mortem bone alteration. Note that many different agents can produce the same morphological bone alterations and their discrimination will be especially difficult in commingled remains where assessment often needs to be made on an element-byelement basis. See STARC Guide no. 1 for information on cranial, postcranial and dental nonmetric traits. Note that traits which exhibit a bilateral expression are often inspected on both sides of the body and subsequently the strongest degree of expression is the one recorded in the database. In cases of commingled remains (e.g. loose teeth or unassociated long bones belonging to multiple individuals) it may be impossible to assess which elements form pairs. In such cases, it is advisable to record nonmetric traits only on one side (either the right or the left) in order to avoid having the same individual twice in the dataset, which would bias the results. MORPHOSCOPIC TRAITS See STARC Guide no. 1 for information on morphoscopic traits. METRICS See STARC Guide no. 1 for cranial, postcranial and dental measurements. STATURE ESTIMATION As explained in STARC Guide no. 1, stature estimation from skeletal remains is based on anatomical and mathematical methods. Anatomical methods are not possible to use 21 ADDITIONAL RESOURCES Online database for recording human commingled remains: Osterholtz AJ. 2018. A FileMaker Pro database for use in the recording of Commingled and/or Fragmentary Human Remains. Mississippi State University: Department of Anthropology and Middle Eastern Cultures. http://hdl.handle.net/11668/14276 Documentation for the above database: Osterholtz AJ. 2019. Advances in documentation of commingled and fragmentary remains. Advances in Archaeological Practice 7: 77–86 REFERENCES Adams BJ. 2014. Commingled remains: Field recovery and laboratory analysis. In: Smith C (ed.) Encyclopedia of Global Archaeology. New York: Springer; p. 1575-1580 Adams BJ, Konigsberg LW. 2004. Estimation of the Most Likely Number of Individuals from commingled human skeletal remains. American Journal of Physical Anthropology 125: 138–151 Adams BJ, Byrd JE. 2006. Resolution of small-scale commingling: A case report from the Vietnam War. Forensic Science International 156: 63–69 Adams BJ, Byrd JE (eds.) 2008. Recovery, Analysis, and Identification of Commingled Human Remains. New York: Springer Adams BJ, Byrd JE (eds.) 2014. Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press Anastopoulou I, Karakostis, FA, Borrini M, Moraitis K. 2018. A statistical method for reassociating human tali and calcanei from a commingled context. Journal of Forensic Sciences 63: 381-385 Anastopoulou I, Karakostis, FA, Moraitis K. 2019. A reliable regression-based approach for reassociating human skeletal elements of the lower limbs from commingled assemblages. Journal of Forensic Sciences 64: 502-506 Beckett J, Robb J. 2006. Neolithic burial taphonomy, ritual and interpretation in Britain and Ireland: A review. In: Gowland R, Knüsel CJ (eds.) The Social Archaeology of Funerary Remains. Oxford: Oxbow Books; p. 57-80 Bidmos MA. 2008. Stature reconstruction using fragmentary femora in South Africans of European descent. Journal of Forensic Sciences 53: 1044-1048 Brickley MB, Buckberry J. 2015. Picking up the pieces: Utilizing the diagnostic potential of poorly preserved remains. International Journal of Paleopathology 8: 51–54 Budd P, Montgomery J, Barreiro B, Thomas RG. 2000. Differential diagenesis of strontium in archaeological human dental tissues. Applied Geochemistry 15: 687-694 Buikstra JE, Gordon CC, St. Hoyme L. 1984. The case of the severed skull. Individuation in forensic anthropology. In: Rathbun TA, Buikstra JE (eds.) Human Identification: Case Studies in Forensic Anthropology. Springfield, IL: Charles C. Thomas; p. 121–135 Byrd JE. 2008. Models and methods for osteometric sorting. In: Adams BJ, Byrd JE (eds.) Recovery, Analysis, and Identification of Commingled Human Remains. New York: Springer; p. 199-220 Byrd JE, Adams BJ. 2003. Osteometric sorting of commingled human remains. Journal of Forensic Sciences 48: 717-724 Byrd JE, LeGarde CB. 2014. Osteometric sorting. In: Adams BJ, Byrd JE (eds.) Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press; p. 167-192 22 REFERENCES Calleja M. 2016. Commingled Tombs and ArcGIS: Analyzing the Mortuary Context and Taphonomy at Bronze Age Tell Abraq. Masters Dissertation, University of Nevada, Las Vegas Castro W, Hoogewerff J, Latkoczy C, Almirall JR. 2010. Application of laser ablation (LA-ICP-SF-MS) for the elemental analysis of bone and teeth samples for discrimination purposes. Forensic Science International 195: 17-27 Chaplin R. 1971. The Study of Animal Bones from Archaeological Sites. London: Seminar Press Christensen AM, Passalacqua NV, Bartelink EJ. 2014. Forensic Anthropology: Current Methods and Practice. San Diego: Academic Press Cox M, Flavel A, Hanson I, Laver J, Wessling R. 2008. The Scientific Investigation of Mass Graves. Cambridge: Cambridge University Press Darrah TH, Prutsman-Pfeiffer JJ, Poreda RJ, Campbell ME, Hauschka PV, Hannigan RE. 2009. Incorporation of excess gadolinium into human bone from medical contrast agents. Metallomics 1: 479-488 Duday H. 1985. Nouvelles observations sur la décomposition des corps dans un espace libre. Méthode d’étude des sépultures. Saint-Germain en Laye: Nouvelles de l’Archéologie. Supplément à MSH Informations Paris; p. 6-13 Duday H. 2005. L’Archéothanatologie ou l’Archéologie de la Mort. In: Dutour O, Hublin J-J, Vandermeesch B (eds.) Objets et Méthodes en Paléoanthropologie. Paris: Comité des Travaux Historiques et Scientifiques; p. 153-215 Duday H. 2009. The Archaeology of the Dead: Lectures in Archaeothanatology. Oxford: Oxbow Books Dupras TL, Schultz JJ, Wheeler SM, Williams LJ. 2012. Forensic Recovery of Human Remains. Archaeological Approaches, 2nd edition. Boca Raton: CRC Press Emberling G, Robb J, Speth J, Wright H. 2002. Kunji cave: early bronze age burials in Luristan. Iranian Antiquities 37: 47–104 Finnegan M. 1988. Variation of Trace Elements Within and Between Skeletons Using Multiple Sample Sites. Paper presented at the 15th Annual Meeting of the Paleopathology Association, Kansas City, MO Gonzalez-Rodriguez J, Fowler G. 2013. A study on the discrimination of human skeletons using X-ray fluorescence and chemometric tools in chemical anthropology. Forensic Science International 231: 407.e1-6 Grupe G. 1988. Impact of the choice of bone samples on trace element data in excavated human skeletons. Journal of Archaeological Science 15: 123-129 Hines DZC, Vennemeyer M, Amory S, Huel RLM, Hanson I, Katzmarzyk C, Parsons TJ. 2014. Prioritized sampling of bone and teeth for DNA analysis in commingled cases. In: Adams BJ, Byrd JE (eds.) Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press; p. 275-305 Holland MM, Cave CA, Holland CA, Bille TW. 2003. Development of a quality, high throughput DNA analysis procedure for skeletal samples to assist with the identification of victims from the world trade center attacks. Croatian Medical Journal 44: 264–272 Just RS, Leney MD, Barritt SM, Los CW, Smith BC, Holland TD, Parsons TJ. 2009. The use of mitochondrial DNA single nucleotide polymorphisms to assist in the resolution of three challenging forensic cases. Journal of Forensic Sciences 54: 887-891 Knüsel CJ, Robb J. 2016. Funerary taphonomy: An overview of goals and methods. Journal of Archaeological Science: Reports 10: 655–673 Konigsberg LW, Adams BJ. 2014. Estimating the number of individuals represented by commingled human remains: A critical evaluation of methods. In: Adams BJ, Byrd JE (eds.) Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press; p. 193-220 Lynch JJ, Byrd J, LeGarde CB. 2018. The power of exclusion using automated osteometric sorting: Pair-matching. Journal of Forensic Sciences 63: 371-380 Moutafi I. 2016. The human remains from Area A. In: Renfrew AC, Philaniotou O, Brodie N, Gavalas G, Boyd MJ (eds.) Kavos and the Special Deposits (The sanctuary on Keros and the origins of Aegean ritual practice: the excavations of 2006– 2008. Vol. II). Cambridge: McDonald Institute for Archaeological Research; p. 483-505 Mundorff AZ, Shaler R, Bieschke ET, Mar-Cash E. 2014. Marrying anthropology and DNA: essential for solving complex commingling problems in cases of extreme fragmentation. In: Adams BJ, Byrd JE (eds.) Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press; p. 257-273 23 REFERENCES Naji S, de Becdeliévre C, Djouad S, Duday H, André A, Rottier S. 2014. Recovery methods for cremated commingled remains: Analysis and interpretation of small fragments using a bioarchaeological approach. In: Adams BJ, Byrd JE (eds.) Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press; p. 33-56 Nikita E. 2014. Estimation of the original number of individuals using multiple skeletal elements. International Journal of Osteoarchaeology 24: 660-664 Nikita E. 2017. Osteoarchaeology: A Guide to the Macroscopic Study of Human Skeletal Remains. San Diego: Academic Press Nikita E, Lahr MM. 2011. Simple algorithms for the estimation of the initial number of individuals in commingled skeletal remains. American Journal of Physical Anthropology 146: 629-636 Osterholtz AJ (ed.) 2016. Theoretical Approaches to Analysis and Interpretation of Commingled Human Remains. New York: Springer Osterholtz AJ, Baustian KM, Martin DL (eds.) 2014a. Commingled and Disarticulated Human Remains: Working Toward Improved Theory, Method, and Data. New York: Springer Osterholtz AJ, Baustian KM, Martin DL. 2014b. Introduction. In: Osterholtz AJ, Baustian KM, Martin DL (eds.) Commingled and Disarticulated Human Remains: Working Toward Improved Theory, Method, and Data. New York: Springer; p. 1-13 Outram AK. 2001. A new approach to identifying bone marrow and grease exploitation: why the “indeterminate” fragments should not be ignored. Journal of Archaeological Science 28: 401-410 Papathanasiou A. 2009. The human osteological material from the Mycenaean tholos at Kazanaki, Volos. Proceedings of the 2nd archaeological meeting of Thessaly and Central Greece, Volos, Greece; p. 151-161 Parsons TJ, Huel R, Davoren J, Katzmarzyk C, Miloš A, Selmanović A, Coble MD, Rizvić A. 2007. Application of novel “miniamplicon” STR multiplexes to high volume casework on degraded skeletal remains. Forensic Science International: Genetics 1: 175-179 Pemmer B, Roschger A, Wastl A, Hofstaetter JG, Wobrauschek P, Simon R, Thaler HW, Roschger P, Klaushofer K, Streli C. 2013. Spatial distribution of the trace elements zinc, strontium and lead in human bone tissue. Bone 57: 184-193 Perrone A, Finlayson JE, Bartelink EJ, Dalton KD. 2014. X-ray fluorescence (XRF) for sorting commingled human remains. In: Adams BJ, Byrd JE (eds.) Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press; p. 145-165 Price TD (ed.) 1989. The Chemistry of Prehistoric Human Bone. Cambridge: Cambridge University Press Primorac D. 2004. The role of DNA technology in identification of skeletal remains discovered in mass graves. Forensic Science International 146: S163–S164 Puerto MS, Egaña S, Doretti M, Vullo CM. 2014. 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; p. 307-335 Roksandic M. 2002. Position of skeletal remains as a key to understanding mortuary behavior. In: Haglund WD, Sorg MH (eds.) Advances in Forensic Taphonomy: Method, Theory, and Archaeological Perspectives. Boca Raton: CRC Press; p. 99-117 Ruff CB. 2008. Biomechanical analyses of archaeological human skeletons. In: Katzenberg MA, Saunders SR (eds.) Biological Anthropology of the Human Skeleton. New York: Wiley Liss; p. 183-206 Ruff CB, Trinkaus E, Walker A, Larsen CS. 1993. Postcranial robusticity in Homo. I. Temporal trends and mechanical interpretation. American Journal of Physical Anthropology 91: 21-53 Schaefer M. 2014. A practical method for detecting commingled remains using epiphyseal union. In: Adams BJ, Byrd JE (eds.) Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press; p. 123-144. Seber GAF. 1973. The Estimation of Animal Abundance and Related Parameters. London: Griffin Shackley SM. 2011. An introduction to X-ray fluorescence (XRF) analysis in archaeology. In: Shackley, SM (ed.) X-Ray Fluorescence Spectrometry (XRF) in Geoarchaeology. New York: Springer; p. 7-44 Siebke I, Steuri N, Furtwängler A, Ramstein M, Arenz G, Hafner A, Krause J, Lösch S. 2019. Who lived on the Swiss Plateau around 3300 BCE? Analyses of commingled human skeletal remains from the Dolmen of Oberbipp. International Journal of Osteoarchaeology 29: 786-796 24 REFERENCES Stevens WD. 2016. Enslaved Labor in the Gang and Task Systems: A Case Study in Comparative Bioarchaeology of Commingled Remains. Doctoral dissertation, University of South Carolina Thomas RM, Ubelaker DH, Byrd JE. 2013. Tables for the metric evaluation of pair-matching of human skeletal elements. Journal of Forensic Sciences 58: 952-956 Tuller H, Hofmeister U. 2014. Spatial analysis of mass grave mapping data to assist in the reassociation of disarticulated and commingled human remains. In: Adams BJ, Byrd JE (eds.) Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press; p. 7-32 Ubelaker DH. 2014. Commingling analysis: Historical and methodological perspectives. In: Adams BJ, Byrd JE (eds.) Commingled Human Remains: Methods in Recovery, Analysis, and Identification. San Diego: Academic Press; p. 1-6 Verdugo C, Kassadjikova K, Washburn E, Harkins KM, Fehren-Schmitz L. 2017. Ancient DNA clarifies osteological analyses of commingled remains from Midnight Terror Cave, Belize. International Journal of Osteoarchaeology 27: 495-499 Vickers S, Lubinski PM, DeLeon LH, Bowen JT. 2015. Proposed method for predicting pair matching of skeletal elements allows too many false rejections. Journal of Forensic Sciences 60: 102-106 Westen AA, Gerretsen RRR, Maat GJR. 2008. Femur, rib, and tooth sample collection for DNA analysis in disaster victim identification (DVI). A method to minimize contamination risk. Forensic Science, Medicine and Pathology 4: 15–21 White TE. 1953. A method of calculating the dietary percentage of various food animals utilized by aboriginal peoples. American Antiquity 18: 393-399 Wittmers Jr LE, Wallgren J, Alich A, Aufderheide AC, Rapp Jr G. 1988. Lead in bone. IV. Distribution of lead in the human skeleton. Archives of Environmental Health: An International Journal 43: 381-391 25 RECORDING SHEETS RECORDING SHEETS RECORDING SHEET FOR COMMINGLED HUMAN SKELETAL REMAINS For cases where the remains have been (partially) sorted by individual, it is advisable to use the form given in STARC Guide no. 1 for articulated skeletons. The forms given here are for individual unassociated skeletal elements. Note that when working with commingled remains, it is generally impractical to use printed forms. Instead, try to fit the information given below in a spreadsheet (e.g. in Excel) whereby each individual element occupies a row and each variable is given in a column. GENERAL INFORMATION Archaeological site: Curation site: Recorder: Date: Burial No: Grave type: Grave size: Field methods for site identification: Field methods for site excavation: Cleaning methods: Restoration methods: 27 RECORDING SHEETS BONE INVENTORY Key: Zones as defined by Knüsel and Outram (2004); record expression per zone as 0 = absent, 1 = present <25%, 2 = present 26-50%, 3 = present 51-75%, 4 = present >76%, or simply as 0 = absent, 1 = present CRANIUM, MANDIBLE, EAR OSSICLES & HYOID Element Zone/Side Frontal Element Zone/Side 1 Vomer - 2 Lacrimal 3 Palatine 4 Ethmoid - Occipital 5 Mandible 1 Temporal 6 2 7 3 8 4 9 5 10 6 11 7 Parietal Sphenoid Zygomatic Maxilla Nasal Inferior nasal concha 28 Expression 12 Malleus 13 Stapes 14 Incus 15 Hyoid - Expression RECORDING SHEETS THORACIC CAGE & VERTEBRAE Element Zone Sternum 1 Rib 1 Left Right Rib 11 1 3 3 1 4 Axis 2 1 3 2 4 C3-7 1 1 2 2 3 3 4 T1-12 1 2 2 3 3 1 4 2 L1-5 3 Expression 1 3 1 Rib 12 Atlas 2 3 Rib 3-10 Zone 2 2 Rib 2 Element 1 2 3 4 SHOULDER GIRDLE Element Zone Clavicle 1 Left Right 2 3 Scapula 4 5 6 7 8 9 29 RECORDING SHEETS UPPER AND LOWER LIMB LONG BONES & PATELLA Element Zone Humerus 1 Radius Element Zone Femur 1 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 1 Patella - 2 Tibia 1 3 2 4 3 5 4 6 5 7 6 8 7 9 8 10 9 11 10 Fibula 1 A&B 2 C 3 D 4 E 5 F 6 G H J 30 Right 2 J Ulna Left Left Right RECORDING SHEETS HAND BONES Element Zone Left Right Scaphoid - Lunate - 2 Triquetral - 3 Pisiform - Trapezium - 2 Trapezoid - 3 Capitate - Hamate - 2 MC1 1 3 MC4 1 Left Right Left Right 1 Proximal phalanx Middle phalanx 1 1 3 2 1 3 2 MC3 Zone MC5 2 MC2 Element Distal phalanx 1 3 2 1 3 2 3 PELVIC BONES Element Zone Os coxa 1 Left Right Element Zone Sacrum 1 2 2 3 3 4 4 5 6 7 8 9 10 11 12 31 RECORDING SHEETS FOOT BONES Element Zone Talus 1 Right Element Zone MT3 1 2 2 3 3 4 Calcaneus Left MT4 1 1 2 2 3 3 MT5 1 4 2 5 3 Navicular - Cuboid - 2 1st Cuneiform - 3 2nd Cuneiform - 3rd Cuneiform - 2 MT1 1 3 2 MT2 Middle phalanx Distal phalanx 1 1 1 3 2 1 3 2 3 32 Proximal phalanx Left Right RECORDING SHEETS UNIDENTIFIED BONE Type Size class Cortical <1 cm No of fragments Weight 1-3 cm 3-5 cm >5cm Trabecular <1 cm 1-3 cm 3-5 cm >5cm Cranial <1 cm 1-3 cm 3-5 cm >5cm Post-cranial <1 cm 1-3 cm 3-5 cm >5cm 33 RECORDING SHEETS DENTAL INVENTORY Key: 1 = Present, not in occlusion, 2 = Present, development completed, in occlusion, 3 = Missing, no associated alveolar bone, 4 = Missing, antemortem loss, 5 = Missing, postmortem loss, 6 = Missing, congenital absence, 7 = Present, damage renders measurement impossible, 8 = Present, unobservable PERMANENT TEETH I1 Maxilla Left Maxilla Right Mandible Left Mandible Right I2 C P3 P4 M1 M2 M3 DECIDUOUS TEETH I1 Maxilla Left Maxilla Right Mandible Left Mandible Right I2 C M1 M2 DENTAL WEAR I1 Maxilla Left Right Mandible Left Right 34 I2 C P3 P4 M1 M2 M3 RECORDING SHEETS SEX ASSESSMENT (ONLY FOR ADULT REMAINS) Key: Record as Female, Probable Female, Ambiguous, Probable Male, Male, Indeterminate Element Trait/Method Sex AGE-AT-DEATH ESTIMATION (FOR NONADULTS) Classify individuals in one of the following categories: fetus = before birth, infant = 0-3 yrs, child = 3-12 yrs, adolescent = 12-20 yrs, nonadult = <18 yrs, indeterminate = unable to estimate age-at-death Element Trait/Method Sex AGE-AT-DEATH ESTIMATION (FOR ADULTS) Classify individuals in one of the following categories: young adult = 20-35 yrs, middle adult = 35-50 yrs, old adult = 50+ yrs, adult = 18+ yrs, indeterminate = unable to estimate age-at-death Element Method Stage Age 35 RECORDING SHEETS PATHOLOGICAL LESIONS Element affected 36 Type of lesion Degree of expression RECORDING SHEETS CRANIOMETRICS Key: All measurements in mm (as defined in Moore-Jansen and Jantz 1989) Measurement Type of Value Maximum cranial breadth Minimum frontal breadth Upper facial breadth Interorbital breadth Biorbital breadth Bizygomatic diameter Nasal breadth Nasal height Upper facial height Orbital height Orbital breadth Frontal chord Basion-bregma height Parietal chord Maximum cranial length Cranial base length Basion-prosthion length Mastoid length Occipital chord Maxillo-alveolar length Maxillo-alveolar breadth Biauricular breadth Foramen magnum breadth Foramen magnum length Chin height Bigonial width Bicondylar breadth Height of the mandibular body Breadth of the mandibular body Mandibular length Maximum ramus height Maximum ramus breadth Minimum ramus breadth 37 RECORDING SHEETS POSTCRANIAL MEASUREMENTS Key: All measurements in mm (as defined in Moore-Jansen and Jantz 1989) Element Measurement Clavicle Maximum length Superior-inferior (vertical) diameter at midshaft Anterior-posterior (sagittal) diameter at midshaft Scapula Height Breadth Humerus Maximum length Maximum midshaft diameter Minimum midshaft diameter Vertical head diameter Ulna Maximum length Physiological length Minimum circumference Anteroposterior (dorsovolar) diameter Mediolateral (transverse) diameter Radius Maximum length Mediolateral (transverse) midshaft diameter Anteroposterior (sagittal) midshaft diameter Os coxa Height Iliac breadth Ischium length Pubis length Sacrum Anterior length Anterosuperior breadth Maximum transverse base diameter Femur Maximum length Subtrochanteric mediolateral (transverse) diameter Subtrochanteric anteroposterior (sagittal) diameter Midshaft circumference Mediolateral (transverse) midshaft diameter Anteroposterior (sagittal) midshaft diameter Bicondylar length Epicondylar breadth Maximum head diameter 38 Left Right RECORDING SHEETS Element Measurement Tibia Length Left Right Left Right Circumference at nutrient foramen Mediolateral (transverse) diameter at nutrient foramen Maximum diameter at nutrient foramen Maximum distal epiphyseal breadth Maximum proximal epiphyseal breadth Fibula Maximum length Maximum midshaft diameter MEASUREMENTS FOR OSTEOMETRIC SORTING Key: Record following Byrd and LeGarde (2014) Element Measurement Humerus Maximum length Epicondylar breadth Capitulum-trochlea breadth Minimum diameter of diaphysis Radius Maximum length Midshaft sagittal diameter Midshaft transverse diameter Maximum shaft diameter at the radial tuberosity Maximum shaft diameter distal to the radial tuberosity Minimum shaft diameter distal to the radial tuberosity Ulna Maximum length Dorso-volar diameter taken perpendicular to the transverse diameter at the same position along the diaphysis Transverse diameter at point of maximum expression of the interosseous crest Minimum diameter of the diaphysis along the portion of the bone that includes the interosseous crest 39 RECORDING SHEETS Element Measurement Femur Maximum length Left Right Left Right Epicondylar breadth Maximum head diameter Anterior-posterior subtrochlear diameter Transverse subtrochlear diameter Tibia Maximum length Maximum breadth of proximal epiphysis Maximum breadth of distal epiphysis Maximum diameter at the nutrient foramen Transverse diameter at the nutrient foramen Minimum anterior-posterior diameter of the shaft Fibula Maximum length Maximum midshaft diameter MEASUREMENTS FOR OSTEOMETRIC ARTICULATION Key: Record following Byrd and LeGarde (2014) Element Measurement Scapula Glenoid fossa max breath Humerus Head A-P breath Capitulum-trochlea breadth Radius Head max diameter Ulna Breadth at distal end of semi-lunar notch Os coxa Max diameter of acetabulum Femur Max head diameter Epicondylar breadth Tibia Max breadth of proximal epiphysis Max breadth of distal epiphysis Talus 40 Min breadth of articular surface RECORDING SHEETS CRANIAL NONMETRIC TRAITS Key: Record as present/absent Trait Expression Trait Metopic suture Squamous ossicle Supranasal suture Frontotemporal articulation Supraorbital foramina Marginal tubercle Supraorbital notches Zygomatico-facial foramen Ethmoidal foramina Divided temporal squama Infraorbital foramina Divided zygomatic bone Zygomatico-facial foramina External auditory torus Zygomaxillary tubercle Squamomastoid suture Maxillary torus Parietal foramina Transverse palatine suture Ossicle at lambda Palatine torus Lambdoid ossicles Lesser palatine foramina Ossicle at asterion Foramen of Vesalius Occipitomastoid ossicle Oval foramen Mastoid foramen Spinous foramen Inca bone Divided occipital condyles Coronal ossicle Occipitomastoid ossicle Ossicle at bregma Divided parietal bone Sagittal ossicle Expression Parietal notch bone 41 RECORDING SHEETS MORPHOSCOPIC TRAITS DENTAL NONMETRIC TRAITS Key: Record based on Hefner (2009) Key: Record in an ordinal scale following the ASUDAS system Trait Expression Inferior nasal aperture Tooth Trait Anterior nasal spine Incisors Winging Shovel-shaped Nasal aperture width Double shoveling Nasal overgrowth Labial curvature Malar tubercle Interruption groove Nasal bone contour Tuberculum dentale Interorbital breadth Peg-shaped incisors Postbregmatic depression Canines Distal accessory ridge Supranasal suture Lower canine root number Transverse palatine suture Bushman canine Premolars Zygomaticomaxillary suture Odontome Upper premolar root number POSTCRANIAL NONMETRIC TRAITS Distosagittal ridge Key: Record as present/absent Tome’s root Lower premolar lingual cusp variation Element Trait Atlas Double atlas facet Cervical vertebrae Transverse foramen bipartite Sternum Sternal foramen Enamel extensions Scapula Bridging of suprascapular notch Hypocone Humerus Supracondyloid process Septal aperture Os coxa Acetabular crease Accessory sacral facets Femur Allen’s fossa Poirier’s facet Plaque Hypotrochanteric fossa Third trochanter Patella Vastus notch Emarginate patella Tibia Squatting facets Talus Medial talar facet Lateral talar extension Double inferior anterior talar facet Calcaneus 42 Double anterior calcaneal facet Expression Molars Carabelli’s trait Upper molar root number Metaconule Deflecting wrinkle Anterior fovea Tuberculum intermedium Tuberculum sextum Lower molar root number Hypoconulid Groove pattern Expression RECORDING SHEETS POST-MORTEM BONE ALTERATION Key: Record based on Fernández-Jalvo and Andrews (2016) Alteration Type Element(s) affected Possible etiology 43 appendix | RECORDING SHEETS ISBN 978-9963-2858-5-3