Critical review of the MNI (minimum number of individuals) as a zooarchaeological unit of quantification

The use of minimum number of individuals (MNI) in the analysis of mammal archaeofaunal assemblages has received intense criticism for its derived nature and its dependence of biasing variables. Some authors have argued that similar taxonomic variability as documented by MNI can be achieved with other less biased measuring units, such as number of identified specimens (NISP) and derivates thereof. The present study is the first experimentally controlled test that shows the degree of bias of estimates of MNI and NISP by different analysts. It shows that the margins of error of both measuring units are independent and that MNI can be more accurately estimated than NISP, despite its “derived” inferential nature.     

Testing the efficacy and reliability of common zooarchaeological sampling strategies: a case study from the Caribbean

Zooarchaeological analysis typically involves implementing a suitable sampling strategy that provides meaningful information representative of a site under investigation. To test the efficacy of how various sampling procedures can influence interpretation of a faunal assemblage, we analyzed in full, a large assemblage of mollusks (minimum number of individuals [MNI] = 58,000+) from the late Ceramic Age (ca. AD 850–1440) site of Coconut Walk on the island of Nevis (northern Lesser Antilles, Caribbean). These remains were recovered from 25 discrete 1 × 1 m units within a larger 5 × 5 m trench that overlaid a dense midden deposit. Using statistical procedures, we compared the results of the number of identified specimens (NISP), MNI, and weight (g) of each mollusk taxon between both units and plana to determine how different sampling strategies would have affected the interpretation of mollusk composition across space and time had different sampling procedures been used. The results reveal that sampling less than 16% (4 out of 25 squares for this study) of a midden is likely to produce results outside of a 95% confidence interval for the population composition, with samples of connected squares having a higher likelihood to lie outside the interval. Discerning change between plana (or layers) may require a larger sample (i.e., 16%+). In addition, samples selected randomly or judgmentally will produce similar results, provided that proper spacing is given between units. The data in this case suggest that the midden was efficiently sampled for midden composition and key species at a rate of 16% of the surface area.     

The shell game: analytic problems in archaeological mollusc quantification

MNI calculations for archaeological molluscs that rely on counts of non-repetitive elements (NREs) are popular for their analytic speed and efficiency. Yet the reliability of this approach as compared to ‘traditionally’ calculated MNI has not been explored. Here, I examine these approaches using data from the Caribbean coastal midden site of Grand Bay, Carriacou. Results suggest that an NRE-based approach to MNI calculation can be unreliable under certain circumstances. In addition, dependence on the NRE-based method may hinder the integration of invertebrate with vertebrate data, obscuring a holistic understanding of prehistoric diet and resource exploitation. These findings are discussed in relation to practical constraints and fulfillment of zooarchaeological goals.     

SHELL FRAGMENTATION AS A TOOL FOR QUANTIFICATION AND IDENTIFICATION OF TAPHONOMIC PROCESSES IN ARCHAEOMALACOLOGICAL ANALYSIS: THE CASE OF THE CANTABRIAN REGION (NORTHERN SPAIN)

This paper presents a quantification method for archaeological molluscs using fragmentation categories. By employing the data from nine prehistoric sites (13 000 to 5000 bp ; 15 700 to 5700 cal bp ) in Cantabrian Spain (northern Spain), it has been shown that traditional methods, based on one or two anatomical elements, systematically produce a loss in the minimum number of individuals (MNI) that is sometimes significant. This paper also proposes using fragmentation categories in the calculation of the fragmentation index. In addition, based on these indices, a reflection is made on the difficulty involved in associating the amount of fragmentation of the shells with the taphonomic processes that have occurred at the archaeological sites, especially in discerning which agents have caused the fragmentation.     

The use of statistical size estimations in minimum number calculations

Analyses of archaeological faunal assemblages often require the use of varied quantitative techniques for reconstructing aspects of the harvested populations. This is particularly important with the increasing use of zooarchaeological data outside of the realm of traditional archaeological inquiry. Detailed size estimations of archaeological faunal specimens can be used to distinguish between the remains of individual animals, aiding in the reconstruction of minimum number calculations. Statistical regression provides a particularly useful technique for generating size estimations which may then be used to calculate such modified MNI values. This paper outlines a methodology for utilising regression estimated sizes in MNI calculations, and tests the methodology in the context of a case study involving fish remains from Aleutian Islands archaeological sites. The use of this methodology to generate modified MNI values results in consistent increases over values calculated using traditional MNI approaches. Such an approach, while more costly in terms of required effort, is felt to be preferable to traditional approaches in contexts where detailed size estimations are usefully calculated to answer other research questions. Copyright © 2005 John Wiley & Sons, Ltd.     

The Efficiency of Flotation Compared with Other Methods for Recovering Assemblages of Terrestrial and Aquatic Gastropods from Archaeological Deposits,with Reference to the Site of Pico Ramos (Basque Country,Spain)

ABSTRACT

We analysed assemblages of shells of land snails and small aquatic and wetland gastropods from a late Mesolithic-early Neolithic deposit in the cave of Pico Ramos (Basque Country). Shells were recovered by hand during excavation and all excavated deposit was processed by flotation, with a 250 μm mesh to recover floating material and a 1.0 mm mesh to recover the heavy fraction. The total assemblage comprised 5780 individuals (MNI) among 33 taxa, the sieved sub-assemblage having the highest proportions of these (2841 MNI in 31 taxa) and the hand-picked sub-assemblage the lowest (698 MNI in 11 taxa). Eleven taxa were absent from the flotation sub-assemblage, although the recovered MNI (2241; 39% of the total MNI) was high. The palaeoecological implications of the results are considered. Adult and large-sized juvenile shells of the edible land snail Cepaea nemoralis are abundant in the hand-picked sub-assemblage, but it is the absence of small-sized juvenile shells in the flotation and wet-sieved sub-assemblages that permits the inference that the species was collected for consumption. Wet sieving is therefore essential, both for interpreting hand-picked samples and in particular for the recovery of representative assemblages of land molluscs.     

Archaeological Quantification of Pottery: The Rims Count Adjusted using the Modulus of Rupture (MR)

Archaeological quantification is a recurrent issue in research about pottery, its typologies and its distribution. We accept the validity of other methods of quantification—sherd count, minimum number of individuals (MNI) or sherd weight—but the methodology that we have proposed for quantification of assemblages of archaeological contexts is the rims count, which has to be transformed into coefficients of reference through a correction using the modulus of rupture (MR). Such correctors are obtained through measuring the percentage of preserved rim of a significant number of sherds of each type and establishing the average of that percentage. This quantification method is easily applicable to all pottery types and it is also statistically reliable. Besides, it can be used in any study in which the gross number of rims is published. Finally, in the case of ceramic transport containers, a second correction can be applied by multiplying the corrected coefficient (number of rims × MR) by its average capacity (AC), another corrector that will allow us to gather statistics according to the litres of transported product. We believe that the rims count (the easiest part to classify) is a fast, relatively easy and very reliable method that needs to be corrected using the MR.     

The Ethnoarchaeology of Reptile Remains at a Lake Turkana Occupation Site, Kenya

The role of reptiles in hominid hunter–gatherer subsistence in Pleistocene Africa has been largely overlooked. This study examines the reptile component of a modern lake shore camp (site 20) assemblage of low archaeological visibility. Site 20, located on the eastern shore of Lake Turkana, has been observed from its creation to burial. The site is an ideal modern analogue for the interpretation of African prehistoric archaeological sites. Patterns of element loss, and patterns of bone modifications, namely burn, cut and slice marks, are examined. From this baseline data, three conclusions are drawn regarding the interpretability of reptile remains found at archaeological lakeside sites: (1) the MNI of crocodile remains at a site is likely to be similar to the original number of individuals which contributed to the deposited assemblage; (2) the MNI estimates of the turtle component will be much less than the original number of individuals in the deposited assemblage; and (3) patterns of bone modifications can be related to hominid food procurement and processing activities.     

Comparing methods for analysing mortality profiles in zooarchaeological and palaeontological samples

In this study, I examine three methods that are currently used for comparing mortality profiles from zooarchaeological and palaeontological samples: (1) histograms with 10% of life‐span age classes; (2) boxplots showing tooth crown height medians; and (3) triangular plots of the proportions of young, prime and old animals. I assess the advantages and disadvantages of each method using data collected on two samples of Northern Yellowstone elk (Cervus elaphus nelsoni) with known, or cementum annuli‐determined, ages at death. One sample was hunted by wolves (n = 96), and the other was hunted by recent humans using rifles (n = 226). I tested each method with the known or cementum annuli age distributions and with age estimation techniques appropriate for archaeological assemblages. Histograms are best used when the relationship between dental eruption/attrition and age is well established so that individuals can be confidently assigned into 10% of life‐span groups, and when more than 30 or 40 individuals are present in the assemblage. Boxplots employ raw crown heights, thus removing the error introduced by assigning specimens to age classes, and therefore they allow the analysis of species where the relationship between dental eruption/attrition and age is unknown. Confidence intervals around the medians allow samples to be statistically compared. Triangular plots are easy to use and allow multiple samples and species to be considered simultaneously, but samples cannot be statistically compared. Modified triangular plots bootstrap samples to provide 95% confidence ellipses, allowing for statistical comparisons between samples. When possible, samples should be examined using multiple methods to increase confidence in the results. Copyright © 2005 John Wiley & Sons, Ltd.     

Number estimation in vertebrate samples

Methods of quantifying bone samples are discussed and the minimum numbers approach criticized. A technique based upon number estimation of game populations is put forward, and some of the practical difficulties of this method are examined. Consideration is given to the underlying statistical assumptions involved in quantified analysis of bone samples, and a method of calculating confidence limits for death population numbers proposed.     

Evaluating the Economic Significance of Sharks,Skates, and Rays (Elasmobranchs) in Prehistoric Economies

Establishing the significance of elasmobranchs in ancient economies is complicated by a variety of biological and taphonomic problems that hinder NISP, MNI, and weight measures and comparisons with other fish or faunal classes. To help address these problems, we present length and live weight estimates for 53 elasmobranch specimens, along with bone weight to edible meat weight ratios for 38 specimens. We argue that quantification is best accomplished using NISP and MNI measures in conjunction with meat weight estimates obtained from modern specimens. These data illustrate the importance of using multiple quantitative measures (i.e., NISP, MNI, weight, etc.) when evaluating the economic significance of various taxa within faunal assemblages.     

A Simulation Study on Specifying a Regression Model for Spatial Data: Choosing between Autocorrelation and Heterogeneity Effects

In this simulation study, regressions specified with autocorrelation effects are compared against those with relationship heterogeneity effects, and in doing so, provides guidance on their use. Regressions investigated are: (1) multiple linear regression, (2) a simultaneous autoregressive error model, and (3) geographically weighted regression. The first is nonspatial and acts as a control, the second accounts for stationary spatial autocorrelation via the error term, while the third captures spatial heterogeneity through the modeling of nonstationary relationships between the response and predictor variables. The geostatistical‐based simulation experiment generates data and coefficients with known multivariate spatial properties, all within an area‐unit spatial setting. Spatial autocorrelation and spatial heterogeneity effects are varied and accounted for. On fitting the regressions, that each have different assumptions and objectives, to very different geographical processes, valuable insights to their likely performance are uncovered. Results objectively confirm an inherent interrelationship between autocorrelation and heterogeneity, that results in an identification problem when choosing one regression over another. Given this, recommendations on the use and implementation of these spatial regressions are suggested, where knowledge of the properties of real study data and the analytical questions being posed are paramount.     

Designs for Detecting Spatial Dependence

The aim of this article is to find optimal or nearly optimal designs for experiments to detect spatial dependence that might be in the data. The questions to be answered are: how to optimally select predictor values to detect the spatial structure (if it is existent) and how to avoid to spuriously detect spatial dependence if there is no such structure. The starting point of this analysis involves two different linear regression models: (1) an ordinary linear regression model with i.i.d. error terms—the nonspatial case and (2) a regression model with a spatially autocorrelated error term, a so-called simultaneous spatial autoregressive error model. The procedure can be divided into two main parts: The first is use of an exchange algorithm to find the optimal design for the respective data collection process; for its evaluation an artificial data set was generated and used. The second is estimation of the parameters of the regression model and calculation of Moran's I , which is used as an indicator for spatial dependence in the data set. The method is illustrated by applying it to a well-known case study in spatial analysis.     

THE ASSESSMENT OF ERROR LIMITS IN THERMOLUMINESCENT DATING

This paper discusses known sources of error, both random and systematic; suggests quantitative limits; and proposes a consistent method for compounding them to yield an overall error in age, both for individual sherds and for the weighted average of a number of sherds from the same context. It is proposed that two error limits should be quoted for the date of a context: first, the standard error of the mean value as calculated from the experimentally observed root-mean-square deviation of the individual samples, and secondly, the expected overall error from the sources discussed in the paper. The former can be considered as a lower limit and relevant when comparing contexts on the same or similar sites whereas the latter is more realistic when comparing TL dates with radiocarbon or archaeologic chronology.     

A Measurement Theory for Time Geography

Hägerstrand's time geography is a powerful conceptual framework for understanding constraints on human activity participation in space and time. However, rigorous, analytical definitions of basic time geography entities and relationships do not exist. This limits abilities to make statements about error and uncertainty in time geographic measurement and analysis. It also compromises comparison among different time geographic analyses and the development of standard time geographic computational tools. The time geographic measurement theory in this article consists of analytical formulations for basic time geography entities and relations, specifically, the space–time path, prism, composite path-prisms, stations, bundling, and intersections. The definitions have arbitrary spatial and temporal resolutions and are explicit with respect to informational assumptions: there are clear distinctions between measured and inferred components of each entity or relation. They are also general to n- dimensional space rather than the strict two-dimensional space of classical time geography. Algebraic solutions are available for one or two spatial dimensions, while numeric (but tractable) solutions are required for some entities and relations in higher dimensional space.     

MULTIVARIATE MIXTURE OF NORMALS WITH UNKNOWN NUMBER OF COMPONENTS: AN APPLICATION TO CLUSTER NEOLITHIC CERAMICS FROM AEGEAN AND ASIA MINOR USING PORTABLE XRF*

Multivariate techniques and especially cluster analysis have been commonly used in archaeometry. Exploratory and model‐based techniques of clustering have been applied to geochemical (continuous) data of archaeological artefacts for provenance studies. Model‐based clustering techniques such as classification maximum likelihood and mixture maximum likelihood have been used to a lesser extent in this context and, although they seem to be suitable for such data, they either present practical difficulties—such as high dimensionality of the data—or their performance gives no evidence that they are superior to standard methods. In this paper standard statistical methods (hierarchical clustering, principal components analysis) and the recently developed model‐based multivariate mixture of normals with an unknown number of components, are applied and compared. The data set provides chemical compositions of 188 ceramic samples derived from the Aegean islands and surrounding areas.     

A Study on the Elemental Composition of Chinese Mise Type Wares from Different Periods and Kilns

A large number of porcelain shards were unearthed at the Hehuaxin and Silongkou kilns around the Shanglin Lake, Ningbo, China, some of which had clear stratification to enable the determination of the date of firing. With the purpose of studying the elemental composition in different cultural periods and between kilns, the chemical compositions of 43 typical shards from these two kilns were determined by energy‐dispersive X‐ray fluorescence (EDXRF). The results show that the mean contents of K₂O, CaO and MnO in the glaze fluctuate in different cultural periods at the Hehuaxin kiln, but not systematically. Principal component analysis (PCA) indicates that samples from the Five Dynasties and the Northern Song dynasty are similar to each other. This shows continuity in the raw materials and firing technology between the two dynasties. Comparing the mise type wares from the Hehuaxin and Silongkou kilns in the Northern Song dynasty, we found that MnO and ZrO₂ in the glaze can be used as fingerprinting elements to distinguish between samples from these two sites. In spite of this, their chemical compositions are broadly similar. Thus, perhaps, their chemical compositions can be taken as typical for those of the mise type wares produced around the Shanglin Lake.     

A Synthesis of Aggregation Methods for Multifacility Location Problems: Strategies for Containing Error

When solving a location problem using aggregated units to represent demand, it is well known that the process of aggregation introduces error. Research has focussed on individual components of error, with little work on identifying and controlling total error. We provide a focussed review of some of this literature and suggest a strategy for controlling total error. Consideration of alternative criteria for evaluating aggregation schemes shows that the method selected should be compatible with the objectives of the analyses in which it is used. Experiments are described that show that two different measures of error are related in a nonlinear way to the number of aggregate demand points (q), for any value of the number of facilities (p). We focus on the parameter q/p and show that it is critical for determining the expected severity of the error. Many practical implementations of location algorithms operate within the range of q/p where the rate of change of error with respect to q/p is highest.     

四川大邑县高山古城遗址出土新石器时代动物骨骼遗存

通过对四川大邑县高山古城遗址2015、2016年出土动物骨骼遗存做了比较详细的鉴定、测量、统计和分析,可以肯定为家畜的仅有猪和狗。从出土动物骨骼的NISP、MNI统计来看,家猪的比重均占绝对优势,NISP占比达93.78%,MNI占比为86.49%;家猪上、下第三臼齿的长度、宽度较小,均处于家猪第三臼齿的长、宽参数范围内;从猪的年龄结构来看,以月龄10~20个月为主,也与一般农耕聚落家猪的宰杀年龄结构相符。鉴于以上三个方面的观察,可以推断高山古城先民饲养家猪的技术比较成熟且稳定,饲养家猪也是先民肉食资源最主要的获取途径。而狗所占的比重不高,占NISP、MNI的比重分别为3.76%和5.41%,可能多用做狩猎伴侣或其他。结果表明宝墩文化时期的高山古城先民以饲养家猪作为获取肉食资源的主要途径,当时的家畜饲养技术比较成熟稳定,而稻-粟混作的农业体系无疑为家畜的饲养奠定了坚实的基础。     

Estimating meagre (Argyrosomus regius) size from otoliths and vertebrae

This study presents a method for predicting meagre (Argyrosomus regius) body size (total length) from otoliths and vertebrae recovered from archaeological sites. The method involves regression equations calculated from a reference collection of 36 meagre skeletons and 113 meagre otoliths (sagitta) and allows the simultaneous estimation of original body size and minimum number of individuals (MNI) from archaeological bone structures. We selected the following measurements to predict meagre body size: greatest dorso-ventral height, greatest mediolateral breadth, and greatest anteroposterior length of the vertebrae centra; maximum anteroposterior length, medial anteroposterior length, and dorso-ventral height of the sagitta. Our results show that the original body size of meagre can be accurately predicted from many bone measurements (r² range: 0.921–0.992). We exemplify the use of the regressions in the assessment of size variation and MNI of meagre from four Portuguese Mesolithic sites. We show that regression results provide additional insight into the significant role that this fish played in the subsistence of coastal fisher-hunter-gatherers, who targeted medium-sized animals but were also capable of acquiring rather larger specimens.