Multivariate analysis in biological anthropology: Some considerations

The relation between modern multivariate statistical analysis and common objectives in biological anthropology is discussed. Many formal statistical assumptions should be satisfied to fully justify multi-variate analysis, but rarely are. Their failure does not have as adverse an effect upon scientific research as some procedural errors in research design. These include the treatment of single specimens as if they are samples, and the failure to separate considerations of size and shape. The superiority of elaborate multivariate distances over simpler quantitative techniques, inferred from the former's consideration of trait intercorrelations, is unwarranted from both theoretical and practical aspects. Based on these points and several comparative examples of analyses, it is argued that simpler, non-statistical approaches such as those of the school of numerical taxonomy have much to offer primato-logical studies. Classic multivariate statistics, while useful for studying closely similar living populations, are considerably more limited in usefulness and significance for paleoanthropological studies.     

Comparing Vectors of Adjusted Means when Covariance Matrices are Unequal

The ability to distinguish between stocks of fish is of vital importance in fisheries management programs. Morphometric measurements are commonly employed to separate biological populations. Unique role of multivariate analysis of covariance in separating fish stocks has been recognized recently. However, fish populations frequently differ in their covariance matrices and sometimes also in vectors of slope (on the covariate) coefficients. In this article the following are presented: (a) a procedure for comparing adjusted (for variations in body sizes of specimens) mean vectors of populations with unequal covariance matrices and slopes, and (b) a study to distinguish between beaked redfishes based on the analysis of a real data set by this procedure. This will also serve as a numerical example.     

Variability in cuticular hydrocarbons and phenotypic discrimination of Ixodes ricinus populations (Acarina: Ixodidae) from Europe

The cuticular hydrocarbon composition and a stepwise discriminant analysis are used to elucidate the phenotypic relationships of 66 populations of Ixodes ricinus in Europe. The method correctly allocates Ixodes persulcatus (outgroup) populations away from the main cluster of I. ricinus samples and separates the samples into ten relatively defined clusters of specimens. Populations from Poland are inseparable from samples collected in Germany, Switzerland and the Italian Alps, while individuals from Slovakia and the Czech Republic come into separate groups of phenotypic similarity. Irish and British specimens are separated but highly related and Spanish populations show an unexpectedly high distance from the remaining clusters.     

The position of the Nazlet Khater specimen among prehistoric and modern African and Levantine populations

The morphometric affinities of the 33,000 year old skeleton from Nazlet Khater, Upper Egypt are examined using multivariate statistical procedures. In the first part, principal components analysis is performed on a dataset of mandible dimensions of 220 fossils, sub-fossils and modern specimens, ranging in time from the Late Pleistocene to recent and restricted in space to the African continent and Southern Levant. In the second part, mean measurements for various prehistoric and modern African and Levantine populations are incorporated in the statistical analysis. Subsequently, differences between male and female means are examined for some of the modern and prehistoric populations. The results indicate a strong association between some of the sub-Saharan Middle Stone Age (MSA) specimens, and the Nazlet Khater mandible. Furthermore, the results suggest that variability between African populations during the Neolithic and Protohistoric periods was more pronounced than the range of variability observed among recent African and Levantine populations. Results also demonstrate a general reduction in the degree of sexual dimorphism during the Holocene. However, this pattern of reduction pattern varies by geographic location and is not uniform across the African continent.     

Intraspecies variability of chaetotactic characteristics of the red mites of Hirsutiella genus (Acari: Trombiculidae)

Numbers of scutal, coxal and sternal setae were counted in 1393 specimens of 5 Hirsutiella species. All observed chaetotactic anomalies were classified into 3 groups: 1) "rare" anomalies, such as absence of one AL, AM, coxal or sternal setae, presence of additional AL, AM, coxal or anterior sternal setae; 2) presence of 1-2 post-posterolateral soutal setae (PPL); 3) presence of additional 1-3 posterior sternal setae. "Rare" anomalies were found in 2.8% of the specimens examined. They were observed usually in northern or alpine populations. Obviously, it is an effect of more frequent ontogenetic failures in rigorous climate conditions being out of optimum for trombiculids. The quota of the individuals with PPL in a set of H. steineri samples varied from 16 to 74%, while in the other samples of this species the presence of the single (and not more) PPL were found very seldom, as well as in the other species examined. In the similar way, the additional posterior sternal setae were found with the frequency up to 100% in the some samples of H. steineri, H. hexasternalis and H. alpina, but in other ones they were present only in few specimens. Since in a series of cases the ratio of the number of specimens with PPL to the number of specimens without PPL, as well as the ratio of the number of specimens with fSt = 2.4 to the number of specimens with fSt = 2.2 is almost equal to 2, one can suppose that the variance of this characters has a genetic basis. All samples having a high frequency of the PPL appearance or the appearance of additional posterior sternal setae were collected in the alpine zone. It is an evident example of the ecogeographical rules, which were reported previously by the author for the quantitative characters in species of the genera Neotrombicula (Stekolnikov, 1996, 1998, 1999) and Hirsutiella (Stekolnikov, 2001). It is possible that the PPL appearance is caused by the higher humidity of the climate, and the appearance of additional posterior sternal setae--by the low temperature. It is supported by the fact that the samples with high friquency of individuals having PPL are specific to damp areas of Western Caucasus, while the populations with high frequency of additional posterior sternal setae appear to be characteristic for regions with more dry climate (Daghestan, Turkey, Chelyabinsk Region of Russia). The fact, that in some populations of H. steineri, the individuals having PPL are prevalent, can serve as a serious argument for the abolishment of the genera Hoffmannina Brennan et Jones, 1959 and Aboriginesia Kudryashova, 1993 distinguished from related groups only by the presence of this setae.     

Is a multivariate consensus representation of genetic relationships among populations always meaningful?

To determine the relationships among closely related populations or species, two methods are commonly used in the literature: phylogenetic reconstruction or multivariate analysis. The aim of this article is to assess the reliability of multivariate analysis. We describe a method that is based on principal component analysis and Mantel correlations, using a two-step process: The first step consists of a single-marker analysis and the second step tests if each marker reveals the same typology concerning population differentiation. We conclude that if single markers are not congruent, the compromise structure is not meaningful. Our model is not based on any particular mutation process and it can be applied to most of the commonly used genetic markers. This method is also useful to determine the contribution of each marker to the typology of populations. We test whether our method is efficient with two real data sets based on microsatellite markers. Our analysis suggests that for closely related populations, it is not always possible to accept the hypothesis that an increase in the number of markers will increase the reliability of the typology analysis.     

Morphological differentiation of introduced pikeperch (Sander lucioperca L., 1758) populations in Tunisian freshwaters

In order to evaluate the phenotypic plasticity of introduced pikeperch populations in Tunisia, the intra‐ and interpopulation differentiation was analysed using a biometric approach. Thus, nine meristic counts and 23 morphological measurements were taken from 574 specimens collected from three dams and a hill lake. The univariate (anova ) and multivariate analyses (PCA and DFA) showed a low meristic variability between the pikeperch samples and a segregated pikeperch group from the Sidi Salem dam which displayed a high distance between mouth and pectoral fin and a high antedorsal distance. In addition, the Korba hill lake population seemed to have more important values of total length, eye diameter, maximum body height and a higher distance between mouth and operculum than the other populations. However, the most accentuated segregation was found in the Lebna sample where the individuals were characterized by high snout length, body thickness, pectoral fin length, maximum body height and distance between mouth and operculum. This study shows the existence of morphological differentiations between populations derived from a single gene pool that have been isolated in separated sites for several decades although in relatively similar environments.     

Another look at shape variation in the distal femur of Australopithecus afarensis: implications for taxonomic and functional diversity at Hadar

Previous studies have recognized two patterns of distal femoral morphology among the specimens from Hadar (Ethiopia) assigned to Australopithecus afarensis. Size and shape differences between the well-preserved large (AL 333-4) and small (AL 129-1a) distal femora have been used to invoke both taxonomic and functional differences within the A. afarensis hypodigm. Nevertheless, prior studies have not analyzed these specimens in a multivariate context, nor have they compared the pattern of shape differences between the fossils to patterns of sexual dimorphism among extant taxa (i.e., the manner in which males and females differ). This study reexamines morphometric differences between the above specimens in light of observed levels of variation and patterns of sexual dimorphism among extant hominoids. Eight extant reference populations were sampled to provide a standard by which to consider size and shape differences between the fossils. Samples include three populations of modern humans, two subspecies of Pan troglodytes, three subspecies of Gorilla gorilla, Pan paniscus, and Pongo pygmaeus. Using size ratios and scale-free "shape" data (both derived from 2-D coordinate landmarks), size and shape differences between the fossils were evaluated against variation within each reference population using an exact randomization procedure. Growth Difference Matrix Analysis (GDMA) was used to test whether the pattern of morphological differences between the fossils differs significantly from patterns of sexual dimorphism observed among the ten extant groups. Overall morphometric affinities of the fossils to extant taxa were explored using canonical variates analysis (CVA).Results of the randomization tests indicate that the size difference between the Hadar femora can be easily accommodated within most hominoid taxa at the subspecific level (though not within single-sex samples). In addition, the magnitude of shape differences between the fossils can be commonly sampled even within most single-sex samples of a single hominoid subspecies. The pattern of morphological differences between the fossils does not differ statistically from any average pattern of femoral shape dimorphism observed among living hominoids. Moreover, contrary to prior claims, and despite a size disparity between the fossils greater than is typically observed within some chimpanzee and human populations, the two Hadar fossils appear to be much more similar to one another in overall shape than either specimen is to any extant hominoid group.     

Assessing the affinities of fossils using canonical variates and generalized distances

Canonical variate analysis and generalized distances are commonly used multivariate statistical techniques for assessing the comparative morphology of living and fossil primates. Some common pitfalls of these methods when used to analyze fossil specimens are: (1) ignoring the possibility that a fossil belongs to a group other than one of the predefined reference samples (i.e., restricted versus unrestricted approaches to classification), (2) misinterpreting probabilities of group membership (i.e., posterior versus typicality probabilities), and (3) failing to understand how sample sizes influence multivariate ordinations in trying to effectively illustrate the morphometric affinities of a fossil (i.e., weighted versus unweighted analyses with fossils entered indirectly or directly into the analysis). To better understand canonincal variate analysis and generalized distances, the workings of these methods are portrayed graphically as a series of rotations and rescaling of data plotted in a multivariate data space. The influence of sample sizes and the importance of higher-numberer canonical variates are emphasized. Simple examples taken from the literature illustrate how the different approaches to including a fosil in canonical variate analysis may affect the multivariate results upon which physical anthropologists base their interpretations of the fossil's morphology.     

A multivariate morphometric study of the Cardamine amara group (Cruciferae) in the Carpathian and Sudeten mountains

MARHOLD, K., 1992. A multivariate morphometric study of the Cardamine amara group (Cruciferae) in the Carpathian and Sudeten mountains. A multivariate morphometric study of the Cardamine amara group based on 55 population samples, collected throughout the Carpathian and Sudeten mountains in Czechoslovakia, Poland, Ukraine and Romania, is presented. Among the multivariate methods, principal component analysis, cluster analysis, classificatory, canonical and stepwise discriminant analysis were used. The analysis of chromosome numbers from all studied populations showed the same diploid chromosome number: 2n = 16. Together with the results of multivariate morphometric analysis this led to the conclusion that the best way of classifying this group in the area studied was to divide it into two subspecies– Cardamine amara subsp. amara and C. amara subsp. opicii. Tetraploid populations of this group occurring in the Alps require further study.     

Comparison of molecular fingerprinting methods for analysis of soil microbial community structure

Denaturing gradient gel electrophoresis (DGGE), terminal-restriction fragment length polymorphism (T-RFLP) analysis, and automated ribosomal intergenic spacer analysis (ARISA) have been widely used as molecular fingerprinting methods for analysis of microbial communities. To find suitable methods, we compared the three fingerprinting methods by analyzing soil fungal communities in four differing land-use types: bare ground, crop fields, grasslands, and forests. We also examined optimal primer pairs for DGGE analysis by comparing single and mixed DNA samples of cultured fungal populations. Principal coordinate analysis (PCO), nonmetric multidimensional scaling method (NMDS), and analysis of similarities (ANOSIM), which are major multivariate statistical analyses for quantifying fingerprint patterns, were compared. All three fingerprinting methods yielded clear discrimination of soil fungal communities among the four land-use types, irrespective of statistical methods. The advantages and disadvantages of the three fingerprinting methods were discussed.     

Craniometric variation in the Eurasian badger, Meles meles

A multivariate examination of cranial variation within and between European populations of Meles meles (L.) revealed that populations from Ireland, Great Britain, Norway, and the Slovak Republic could be differentiated both by cranial form and by the degree of sexual dimorphism exhibited. Irish material was characterized by low sexual dimorphism, particularly when compared to Slovak specimens. Badgers from the British Isles had larger skulls than other samples and were more similar to each other than they were to badgers from mainland Europe. Size played a greater role in differentiating samples of female badgers than it did in males. Significant variation occurred within the British Isles, with individual samples being highly differentiable. There was, however, little relationship between morphological similarity and geographic proximity. We contend that macrogeographical (between-country) variation in the species is primarily determined by historical factors and adaptation to current conditions, while microgeographic (within-country) variation is a result of selectively neutral processes.     

Why Ecologists Need Systematics: Importance of Systematics to Ecological Research

Ecologists are concerned with population dynamics of organisms and with the spatial patterns of single or multiple populations. The goal of the ecologist is usually to explain the observed patterns in terms of processes. Field samples of nematodes from different habitats may contain similar but not identical specimens of a nominal taxon, and the systematist can help the ecologist decide whether the specimens are ecophenotypes of a single taxon or represent distinct species. A correct decision may be important or trivial, depending on the parameters and goals of the ecological study. When a precise identification is crucial to the success of the study, new biochemical methodologies of systematists may provide rapid and accurate diagnoses. Systematists can provide additional help in the assignment of taxa to trophic groups. For clarifying host-parasite associations, often a goal in ecological investigations, modern analytical methods of systematists can facilitate the ordering of systematic relationships.     

Extensive variation in intracellular symbiont community composition among members of a single population of the wood-boring bivalve Lyrodus pedicellatus (Bivalvia: Teredinidae)

Shipworms (wood-boring bivalves of the family Teredinidae) harbor in their gills intracellular bacterial symbionts thought to produce enzymes that enable the host to consume cellulose as its primary carbon source. Recently, it was demonstrated that multiple genetically distinct symbiont populations coexist within one shipworm species, Lyrodus pedicellatus. Here we explore the extent to which symbiont communities vary among individuals of this species by quantitatively examining the diversity, abundance, and pattern of occurrence of symbiont ribotypes (unique 16S rRNA sequence types) among specimens drawn from a single laboratory-reared population. A total of 18 ribotypes were identified in two clone libraries generated from gill tissue of (i) a single specimen and (ii) four pooled specimens. Phylogenetic analysis assigned all of the ribotypes to a unique clade within the gamma subgroup of proteobacteria which contained at least five well-supported internal clades (phylotypes). By competitive quantitative PCR and constant denaturant capillary electrophoresis, we estimated the number and abundance of symbiont phylotypes in gill samples of 13 individual shipworm specimens. Phylotype composition varied greatly; however, in all specimens the numerically dominant symbiont belonged to one of two nearly mutually exclusive phylotypes, each of which was detected with similar frequencies among specimens. A third phylotype, containing the culturable symbiont Teredinibacter turnerae, was identified in nearly all specimens, and two additional phylotypes were observed more sporadically. Such extensive variation in ribotype and phylotype composition among host specimens adds to a growing body of evidence that microbial endosymbiont populations may be both complex and dynamic and suggests that such genetic variation should be evaluated with regard to physiological and ecological differentiation.     

Preparation of uniform, intact DNA samples from resected tumor tissues for pulsed-field gel electrophoretic analyses.

Preparation of high molecular weight DNA from resected tumor tissues suitable for pulsed-field gel electrophoresis (PFGE) can be complicated by the presence of nonviable cells and lymphocytes. We have developed a simple procedure to reduce the level of degraded DNA in PFGE DNA samples prepared from resected tumor tissues. The procedure employs a single, three component Percoll step gradient centrifugation and can be performed on several tumor samples simultaneously. Analyses of DNAs from 15 tumor specimens (7 solid tumors and 8 aspirated fluids) demonstrate that the technique enriches the integrity of PFGE DNA samples. Morphologic evaluation of 9 specimens suggested that both cellular debris and contaminating normal lymphocytes are removed from starting cell populations during the enrichment procedure. Fractionation of cells also reduced cell clumping, allowing for the formation of more uniform PFGE DNA samples.     

Historical museum collections clarify the evolutionary history of cryptic species radiation in the world's largest amphibians

Inaccurate taxonomic assessment of threatened populations can hinder conservation prioritization and management, with human‐mediated population movements obscuring biogeographic patterns and confounding reconstructions of evolutionary history. Giant salamanders were formerly distributed widely across China, and are interpreted as a single species, Andrias davidianus. Previous phylogenetic studies have identified distinct Chinese giant salamander lineages but were unable to associate these consistently with different landscapes, probably because population structure has been modified by human‐mediated translocations for recent commercial farming. We investigated the evolutionary history and relationships of allopatric Chinese giant salamander populations with Next‐Generation Sequencing methods, using historical museum specimens and late 20th‐century samples, and retrieved partial or near‐complete mitogenomes for 17 individuals. Samples from populations unlikely to have been affected by translocations form three clades from separate regions of China, spatially congruent with isolation by either major river drainages or mountain ranges. Pliocene–Pleistocene divergences for these clades are consistent with topographic modification of southern China associated with uplift of the Qinghai‐Tibet Plateau. General Mixed Yule Coalescent model analysis indicates that these clades represent separate species: Andrias davidianus (Blanchard, 1871) (northern Yangtze/Sichuan), Andrias sligoi (Boulenger, 1924) (Pearl/Nanling), and an undescribed species (Huangshan). Andrias sligoi is possibly the world's largest amphibian. Inclusion of additional reportedly wild samples from areas of known giant salamander exploitation and movement leads to increasing loss of biogeographic signal. Wild Chinese giant salamander populations are now critically depleted or extirpated, and conservation actions should be updated to recognize the existence of multiple species.     

基于诊断引物的赤眼蜂鉴定和检测技术

通过对松毛虫赤眼蜂Tichogramma dendrolimi Matsumura和玉米螟赤眼蜂T.ostriniae Pang et Chen（膜翅目：赤眼蜂科）核内可转录第二间隔区（简称：ITS2）的克隆、测序,并获取和分析了GenBank中已登录的同源序列,然后设计了松毛虫赤蜂的特异引物以用于该蜂的分子鉴定和检测,经过反复筛选发现：采用鉴定引物通过PCR扩增不仅可以区分鑫头样品,单头样品（雌蜂或雄峰）,而且可鉴定幼期虫和卵,这用传统方法是无法办到的。该鉴定技术比基于形态学鉴定检测技术用来鉴定了从中国大陆不同地域和寄主上采集到的12个样品,结果表明：该方法可用于赤眼蜂田间分子监测和实验室拟寄生行为研究。     

PCR-BASED TECHNIQUE FOR IDENTIFICATION AND DETECTION OF TRICHOGRAMMA SPP. (HYMENOPTERA: TRICHOGRAMMATIDAE) WITH SPECIFIC PRIMERS

Abstract The rDNA-ITS2 regions of T. dendrolimi Matsumura and T. ostriniae Pang et Chen (Hymenoptera: Trichogrammatidae) were cloned and sequenced. The homologous sequences available in GenBank were retrieved and analyzed, and then specific primers were designed for molecular identification and detection of T. dendrolimi . Repeated screening showed that PCR amplification by the diagnostic primers enabled the differentiation of not only bulk samples and single adult (male or female), but also eggs and juveniles, which was not possible by conventional methods. The advantage of this system over morphology-based systems is that non-specialists are able to identify individuals or trace specimens efficiently. The derived molecular detection technique was then used to identify 12 specimens collected from different localities on the Chinese mainland; the results showed that this protocol could be applied to molecular monitoring of Trichogramma species in the field. Finally, 1132s of 6 geographical populations of T. dendrolimi (TdCHA, TDJL, TdXZ, TdKH, TdCZ and TdYBL) were cloned and sequenced. The multialignment analysis of intraspecific ITS2 sequences showed that the diagnostic primers have their own theoretical bases.     

Multivariate analyses of Canadian and Japanese populations ofGlycyphagus destructor (Schrank) (Acarina,Glycyphagidae)

Summary Two multivariate statistical procedures were used to determine the basic trends of morphologic and geographic variations between males of a common stored-grain mite,Glycyphagus destructor (Schrank) collected from Canada and Japan. All analyses were carried out on physical measurements of 25 morphological features. Three principal component analyses bases on the Canadian (50 specimens), Japanese (50 specimens) and the combined populations from the 2 countries (100 specimens) revealed that the first component, accounting for 40% of the variability in all 3 solutions represented a measure of the morphologic dimension of the body. The second component, which explained over 18% of the variability, is a measure of the leg dimension. Smaller clusters of variates characteristic of the population from each county were also evident. Discriminant analysis, performed with the Canadian and Japanese populations, identified the variates that differed between the 2 populations and provided an approximate appraisal of interrelations. The general conclusion based on these analyses is that the Canadian and the Japanese populations are morphologically distinct. The difference is most evident in the diameters of genu 2, lengths of the sensory seta W_I, lengths of the body, and the distances between the vertical external setae. Contribution No. 446, from Research Station, Canada Department of Agriculture, Winnipeg, Manitoba, Canada.