Canonical Variate Analysis Research Papers (original) (raw)

Exopolymers from a diverse collection of marine and freshwater bacteria were characterized by pyrolysis-mass spectrometry (Py-MS). Py-MS provides spectra of pyrolysis fragments that are characteristic of the original material. Analysis of the spectra by multivariate statistical techniques (principal component and canonical variate analysis) separated these exopolymers into distinct groups. Py-MS clearly distinguished characteristic fragments, which may be derived from components responsible for functional differences between polymers. The importance of these distinctions and the relevance of pyrolysis information to exopolysaccharide function in aquatic bacteria is discussed.

- by
- •
- Chemistry, Mass Spectrometry, Medicine, Multidisciplinary

Thirteen morphological traits were measured on a total of randomly selected 1672 female mature goats from four vegetation zones ranging from South towards North Benin. Four populations were characterized as distinct and may be referred to as ecotypes. The cluster analysis showed that the two southerly populations were closest to each other. Populations from northernmost vegetation zones always had larger features and less additional teats than those from southernmost zones. They also tended to have wattles and dropping ears. In the canonical discriminant analysis, the first two canonical variates were significant and accounted for 92% of the variability among vegetation zones. Measurements that were best able to separate the populations, as judged from the canonical variate analysis, were: height at withers, neck length, rump height, tail length and the ratio ear length/thoracic depth. Discriminant function models correctly allocated more than 70% of the individual goats to their a priori vegetation zones. These results provide supporting evidence for spatial variation in population of goats in Benin. They also underscore a need for further on-farm performance experiments that will lead to the characterization of these populations on a sound ecological basis and allow formulation of sustainable breeding and management strategies.

- by luc dossa
- •
- Animal Production, Discriminant Analysis, Cluster Analysis, Canonical Variate Analysis

Flash profile is a quick sensory profiling technique designed to meet industrial needs. It is based on the combination of free choice profiling and a comparative evaluation of the whole product set. This study aimed at comparing Flash profile with conventional profiling. This has been done with two product spaces: strawberry blended yoghurts and apricot ''fromages frais'', both from the French market. Multidimensional data treatment such as Generalised Procrustes Analysis and Canonical Variate Analysis (CVA) allowed us to assess the correspondence between the sensory maps obtained with both methods. In the case of strawberry yoghurts, very similar results were obtained with the two methods. On the other hand the sensory positioning of the apricot fresh cheeses somewhat differed between the methods. For both product sets, Flash profile was slightly more discriminating than the conventional profile. Advantages and limits of Flash profile are discussed. #

- by Julien Delarue
- •
- Marketing, Descriptive Analysis, Canonical Variate Analysis, Dairy Products

Partial least squares regression Partial least squares path modeling PLS Symmetric PLS Asymmetric PLS Task PLS Behavior PLS Seed PLS Multi-block PLS Multi-table PLS Canonical variate analysis Co-inertia analysis Multiple factor analysis... more

- by Anjali Krishnan
- •
- Algorithms, Experimental Design, Brain, Discriminant Analysis

The analysis of shape is a fundamental part of much biological research. As the field of statistics developed, so have the sophistication of the analysis of these types of data. This lead to multivariate morphometrics in which suites of measurements were analyzed together using canonical variates analysis, principal components analysis, and related methods. In the 1980s, a fundamental change began in the nature of the data gathered and analyzed. This change focused on the coordinates of landmarks and the geometric information about their relative positions. As a by-product of such an approach, results of multivariate analyses could be visualized as configurations of landmarks back in the original space of the organism rather than only as statistical scatter plots. This new approach, called "geometric morphometrics", had benefits that lead Rohlf and Marcus (1993) to proclaim a "revolution" in morphometrics. In this paper, we briefly update the discussion in that paper and summarize the advances in the ten years since the paper by Rohlf and Marcus. We also speculate on future directions in morphometric analysis.

- by F. James Rohlf
- •
- Zoology, Principal Component Analysis, Canonical Variate Analysis, Shape Analysis

Sulphur dioxide (SO 2 ) fumigation is a commonly adopted commercial treatment to retain the postharvest quality of litchi (Litchi chinensis Sonn.) fruit. The present investigations aim to overcome the undesirable after-effect of (SO 2 ) fumigation during long-term storage and distribution. The fruit of litchi cv. Mauritius, harvested during the early season, were dipped for 2 min at 8 • C in postharvest treatment solutions of ethylenediaminetetraacetic acid, calcium disodium salt hydrate (EDTA) (0.1%), phosphoric acid (0.1%) or 4-hexylresorcinol (0.1%). After treatment, fruit were packed in three types of biorientated polypropylene packaging: (1) BOPP-1, (2) BOPP-2 and (3) BOPP-3, heat-sealed to modify the atmosphere around the fruit and placed in commercial cardboard cartons. After treatment, fruit were held at 2 • C, 95% RH for 34 days to simulate refrigerated sea shipment conditions and thereafter, at market shelf conditions of 14 • C for 2 days. Standard commercial SO 2 fumigation was included as a comparative control. Weight loss, firmness, pericarp browning, severity of decay, Hunter colour values, titratable acidity and soluble solids concentrations were determined after cold storage followed by simulated market shelf conditions. The modified atmosphere created (17.0% O 2 and 6.0% CO 2 ) inside the BOPP-3 and the high RH around the fruit minimised the rate of transpiration, preventing weight loss and deterioration of fruit quality. Sensory analysis was carried out to determine pericarp and aril colour, firmness, flavour qualities, taste, odour and juiciness. Multivariate canonical variate analysis (CVA) of the results indicated that fruit packed in BOPP-3 retained colour and excellent eating qualities during long-term storage. The BOPP-3 packaging can be recommended as a safe, cost-effective alternative for extending the storage life of litchi cv. Mauritius during sea shipment.

- by Lise Korsten
- •
- Seasonality, Weight Loss, Cost effectiveness, Polypropylene
- by Christine Janis
- •
- Zoology, Functional Analysis, Late Miocene, Canonical Variate Analysis

Little is known about the effects of grazing on vegetation composition on the Arabian Peninsula. The aim of this study therefore was to analyse the vegetation response to environmental conditions of open woodlands along an altitudinal and a grazing gradient in the Jabal al Akhdar mountain range of Oman. The species composition, vegetation structure, grazing damage and several environmental variables were investigated for 62 samples using a nested plot design. Classification analysis and a Canonical Variate Analysis (CVA) were used to define vegetation types and to identify underlying environmental gradients. The relationship between environmental variables and diversity was analysed using correlation coefficients and a main-effects ANOVA. The plant species richness followed a unimodal distribution along the altitudinal gradient with the highest number of species at the intermediate altitudinal belt. The cluster analysis led to five vegetation groups: The Sideroxylon mascatense-Dodonaea viscosa group on grazed and the Olea europaea-Fingerhuthia africana group on ungrazed plateau sites at 2000 m a.s.l., the Ziziphus spina-christi-Nerium oleander group at wadi sites and the Moringa peregrina-Pteropyrum scoparium group at 1200 m a.s.l, and the Acacia gerrardii-Leucas inflata group at 1700 m a.s.l. The CVA indicated a clear distinction of the groups obtained by the agglomerative cluster analysis. The landform, altitude and grazing intensity were found to be the most important variables distinguishing between clusters. Overgrazing of the studied rangeland is an increasing environmental problem, whereas the plant composition at ungrazed sites pointed to a relatively fast and high regeneration potential of the local vegetation.

- by Annette Patzelt and +1
- •
- Earth Sciences, Arid environments, Biological Sciences, Environmental Sciences

Please cite this article as: Chope, G.A., Terry, L.A., Use of canonical variate analysis to differentiate onion cultivars by mineral content as measured by ICP-AES, Food Chemistry (2009), Use of canonical variate analysis to differentiate... more

- by Gemma Chope
- •
- Calcium, Food Chemistry, Multidisciplinary, Multivariate Analysis
- by Sara Morales-ojeda
- •
- Water, Water quality, Environmental Monitoring, Multidisciplinary

The gelling properties of pectins are related not only to the degree of esterification (DE), but also to the distribution of the ester groups. In this study, we have examined an experimentally designed series of 31 pectins originating from the same mother pectin and de-esterified using combinations of two different enzymatic mechanisms. The potential of using infrared (IR), Raman, and near infrared (NIR) spectroscopies combined with chemometrics for reliable and rapid determination of the DE and distribution patterns of methyl ester groups in a designed set of pectin powders was investigated. Quantitative calibration models using partial least squares (PLS) regression were developed and compared. The calibration models for prediction of DE obtained on extended inverse signal correction (EISC)-treated spectra of all three spectroscopic methods yielded models with cross-validated prediction errors (RMSECV) between 1.1%p and 1.6%p DE and correlation coefficients of 0.99. A calibration model predicting degree of random de-esterification (R) and block de-esterification (B) was developed for each spectroscopic method, yielding RMSECV values between 4.4 and 6.7 and correlation coefficients (r) between 0.79 and 0.92. Variable selection using interval PLS (iPLS) significantly improved the prediction of R for IR spectroscopy, yielding RMSECV of 3.5 and correlation coefficients of 0.95. All three spectroscopic methods were able to distinguish the spectral patterns of pectins with different enzyme treatments in simple classification models by principal component analysis (PCA). Extended canonical variate analysis revealed one specific signal in the Raman (1045 cm À1 ) spectrum and one significant area (1250-1400 cm À1 ) in the IR spectrum which are able to classify the pectin samples according to the four different enzyme treatments. In both Raman and IR spectra, the signal intensity decreased in the sequence R-B > B > B-R > R > re-methylated pectin.

A new species, formerly identified as P. maculatus La Cepède, 1803, is described. It differs from its congeners, except P. maculatus and P. mysteriosus Azpelicueta, 1998 by the color pattern with dark brown blotches in the flanks. The new species is morphometricaly similar to P. argenteus Perugia, 1891 and P. mysteriosus, differing, respectively, by the blotched color pattern and by the smaller maxillary barbel. The described new species differs from P. maculatus for presenting the supraoccipital process more robust, with base as broad as long (versus longer than broad); anterior nostrils 25% of distance between the margin of snout to posterior nostrils (versus 33% of this distance); small saws in the pectoral spine more developed and present in more than half of the anterior margin of spine (versus small saws present in less than half of that margin); very thin skin covering the head, allowing the grooves in the bones surfaces conspicuous; 22-27 (mode=24) gill rakers in the first branchial arch [versus 19-24 (mode=21)]. This new species was also discriminated from P. maculatus in the size free canonical variates analysis by presenting larger values of interorbital distance, mouth width and maxillar barbel length and smaller base length of adipose-fin.

- by Oscar Shibatta
- •
- Zoology, Canonical Variate Analysis

A modification of the standard Canonical Variates Analysis (CVA) method to cope with collinear high-dimensional data is developed. The method utilizes Partial Least Squares regression as an engine for solving an eigenvector problem involving singular covariance matrices. Three data sets are analyzed to demonstrate the properties of the method: a two-group problem with near infrared spectroscopic data consisting of 60 samples and 376 variables, a multi-group problem with fluorescence spectroscopic data (1023 variables) consisting of 83 samples from six groups and a three-group problem with physical-chemical data (10 variables) consisting of 41 samples from three groups. It is demonstrated that the modified CVA method forces the discriminative information into the first canonical variates as expected. The weight vectors found in the modified CVA method possess the same properties as weight vectors of the standard CVA method. By combination of the suggested method with, for example, Linear Discriminant Analysis (LDA) as a classifier, an operational tool for classification and discrimination of collinear data is obtained.

- by Lars Nørgaard and +2
- •
- Analytical Chemistry, Chemometrics, Spectroscopy, Discriminant Analysis

The Neanderthal taxonomic position is a matter of wide disagreement among paleoanthropologists. Some workers consider this fossil human group to represent a different species, Homo neanderthalensis, while others see it as a subspecies of Homo sapiens. This study developed two models of morphological variation to be applied to a comparison between Neanderthals and modern humans: modern human populations provided a measure of intra-specific variation, while the species and subspecies of Pan provided measures of both intra-and inter-specific morphological differences. Although such an approach has been advocated strongly, it has not been systematically undertaken until recently. The techniques of geometric morphometrics were used to collect data in the form of three-dimensional coordinates of craniofacial landmarks. The data were processed using generalized procrustes analysis, and analyzed by an array of multivariate statistical methods, including principal components analysis, canonical variates analysis and Mahalanobis D 2 . The morphological distances between Neanderthals and modern humans, and between Neanderthals and Late Paleolithic/early anatomically modern specimens, are consistently greater than the distances among recent human populations, and greater than the distances between the two chimpanzee species. Furthermore, no strong morphological similarities were found between Neanderthals and Late Paleolithic Europeans. This study does not find evidence for Neanderthal contribution to the evolution of modern Europeans. Results are consistent with the recognition of Neanderthals as a distinct species.

- by Katerina Harvati
- •
- Comparative Anatomy, Evolutionary Biology, Archaeology, Anthropology

The first known fossil ape from the early-middle Miocene of Fejej, Ethiopia, is described here. The specimen, FJ-18SB-68, is a partial ulna from a locality dated by 40 Ar/ 39 Ar and paleomagnetic methods to a minimum age of 16.18 MYA. Compared to a variety of extant and fossil ulnae, FJ-18SB-68 is most similar to Turkanapithecus, Proconsul, and Pliopithecus, and appears to have been an arboreal quadruped with substantial forearm rotational mobility.

- by John Kappelman
- •
- Evolutionary Biology, Archaeology, Anthropology, Locomotion

The paleoanthropological remains from Grotta di San Teodoro near Acquedolci (province of Messina, Italy) represent the oldest and largest skeletal collection yet found documenting human settlement of Sicily. The sample, attributed to the Late Epigravettian (between 14,000 and 10,000 years B.P.), consists of seven variously complete adult individuals (San Teodoro 1-7). We compare the cranial sample to an array of both prehistoric and recent samples using multivariate techniques including D 2 distance analysis, canonical variate analysis, cluster analysis, and multidimensional scaling. Overall, the San Teodoro cranial sample displays a morphometric pattern close to Western European groups of similar antiquity, in particular those from Central and Southern Italy. The morphometric affinities indicate that these people probably came from peninsular Italy by sea during the Late Epigravettian epoch. An alternative hypothesis is that they descended from immigrants that arrived by land during a low sea level episode corresponding to the maximum Wü rmian regression, about 18,000 years B.P, with gene flow accounting for the morphological homogeneity with the populations of peninsular Italy. The San Teodoro skeletal sample provides the first reliable evidence for human settlement of Sicily.

- by Giandonato Tartarelli
- •
- Evolutionary Biology, Archaeology, Paleontology, Anthropology

- by Dean Adams
- •
- Zoology, Principal Component Analysis, Canonical Variate Analysis, Shape Analysis
- by Ernesto Capanna
- •
- Principal Component Analysis, Taxonomy, Biological Sciences, Phylogeny

Wines from three important wine-producing Southern Italian regions (Basilicata, Calabria and Campania) were characterized according to their content in macro-, micro-elements and lanthanides. The elemental composition was used in multivariate statistical analysis to discriminate the wines according to geographical origin. A total of 61 elements were determined in 120 red wines produced in the years 2000e2002, and precisely 40 samples for each region. Results indicated that element analysis provides a good prospect for discriminating wines by regions, even if the element composition is not dependent on the year of production of wine. A Canonical Variate Analysis showed that the discrimination of wines according to their provenance is based on the following elements:

- by Fernanda Galgano and +1
- •
- Animal Production, Cluster Analysis, Canonical Variate Analysis, Red Wine

The objectives of this study were to determine the effects of fire retardant chemical (Phos-Chek G75-F*) and fire suppres- sant foam (Silv-Ex) application, alone and in combination with fire, on Great Basin shrub steppe vegetation. We mea- sured growth, resprouting, flowering, and incidence of galling insects on Chrysothamnus viscidiflorus and Artemisia tridentata. These characteristics were not affected by any chemical

- by Diane Larson
- •
- Wildland Fire, Canonical Variate Analysis, Species Richness, Fire Suppression

Morphological changes in Senegalese sole adults (Solea senegalensis; Kaup 1858) reared at two stocking density conditions (low stocking density, 60% of bottom occupation and high stocking density, 180% of bottom occupation) were investigated using geometric morphometrics. Canonical variate analysis on weight matrix scores, including the uniform component (W'), at the end of the experiment revealed differences in body shape between experimental groups. Low density group presented a similar change pattern to High density group but less intense. Differences were discussed in terms of the effect of a high stocking density on the shape of soles which were very close to commercialization. Results did not indicate a significant size-related shape, likely due to fish age.

- by Rosa Flos
- •
- Aquaculture, Fisheries Sciences, Canonical Variate Analysis, Stocking Density

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... more

- by Gene H Albrecht
- •
- Human Evolution, Multivariate Statistics, Multivariate Analysis, Group membership

The circuit changes that mediate parkinsonian tremor, while likely differing from those underlying akinesia and rigidity, are not precisely known. In this study, to identify a specific metabolic brain network associated with this disease manifestation, we used FDG PET to scan nine tremor dominant Parkinson's disease (PD) patients at baseline and during ventral intermediate (Vim) thalamic nucleus during deep brain stimulation (DBS). Ordinal trends canonical variates analysis (OrT/CVA) was performed on the within-subject scan data to detect a significant spatial covariance pattern with consistent changes in subject expression during stimulation-mediated tremor suppression. The metabolic pattern was characterized by covarying increases in the activity of the cerebellum/dentate nucleus and primary motor cortex, and, to a less degree, the caudate/ putamen. Vim stimulation resulted in consistent reductions in pattern expression (p<0.005, permutation test). In the absence of stimulation, pattern expression values (subject scores) correlated significantly (r=0.85, p<0.02 with concurrent accelerometric measurements of tremor amplitude.

- by jorge montero
- •
- Water, Water quality, Environmental Monitoring, Multidisciplinary

Limnological data (e.g., water chemistry, lakewater temperature, vegetation zone and degree of human impact) were collected from lakes spanning the Russian arctic treeline in three regions: on the Taimyr Peninsula and near the mouth of the Lena River, both in central arctic Siberia, and near the mouth of the Pechora River, western arctic Russia. Pearson correlation and canonical variates analyses revealed similar environmental gradients in all three regions. Variables expressing ionic composition of the water (i.e., cations, anions, dissolved inorganic carbon and conductivity) were highly intercorrelated, as were nutrients, chlorophyll a, particulate organic matter and metal (i.e., Fe and Mn) concentrations. Lakewater transparency was related to water colour (i.e., Fe, Mn and dissolved organic carbon) and productivity. Regional differences among the lakes were strong and appeared to reflect differences in geology, hydrology and human impact. For example, Na and Cl concentrations were related to proximity to the ocean in the Lena and Pechora River regions but not in the more inland Taimyr region. Extensive mining and smelting at Norilsk, on the Taimyr Peninsula, has apparently resulted in elevated major ion and metal concentrations in lakes closer to the city. Surface water temperatures, nutrients, and related variables were particularly useful for distinguishing lakes in different vegetation biomes. Forest lakes were typically warmer, with slightly elevated concentrations of dissolved organic carbon (DOC). Lakes in the forest-tundra zone often had higher concentrations of particulate organic matter, Fe and Mn. Tundra lakes were characterized by low nutrient and DOC concentrations. These data will facilitate the development of models that predict the outcome of future climatic change on arctic and subarctic aquatic ecosystems, as well as provide baseline data for future limnological studies in these remote regions.

- by John Smol
- •
- Earth Sciences, Climate Change, Limnology, Statistical Analysis

Difficulties in delimiting well-defined entities in the dogroses (Rosa sect.Caninae) has resulted in very variable taxonomic treatments. The present study was undertaken to provide a background for taxonomy as well as plant breeding. Morphometric diversity was analysed on seedlings obtained from field collections in South Sweden of three species,Rosa dumalis, R. rubiginosa andR. villosa. A canonical variates analysis showed that the three species are relatively distinct whereas two subspecies ofR. dumalis were less well discriminated. Analyses of variance demonstrated that intraspecific variation is pronounced inR. dumalis and, to a lesser extent, inR. villosa.

- by Gun Werlemark
- •
- Evolutionary Biology, Plant Biology, Plant Systematics, Plant Breeding

This paper investigates the use of a flexible forecasting method based on non-linear Markov modelling and canonical variate analysis, and the use of a prediction algorithm to forecast conditional volatility. We assess the dynamic behaviour of the model by forecasting volatility of a stock index. It is found that the non-linear non-parametric model based on canonical variate analysis forecasts stock index volatility significantly better than the GJR-GARCH(1,1)-t model due to the flexibility in accommodating multiple dynamic patterns in volatility which are not captured by its parametric counterpart.

1,3, 2 * SARITA JANE BENNETT and SIMONETTA BULLITTA 1Centre for Legumes in Mediterranean Agriculture, University of Western Australia, 35 Stirling Hwy, 2 Crawley, WA 6009, Australia; C.N.R., Centro di Studio sui Pascoli Mediterranei,V. De... more

A preliminary investigation on the morphometric variations among four Rhinolophus species was carried out using voucher specimens from the Universiti Malaysia Sarawak (UNIMAS) Zoological Museum and the Department of Wildlife and National Park (DWNP), Kuala Lumpur. A total of 19 individuals from R. acuminatus, R. affinis, R. creaghi and R. stheno were morphologically analysed where 27 linear measurements of body, skull and dental were appropriately recorded. The data were subjected to Discriminant Function Analysis (DFA) and Canonical Variate Analysis (CVA) using SPSS Version 15.0 and Cluster Analysis of Euclidean distance using Minitab Version 14.4. The highest character loadings observed in Function 1, Function 2 and Function 3 were the fifth digit metacarpal length (D5MCL), the fourth digit metacarpal length (D4MCL) and the palatal length (PL) with the standardized canonical discriminant function coefficient value of 22.384, 14.235 and 8.122, respectively. These three characters are identified as the best morphological predictor in for differentiating the four species of Rhinolophus in this study. Thus, the morphometric approach which as being more cost-effective could be useful in addition to DNA sequencing for aiding in species identification.

Background: Analysis of Cerebrospinal Fluid (CSF) samples holds great promise to diagnose neurological pathologies and gain insight into the molecular background of these pathologies. Proteomics and metabolomics methods provide invaluable information on the biomolecular content of CSF and thereby on the possible status of the central nervous system, including neurological pathologies. The combined information provides a more complete description of CSF content. Extracting the full combined information requires a combined analysis of different datasets i.e. fusion of the data. Results: A novel fusion method is presented and applied to proteomics and metabolomics data from a pre-clinical model of multiple sclerosis: an Experimental Autoimmune Encephalomyelitis (EAE) model in rats. The method follows a mid-level fusion architecture. The relevant information is extracted per platform using extended canonical variates analysis. The results are subsequently merged in order to be analyzed jointly. We find that the combined proteome and metabolome data allow for the efficient and reliable discrimination between healthy, peripherally inflamed rats, and rats at the onset of the EAE. The predicted accuracy reaches 89% on a test set. The important variables (metabolites and proteins) in this model are known to be linked to EAE and/or multiple sclerosis.

- by Sybren Wijmenga
- •
- Multiple sclerosis, Metabolomics, Proteomics, Biological Sciences

Partial Least Squares (PLS) methods are particularly suited to the analysis of relationships between measures of brain activity and of behavior or experimental design. In neuroimaging, PLS refers to two related methods: (1) symmetric PLS or Partial Least Squares Correlation (PLSC), and (2) asymmetric PLS or Partial Least Squares Regression (PLSR). The most popular (by far) version of PLS for

- by Atul Krishnan
- •
- Computer Science, Algorithms, Experimental Design, Medicine

The time it takes for a human participant to decide whether a given stimulus is an element of a remembered set increases approximately linearly with the number of elements in the set. Here we tested for and detected a spatial pattern of brain activity whose magnitude of expression during this memory search process correlates with set size. We then tested the idea that memory search simply involves a reactivation of neurons involved in remembering the set by statistically comparing the patterns of brain activity corresponding to memory search and set size dependent working memory maintenance. These patterns were significantly different, suggesting that memory search and working memory maintenance are mediated by distinct neural mechanisms.

- by Joseph Flynn
- •
- Working Memory, Space perception, Memory, Brain Mapping

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document and are linked to publications on ResearchGate, letting you access and read them immediately. This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues.

- by Giandonato Tartarelli and +1
- •
- Evolutionary Biology, Archaeology, Paleontology, Anthropology

We consider the problem of optimally separating two multivariate populations. Robust linear discriminant rules can be obtained by replacing the empirical means and covariance in the classical discriminant rules by S or MM-estimates of location and scatter. We propose to use a fast and robust bootstrap method to obtain inference for such a robust discriminant analysis. This is useful since classical bootstrap methods may be unstable as well as extremely time-consuming when robust estimates such as S or MM-estimates are involved. In particular, fast and robust bootstrap can be used to investigate which variables contribute significantly to the canonical variate, and thus the discrimination of the classes. Through bootstrap, we can also examine the stability of the canonical variate. We illustrate the method on some real data examples.

- by Stefan Van Aelst
- •
- Statistics, Discriminant Analysis, Canonical Variate Analysis, Bootstrap Method

Pollen grains of four cultivars of olive (Oka europaea L.) were examined using scanning electron microscope linked to an image analyser (IA-SEM). For each cultivar 20 pollen grains from each one of five test trees were studied, with a total of 100 replications. For exine pattern more than 400 measurements were done for each detail of each cult&u. The details analysed were: lumina area, muri width, lumina perimeter, shape factor, equivalent circle diameter (ECD), maximum (L), minimum (W) and mean diameter, L/W ratio, size index, maximum, minimum and mean Martin's radius. The whole data matrix was subjected to principal components analysis. This technique reduced the dimensionality of the data set and determined the dominating variables. Successively, the canonical variate analysis, on the reduced data, separated the samples according to their varietal origin. The application of multivariate approaches appeared a very useful method in identifying the discriminant exine pattern features of taxonomic value:

- by Jorge Herrera-silveira
- •
- Water, Water quality, Environmental Monitoring, Multidisciplinary

This study aimed to determine the factors which regulate soil microbial community organisation and function in temperate upland grassland ecosystems. Soil microbial biomass (C mic), activity (respiration and potential carbon utilisation) and community structure (phospholipid fatty acid (PLFA) analysis, culturing and community level physiological pro®les (CLPP) (Biolog w)) were measured across a gradient of three upland grassland types; Festuca±Agrostis±Galium grassland (unimproved grassland, National Vegetation Classi®cation (NVC) Ð U4a); Festuca± Agrostis±Galium grassland, Holcus±Trifolium sub-community (semi-improved grassland, NVC Ð U4b); Lolium±Cynosurus grassland (improved grassland, NVC Ð MG6) at three sites in different biogeographic areas of the UK over a period of 1 year. Variation in C mic was mainly due to grassland type and site (accounting for 55% variance, v, in the data). C mic was signi®cantly P , 0:001 high in the unimproved grassland at Torridon (237.4 g C m 22 cf. 81.2 g C m 22 in semi-and 63.8 g C m 22 in improved grasslands) and Sourhope (114.6 g C m 22 cf. in 44.8 g C m 22 semi-and 68.3 g C m 22 in improved grasslands) and semi-improved grassland at Abergwyngregyn (76.0 g C m 22 cf. 41.7 g C m 22 in un-and 58.3 g C m 22 in improved grasslands). C mic showed little temporal variation v 3:7%: Soil microbial activity, measured as basal respiration was also mainly affected by grassland type and site n 32%: In contrast to C mic , respiration was signi®cantly P , 0:001 high in the improved grassland at Sourhope (263.4 l h 21 m 22 cf. 79.6 l h 21 m 22 in semi-and 203.9 l h 21 m 22 unimproved grasslands) and Abergwyngregyn (198.8 l h 21 m 22 cf. 173.7 l h 21 m 22 in semi-and 88.2 l h 21 m 22 unimproved grasslands). Microbial activity, measured as potential carbon utilisation, agreed with the respiration measurements and was signi®cantly P , 0:001 high in the improved grassland at all three sites (A 590 0.14 cf. 0.09 in semi-and 0.07 in unimproved grassland). However, date of sampling also had a signi®cant P , 0:001 impact on C utilisation potential v 24:7% with samples from April 1997 having highest activity at all three sites. Variation in microbial community structure was due, predominantly, to grassland type (average v 23:6% for bacterial and fungal numbers and PLFA) and date of sampling (average v 39:7% for bacterial and fungal numbers and PLFA). Numbers of culturable bacteria and bacterial PLFA were signi®cantly P , 0:001 high in the improved grassland at all three sites. Fungal populations were signi®cantly P , 0:01 high in the unimproved grassland at Sourhope and Abergwyngregyn. The results demonstrate a shift in soil microbial community structure from one favouring fungi to one favouring bacteria as grassland improvement increased. Numbers of bacteria and fungi were also signi®cantly P , 0:001 higher in August than any other sampling date. Canonical variate analysis (CVA) of the carbon utilisation data signi®cantly P , 0:05 differentiated microbial communities from the three grassland types, mainly due to greater utilisation of sugars and citric acid in the improved grasslands compared to greater utilisation of carboxylic acids, phenolics and neutral amino acids in the unimproved grasslands, possibly re¯ecting substrate availability in these grasslands. Differences in C mic , activity and community structure between grassland types were robust over time. In addition, broad scale measures of microbial growth and activity (C mic and respiration) showed little temporal variation compared to measures of soil microbial community structure, which varied quantitatively with respect to environmental variables (temperature, moisture) and plant productivity, hence substrate supply.

- by Sue Grayston
- •
- Biology, Biological Sciences, Environmental Sciences, Community Structure

- by Noor Haliza Hasan
- •
- Biology, DNA sequence design, Cluster Analysis, Species identification

The Eurasian woodwasp, Sirex noctilio, causes considerable tree mortality in commercial pine plantations in southern KwaZulu-Natal, South Africa. Broad-scale visual assessments of infestation provided by forest managers are currently used to measure forest health and vitality. The effectiveness of visual assessments is questionable because they are qualitative, subjective and dependent on the skill of the surveyor. Remote sensing technology provides a synoptic view of the canopy and thus offers an alternative to the conventional methods of monitoring forest health and vitality. In this study, high resolution (0.5 × 0.5m) digital multispectral imagery (DMSI) was acquired over commercial Pinus patula trees of varying age classes, which had been ground assessed and ranked on an individual tree crown basis using a severity scale. The severity scale was based on a hierarchy of decline symptoms that are visibly apparent on the infested tree and are represented in this study as the green, ...

- by Uncle BOB
- •
- Sociology, Forestry, Remote Sensing, Tree Mortality

Geometric morphometric methods of capturing information about curves or outlines of organismal structures may be used in conjunction with canonical variates analysis (CVA) to assign specimens to groups or populations based on their shapes. This methodological paper examines approaches to optimizing the classification of specimens based on their outlines. This study examines the performance of four approaches to the mathematical representation of outlines and two different approaches to curve measurement as applied to a collection of feather outlines. A new approach to the dimension reduction necessary to carry out a CVA on this type of outline data with modest sample sizes is also presented, and its performance is compared to two other approaches to dimension reduction. Two semi-landmark-based methods, bending energy alignment and perpendicular projection, are shown to produce roughly equal rates of classification, as do elliptical Fourier methods and the extended eigenshape method ...

- by Kristen Covino
- •
- Bioinformatics, Zoology, Principal Component Analysis, Life Sciences
- by Eva Schlecht and +1
- •
- Earth Sciences, Arid environments, Biological Sciences, Environmental Sciences

CITATIONS 18 READS 157 4 authors, including: Some of the authors of this publication are also working on these related projects: Elliptic Fourier Analysis of Bumblebee male genitalia View project SAPOLL -SAuvons nos POLLinisateurs -Plan d'action transfrontalier en faveur des pollinisateurs sauvages / SAmenwerken voor POLLinators -Grensoverschrijdend actieplan voor wilde bestuivers View project Thibaut De Meulemeester

- by Denis Michez
- •
- Evolutionary Biology, Geology, Paleontology, Geometric Morphometrics

In this paper, we propose an intelligent data analysis method for modelling and prediction of daily electricity consumption in buildings. The objective is to enable a building management system to be used for forecasting and detection of abnormal energy use. First, an outlier detection method is proposed to identify abnormally high or low energy use in building. Then a canonical variate analysis is employed to describe latent variables of daily electricity consumption profiles, which can be used to group the data sets into different clusters. Finally, a simple classifier is used to predict the daily electricity consumption profiles. A case study, based on a mixed use environment, was studied. The results demonstrate the method proposed in this paper can be used in conjunction with a building management system to identify abnormal utility consumption and notify building operators in real time.

- by Chris Bowers
- •
- Engineering, Statistical Analysis, Energy Consumption, Energy Management

The difference between the culture-stratigraphic entities, the MSA I and MSA II, in the Klasies River sequence is explored by statistical analysis of the end products. Technological analysis of the cores, end products and waste products suggested that the MSA I and MSA II represent distinct technological traditions aimed at producing different end products. To quantify the difference between the end products, points and blades, extensive univariate and multivariate statistical analyses of continuous variables have been undertaken. Biplot methodology is adopted for enhancing the statistical analysis. Canonical variate analysis biplots are constructed and alpha-bags added for visual displays of the overlap and separation among the different groups. It is demonstrated that the platform thickness relative to length is useful in discriminating between the end products of the MSA I and MSA II. Statistical analyses support a clear distinction between the MSA I and MSA II.

- by sarah wurz
- •
- Archaeology, Geochemistry, Multivariate Statistics, Archaeological Science

Frequently multivariate experimental data taken over multiple occasions is found to produce a multivariate multi-table experiment. Neither the separate analysis of each occasion, using MANOVA or canonical variate analysis, nor the joint analysis using the STATIS-ACT methodology for multiple tables, are adequate to capture the real structure of the data matrices, since the first account for group structure but not time evolution and the second confuses the "between" and "within" group variabilities. A method baptized Canonical STATIS, based on STATIS-ACT methodology, is proposed. The method permits to account for data group structure as well as time evolution on the various occasions by obtaining common or stable canonical variates across multiple occasions or data sets. A simultaneous representation of groups, variables and occasions (biplot) is obtained, thereby widening the capabilities of classical methods.

- by Jose Luis Vicente Villardon
- •
- Econometrics, Statistics, Structured data, Canonical Variate Analysis

Seventy-three Southern Italian red wines were characterized according to their content in total polyphenols, trans- and cis-resveratrol and biogenic amines. These quality parameters were used in multivariate statistical analysis to discriminate the wines according to their specific geographical origin. The results indicated that total polyphenols, resveratrol isomers and biogenic amines provide a good prospect for discriminating wines by regions. The discrimination was also possible to a lesser extent by cultivar. In particular, canonical variate analysis suggested that the discrimination of wines according to their provenance is based on the following parameters: cis-resveratrol, total polyphenols, spermidine and tryptamine for Basilicata region; agmatine and trans-resveratrol for Calabria and Campania regions; cadaverine, ethanolamine, histamine, putrescine and tyramine for Puglia region.

Populations of Ocyurus chrysurus were compared genetically and morphometrically along the West Atlantic coast to test the null hypothesis of population homogeneity in the area. Brazilian populations were found to be differentiated in shape (canonical variates analysis; F [48,515] = 10.84, p < 0.0001). Analyses of mitochondrial DNA sequences (663 bp of the control region) did not show any differences between Brazilian populations but could detect differences between Brazilian and Caribbean (Belize) populations. The samples from Pernambuco differed significantly from the other Brazilian populations in allozyme frequencies (11 loci; F ST = 0.167; p < 0.05), but this may have resulted from the small number of samples analysed for that population. Sequence variation of Belize samples departed from neutral expectations (Fu's FS = -8.88; p < 0.001). A mismatch distribution analysis points to an ancient population expansion in that area. We conclude that the genetic data do not allow the rejection of the null hypothesis of panmixia for Brazilian yellowtail snapper populations which should be treated as a single genetic stock, with a latitudinal gradient on their morphology which probably results from phenotypic plasticity. On the other hand, there is a severe restriction to gene flow between O. chrysurus populations from the Caribbean and from the southwestern Atlantic.

- by Anderson Aguiar
- •
- Genetics, Molecular Biology, Gene Flow, Phenotypic Plasticity

Metabolomic technologies produce complex multivariate datasets and researchers are faced with the daunting task of extracting information from these data. Principal component analysis (PCA) has been widely applied in the field of metabolomics to reduce data dimensionality and for visualising trends within the complex data. Although PCA is very useful, it cannot handle multifactorial experimental designs and, often, clear trends of biological interest are not observed when plotting various PC combinations. Even if patterns are observed, PCA provides no measure of their significance. Multivariate analysis of variance (MANOVA) applied to these PCs enables the statistical evaluation of main treatments and, more importantly, their interactions within the experimental design. The power and scope of MANOVA is demonstrated through two different factorially designed metabolomic investigations using Arabidopsis ethylene signalling mutants and their wild-type. One investigation has multiple experimental factors including challenge with the economically important pathogen Botrytis cinerea and also replicate experiments, while the second has different sample preparation methods and one level of replication ÔnestedÕ within the design. In both investigations there are specific factors of biological interest and there are also factors incorporated within the experimental design, which affect the data. The versatility of MANOVA is displayed by using data from two different metabolomic techniques; profiling using direct injection mass spectroscopy (DIMS) and fingerprinting using fourier transform infra-red (FT-IR) spectroscopy. MANOVA found significant main effects and interactions in both experiments, allowing a more complete and comprehensive interpretation of the variation within each investigation, than with PCA alone. Canonical variate analysis (CVA) was applied to investigate these effects and their biological significance. In conclusion, the application of MANOVA followed by CVA provided extra information than PCA alone and proved to be a valuable statistical addition in the overwhelming task of analysing metabolomic data.

- by Luis Mur
- •
- Analytical Chemistry, Principal Component Analysis, Metabolomics, Experimental Design