SHORT COMMUNICATION PRINCIPAL COMPONENT ANALYSIS AND POLLUTION BY TRACE ELEMENTS IN A MUSSEL SURVEY

The total content of nine trace elements(Mn,Fe,Co,Ni,Cu,Zn,Cd,Hg,Pb)in the soft part of mussels(Mytilus galloprovincialis Lamarck)sampled in two sites was considered.Wild,polluted molluscs weresampled in Muggia Bay(Gulf of Trieste,Northern Adriatic Sea)in the proximity of an important sewerof the city.Edible,unpolluted mussels were simultaneously sampled in a hatchery just off the Bay.Principal component analysis has been applied to correlation matrices obtained from data matrices fromthe literature.The nine variables were reduced to three or two principal components,which explained70-80% of the total variance.The unrotated and orthogonally rotated matrix of the correlations ofvariables with principal components showed that the clusters of elements are positively associated to thefirst two eigenvectors.The origin of some toxic elements in the soft part of mussels from Muggia Bay isdiscussed.The projection on the first two eigenvectors of all data as component scores allow a nearlycomplete separation of polluted from unpolluted molluscs.     

广州市社会空间的因子生态再分析

运用因子分析、聚类分析方法对广州市中心区1990年人口普查数据进行了研究.从9类47个变量中抽取出5个形成广州市社会区类型的主因子,根据各主因子的得分,把广州市中心区划分为七类社会区,并把形成广州市社会空间分异现象的原因归结为五类.通过与1985年类似研究结果的比较,发现广州市社会空间结构的分异现象更趋明显,具体反映在人口“外溢”、居住条件改善及新开发区的形成等方面.最后就如何引导广州市社会空间结构演变进行了讨论.     

基于主成分分析的关中地区农业粮食生产变化的影响因素研究

针对陕西省关中区域1978—2017年的农业生产数据,在分析关中40 a农业粮食生产的趋势变化后,运用主成分分析法,对影响关中农业生产中的地理环境和生产投入等主要因素进行了评价研究。结果表明：（1） 关中农业粮食生产的趋势变化呈现周期为3～7 a的循环增长方式,平均每周期峰值增长率为4.5%。（2） 主成分分析研究后得出,第一主成分全是地理因素指标,方差贡献率达到0.554,对关中地区农业粮食生产起着非常显著的决定影响作用,包括受灾农田面积（不含病虫害）、主要粮食作物播种面积、成灾农田面积（不含病虫害）、有效灌溉耕地面积、耕地面积;第二主成分方差贡献率为0.25,是影响粮食生产的重要因素和农业生产的生命补给。包括农业用电量、化肥、农用机械等生产资料投入和主要粮食作物稳产面积、劳动力投入因素指标;第三主成分为农药应用量,方差贡献率为0.068,影响较小。主成分累计方差贡献为0.872。通过对关中地区农业粮食生产变化的影响因素分析,可以为政府部门提出数据支撑和相关性的建议。     

RELATIONSHIPS BETWEEN THE MORPHOMETRY OF ALPINE AND SUBALPINE BASINS AND REMOTELY SENSED ESTIMATES OF LAKE TURBIDITY,GLACIER NATIONAL PARK,MONTANA

Relationships between characteristics of alpine and subalpine drainage basin form and estimates of lake turbidity were examined. The study was designed to provide insight into sediment removal and delivery systems in an environment where direct process studies are difficult because of logistical constraints. Remote sensing techniques were developed to assess lake turbidity levels through the acquisition of spectral responses by the Landsat Thematic Mapper. Topographic maps were interpreted to derive measurements of morphometric characteristics of 25 study basins within Glacier National Park, Montana. The morphometric variables were chosen to describe basin characteristics related to sediment removal and transport processes. Correlation analysis, principal components analysis (PCA), hierarchical clustering, and analysis of variance (ANOVA) were used to assess characteristics of basin form and the relationship to levels of lake turbidity. A significant correlation between the first principal component, which accounted for the general morphometric character of basins, and estimated lake turbidity was discovered. The areal extent of glacial ice and total stream length within the basins were the most significant variables related to estimated levels of lake turbidity.     

Autotrophic response to lake age,conductivity and temperature in two West Greenland lakes

Predicted changes in future climate necessitate a better understanding of climate impacts on lake biota, and the role of within-lake processes in modifying biotic response. Therefore we examined two climate-related variables (lake-water conductivity and GRIP temperatures) and lake ontogeny (lake age), to determine their influence on lake autotrophic communities in two neighbouring closed-basin lakes from West Greenland spanning the past 8,000 years. Using sedimentary pigments as proxies for lake autotrophic communities, we used synchrony and variance partitioning analyses (VPA) to test three specific hypotheses (a) that lake primary production would increase with lake age, (b) that climate would be the dominant process controlling autotrophic communities in these pristine lakes and (c) that the response of autotrophs to conductivity and temperature would vary depending upon the age of the lake. The results supported our first hypothesis, showing that lakes changed significantly with age, exhibiting an increase and decline in production in the first millennium of their existence, followed by a steady increase in production and increasingly frequent abrupt switches between mixed and meromictic states. The highly synchronous detrended response (r = 0.769) of lake autotrophs in the two study lakes, supported our hypothesis that climate was the dominant factor controlling lake autotrophs. However, VPA revealed that our climate-related variables (temperature and conductivity) explained only small amounts of variance alone (<12.9%) because covariance among them hindered efficient partitioning. In support of our third hypothesis, autotrophs responded significantly to temperature and conductivity in interaction with lake age (>50% variance explained) and with each other (>28% variance explained), such that autotrophic response changed as lakes aged. In spite of this, lakes sometimes responded independently, as a result of differences in the relative proportion of benthic to pelagic production and because of differences in lake morphometry. Together these results show that long-term control of lake autotrophs by climate and lake age is modified on shorter timescales by non-linear responses related to within-lake processes, and by the interaction of different climate variables with each other and with lake age.     

SOIL AND VEGETATION PATTERNS IN BARRIER-ISLAND DUNE ENVIRONMENTS

Plant species distributions often have been attributed to landform characteristics or their associated geomorphic processes. This complicates interpretation of vegetation patterns in that geomorphic processes shape, and are shaped by, landforms. To characterize the biogeographic impacts of this interaction, I used principal components analysis (PCA) to examine hypotheses regarding the structure of variation among soil properties in active barrier-island dune systems. Dune soils and vegetation were sampled on two well-recognized barrier-island morphologies. On low-profile, wave-dominated microtidal barrier islands (South Core Banks, North Carolina) frequent overwash exerts a greater control on the distribution of soil properties. On mixed-energy mesotidal barrier islands (Sapelo Island, Georgia), overwash is less frequent, and the distribution of soil properties is shaped by a complex dune topography. Nontrivial principal components on both islands captured an equivalent amount of variance in the soil data. However, there were inter-island differences in the dimensionality of these nontrivial principal components, and differences in the distribution of variance and factor loadings. Suites of topography-modifying species, unique to each island, were uniform in the strength of their individual correlation with local edaphic variability. I posit that soil variance structure is a useful criterion to distinguish the relative influence on vegetation patterns of soil properties expressed through landforms (Sapelo Island) versus sediment transport processes (South Core Banks). [Key words: dune vegetation, barrier islands, principal components analysis (PCA), overwash.]     

DISCRIMINANT PRINCIPAL COMPONENTS ANALYSIS

When the number of variables exceeds the number of samples, one method of multivariate discriminationis to use principal components analysis to reduce the dimensionality and then to perform canonicalvariates analysis (PC-CVA). This paper proposes an alternative approach in which discriminant analysisis carried out by a weighted principal component analysis of the group means (DPCA). This method doesnot require prior data reduction and produces discriminant factors that are orthogonal in the original dataspace. The theory and performance of the two methods are compared. Although the individual factors ofDPCA are found to be less discriminating than PC-CVA, the overall discrimination, calculated bymultivariate analysis of variance, and the predictive value, estimated by the leaving-one-out error rate,are broadly comparable.     

COMMENTS ON THE NIPALS ALGORITHM

The Non-linear lterative Partial Least Squares(NIPALS)algorithm is used in principal componentanalysis to decompose a data matrix into score vectors and eigenvectors(loading vectors)plus a residualmatrix.N1PALS starts with some guessed starting vector.The principal components obtained by NIPALSdepends on the starting vector;the first principal component could not always be computed.Wold hassuggested a starting vector for NIPALS,but we have found that even if this starting vector is used,thefirst principal component cannot be obtained in all cases.The reason why such a situation occurs isexplained by the power method.A simple modification of the original NIPALS procedure to avoid gettingsmaller eigenvalues is presented.     

The Organization of Production in High Technology Industries: An Empirical Assessment*

Recent attention given to the concept of vertical integration and disintegration of production processes has identified the existence of a continuum, with small, flexibly specialized producers at one end and large, mass production units at the other end. Firms along the continuum constantly strive for the optimum combination of economies of scale and scope. This paper attempts to identify the organizational characteristics of high technology firms in one industrial complex of the U.S. Manufacturing Belt. A principal components analysis is performed on a number of theoretically relevant variables for a sample of high technology establishments in Northeast Ohio. The results indicate that the region contains a variety of organizational forms ranging from isolated workshops to large, vertically integrated assembly plants. These findings conform to expectations derived from recent conceptualizations in the literature. Additional empirical research will further enhance understanding of the processes that shape the organization of production in territorial complexes.     

Effects of soil surface components on soil hydrological behaviour in a dry Mediterranean environment (Southern Spain)

Various researchers have studied the spatial pattern of soil surface components such as vegetation, rock fragments, bare soil, litter and surface crusts, as a key factor of hydrological behaviour in Mediterranean settings with heterogeneous patches of vegetation cover and strong human impact. The studies indicate that there is a mosaic of patches that generate run-off or infiltrate overland flow, distributed in various ways along hillslopes. Few of these studies, however, have looked at areas underlain by metamorphic rocks such as phyllites or schists. This study analysed the temporal and spatial variability of the effects of soil surface components on hydrological processes in a small dry Mediterranean catchment underlain by metamorphic rocks. A systematic sampling of multiple sites throughout a hydrological year was carried out.We related the hydrological behaviour of soil surface components to 1) their position along the hillslope, 2) the distance of existing vegetation tussocks from the line of run-off, 3) rainfall intensity and 4) the main physical/chemical soil properties affecting infiltration processes. Statistical analysis was used to check the validity of the relationships. The results show that soil surface components have highly variable effects, in both space and time, on soil hydrological behaviour. These effects particularly depend on the location along the line of maximum slope and the intensity of preceding rainfall, whose interaction defines soil hydrological status. These results are similar to those for other Mediterranean settings with different lithology, in that the succession of contributing patches are hydrologically interconnected along a hillslope. The variables used and the grouping of explanatory variables through principal component analyses were found to be suitable for discussing the spatial distribution of soil surface components in the hydrologically dynamic environment of the study area.     

中国北方地区农业干旱脆弱性评价

在调查和分析中国北方地区农业生产现状的基础上,选择水资源、农业经济、社会和防旱抗旱能力4个准则层共16个指标构建农业干旱脆弱性评价指标体系。运用主成分分析法对高维变量系统进行有效降维,根据方差贡献率建立中国北方地区农业干旱脆弱性评价模型,获得中国北方地区水资源、农业经济、社会和防旱抗旱能力脆弱性评价结果,获得各省(自治区、直辖市)干旱脆弱性分级阈值和区划。结果表明:(1)通过主成分分析得到农业经济、水资源、防旱抗旱能力、社会等5个主成分,方差贡献率分别为41.99%、19.25%、14.06%、8.07%和5.32%。(2)中国北方地区农业干旱脆弱性从小到大依次为北京、天津、山东、辽宁、吉林、山西、内蒙古、安徽、河北、河南、陕西、宁夏、青海、黑龙江、新疆和甘肃。     

全球食物供给非均衡性演变及贸易影响

降低食物供给的非均衡性是实现“零饥饿”的重要途径。论文在分析1986—2018年全球食物生产时空演变特征的基础上,基于洛伦兹曲线和锡尔系数,对全球与区域食物生产非均衡性及贸易影响进行了测度。结果表明：(1) 1986—2018年,全球各类食物产量均呈增加趋势,产出结构发生了显著变化,谷物在植物性食物中的比重减小,油料、蔬菜和水果所占比重增加。不同区域和国家食物营养来源具有显著差异,且尺度越小,差异越大。(2)谷物、糖料和水果生产的集中程度相对较低,蔬菜、油料、薯类、水产品生产集中程度较高。豆类、蔬菜、油料、糖料和嗜好类人均产量非均衡性呈增加趋势,薯类、水果、畜禽类和水产品的人均产量非均衡性呈降低趋势。人均热量产出非均衡较低,脂肪和蛋白质产出的非均衡性较高。(3)贸易对全球人均食物及营养供给的非均衡性具有显著的降低作用,且其影响程度逐年增加。贸易对糖类、油料和嗜好类非均衡性降低作用最显著,对脂肪供给非均衡性的降低作用较大。制定以营养为导向的食物生产和贸易政策,对提升食物供给均衡性具有重要意义。     

Environmental dust: A tool to study the patina of ancient artifacts

Dust is a significant, albeit under-recognized, component of ancient patinas that accumulates on exposed surfaces of artifacts. Dust storms are ubiquitous in the Levant; however, often unnoticed substances such as minerals, microfossils and pollen can be found within the patina of an artifact, preserving its geological signature. Modern anthropogenic aerosol sources are often characterized by the presence of heavy metals. Pollen from fruits and shrubs that are not indigenous, if found in the patina, can be used to differentiate recent artifacts from those of antiquity. Archaeological materials that are exposed to local environmental and depositional processes in a tel, a cave or soil, may accrete in a patina over time and may have some dust components reflective of the environmental record. The scores of unprovenanced looted antiquities have necessitated the need to differentiate a genuine artifact from a modern fraud. Since the geological component of the dust in the Levant is known, and the climate and its attendant wind patterns apparently were quite constant during recent millennia times, the dust in the patinas of a true artifact is easy to differentiate from patinas containing modern dust. Both contemporary and historical dust can serve as tools to authenticate an artifact.     

主成分分析在农田土壤环境评价中的应用

本文尝试利用主成分分析方法对农田土壤污染物进行识别,并对土壤环境质量进行分级。结果表明,利用主成分分析可以有效地识别土壤污染物中的主要成分,揭示土壤污染物的数据结构和相互间的关系。主成分分析方法可用于定量化的土壤复合污染研究或对历史数据较为缺乏的地区进行土壤环境质量评价。在污染物检测指标数量较大时,可以在一定程度上简化农田土壤重点污染物的定量化识别过程。     

Climatic and atmospheric teleconnection indices and western Arctic sea ice variability

This study examines the sea ice cover minima in the western Arctic in the context of several climatic mechanisms known to impact its variability. The September latitude of western Arctic sea ice is measured along 11 equally-spaced longitudes extending from 176º?W to 126º?W in the Chukchi and Beaufort Seas, 1953–2010. Indices of seasonal atmospheric and oceanic teleconnections and annual mean Northern Hemisphere temperatures (NHT) and CO₂ concentration are orthogonalized using rotated principal component analysis, forming predictors regressed onto the sea ice latitude data at each longitude using stepwise multiple linear regression. Prior to 1998, small amounts of September ice edge variance are explained by teleconnections such as the Arctic Dipole, Arctic Oscillation, and Pacific-North American Pattern. NHTs begin explaining large amounts of ice edge variance starting in 1998. For the 1953–2010 period, up to 68% of the ice edge variance is explained at 161°?W in the Chukchi Sea, mostly by NHTs. With the exception of the three easternmost longitudes (136–126°?W), the teleconnections and NHTs explain over 50% of the regional ice edge variance. Increases in both NHTs and ice retreat since the mid-1990s account for the large explained variances observed in regression analyses extending into recent years.     

Reproductive phenology of Isolatocereus dumortieri (Cactaceae) in semiarid scrub in central Mexico: Effect of rain during the dry season

The giant cactus Isolatocereus dumortieri is a dominant species of the semiarid scrub of central Mexico. Its reproductive period is during the dry season, and it produces essential resources (pollen, nectar and fruits) for a great variety of animal species. We related the production of reproductive structures with water variables in the soil–plant system. The main goal of the present study was to evaluate the consequences of a change in rainfall pattern on the phenology of I. dumortieri. We watered some plants in the dry season to simulate heavy rain events, and to test the hypothesis that water availability during the reproductive season has a negative effect on fruiting and a positive effect on vegetative growth. The seasonality of rain events caused variations in soil water potential and plant osmotic potential, and we found that both variables influenced fruit production. The regression models relating the number of fruits with soil water potential, rainfall and osmotic potential were significant for all three study periods. The highest production of reproductive structures occurred in the driest year (2009), during which there was an ENSO event. Watering did not have a significant effect on osmotic potential or growth in the cacti. However, the watering × time interaction had a negative effect on the number of immature fruits. That is, at the end of the experiment, the plants that received the most water showed a decrease in the number of fruits. Plant growth during the rainy period was significantly greater than during the dry period. The results support the hypothesis that a change in rainfall pattern during the dry season has an effect on fruit production. Vegetative growth, however, occurred only during the wet season.     

Development and application of sedimentary pigments for assessing effects of climatic and environmental changes on subarctic lakes in northern Sweden

A surface-sediment survey of pigments in 100 lakes in the Scandes Mountains, northern Sweden, was combined with a reconstruction of Holocene sedimentary pigments from Lake Seukokjaure to assess the major factors regulating phototrophic communities, and how these controls may have changed during the period from the deglaciation (~9700 cal. years BP) to the present. The study area covers a pronounced gradient of temperature and precipitation, and encompasses the subarctic tree line, an important ecotonal boundary in this region. Lake Seukokjaure is located in a presently treeless basin close to the modern tree line. The spatial survey of sedimentary pigments was analyzed using principle components analysis (PCA) and redundancy analysis (RDA). PCA explained 73–83% of variance in pigment abundance and composition, whereas RDA explained 22–32% of variation in fossil assemblages. Dissolved organic carbon (DOC) content of lake water, sediment δ¹³C, maximum lake depth, elevation and lake-water conductivity were all identified as environmental variables with significant association with pigment abundances in the spatial survey, although phototrophic communities of lakes situated in different vegetation zones (alpine, birch, conifer/birch) were incompletely distinguished by the ordinations. In the RDAs, the primary pigment variability occurred along a production gradient that was correlated negatively to water-column DOC content and δ¹³C signature of sediments. This pattern suggested that the important controls of primary production were light regime and terrestrial supplies of ¹³C-depleted carbon. In contrast, depth, elevation and conductivity were found to be more important for the differentiation of the phototrophic community composition. Application of these spatial survey results to the Holocene sediment record of Lake Seukokjaure demonstrated the importance of DOC for the temporal development of the lake, from an early state of high production to a period of slight oligotrophication. In general, the algal changes were regulated by the interaction of DOC and conductivity, although transitions in the phototrophic community during the late Holocene were less easily interpreted. Terrestrial vegetation development thus appears to be of utmost importance for the regulation of primary production in oligotrophic alpine and subarctic lakes and climate impacts on lakes, whereas other basin-specific factors may control the ontogeny of algal community composition.     

BODY DIAGONALIZATION OF CORE MATRICES IN THREE-WAY PRINCIPAL COMPONENTS ANALYSIS:THEORETICAL BOUNDS AND SIMULATION

In contrast with conventional PCA,a direct superposition and joint interpretation of loading plots is notpossible in three-way PCA,since there may be data variance which is described by unequal componentsof different modes.The contributions to variance of all possible combinations of components aredescribed in the core matrix.Body diagonalization,which is achieved by appropriate rotation ofcomponent matrices,is an essential tool for simplifying the core matrix structure.The maximum degreeof body diagonality which may be obtained from such transformations is analysed from both themathematical and simulation viewpoints.It is shown that,at least in the average case,high degrees canbe expected,which makes the procedure reasonable for many practical applications.Furthermore,simulation as well as theoretical derivation show that the success of body diagonality depends on the so-called polarity of the core array.The methodology is illustrated by a three-way data example fromenvironmental chemistry.     

THE EFFECT OF MEAN CENTERING ON PREDICTION IN MULTIVARIATE CALIBRATION

Traditionally,one form of preprocessing in multivariate calibration methods such as principal componentregression and partial least squares is mean centering the independent variables(responses)and thedependent variables(concentrations).However,upon examination of the statistical issue of errorpropagation in multivariate calibration,it was found that mean centering is not advised for some datastructures.In this paper it is shown that for response data which(i)vary linearly with concentration,(ii)have no baseline(when there is a component with a non-zero response that does not change inconcentration)and(iii)have no closure in the concentrations(for each sample the concentrations of allcomponents add to a constant,e.g.100%)it is better not to mean center the calibration data.That is,the prediction errors as evaluated by a root mean square error statistic will be smaller for a model madewith the raw data than a model made with mean-centered data.With simulated data relativeimprovements ranging from 1% to 13% were observed depending on the amount of error in thecalibration concentrations and responses.