Showing 23 results for Factor Analysis
R. Honarnejad, M. Shoai-Deylami,
Volume 8, Issue 2 (7-2004)
Abstract
The results of diallel analysis of 7 parents and 21 F2 progenies of Burley tobacco cultivars, which were transplanted in 2002 as RCBD with 3 replications in the Tobacco Research Center, Rasht, Iran, showed significant genetic differences among genotypes and high GCA and SCA for most of the traits. Therefore the role of additive and non-additive (dominant) effects of genes on the formation of the corresponding characteristics was important. The graphical analysis of progenies of diallel crosses showed partial dominant effect for nicotine percentage in leaves. The distribution of parents around regression line showed most dominant genes in cultivars B.CDL 28, B.Banket, and B.21 while cultivars B.14 and B.TN 86 hade most recessive genes for this characteristic. Also the low and high percentages of nicotine were controlled by dominant and recessive genes, respectively. Estimated simple (phenotypic) relationships between characteristics showed significant correlation between dry leaf yield of tobacco cultivars and its components such as leaf area index (LAI) (r = 0.482**), time to flowering (r = 0.440*), appearance of leaves (r = 0.648**) and percent age of dry matter of leaves. The path coefficient analysis showed very high direct influence of dry matter percent age of leaves, appearance of leaves, and LAI in dry leaf yield. These characteristics would be a favorite selection index for increasing tobacco yield, since characteristics such as leaves per plant, plant height and time to flowering have no significant direct influence on dry leaf yield of tobacco cultivars. These yield components explain 82% of variance of tobacco yield.
Results of factor analysis, using principal Component Analysis (PCA) with Varimax rotation showed that characteristics such as leaves per plant, LAI, and plant height with high positive and significant factor loading as a morphological factor explain 44% of data variance. The second factor including such traits as time to flowering, appearance, and percent of dry matter of leaves with high positive and significant factor loading, form a physiological factor. These two factors together explains 65% of variance of dry leaf yield of tobacco cultivars.
A. Sarkhosh, Z. Zamani, M. R. Fatahi Moghaddam, A. Ebadi, A. Saie, S. Z. Tabatabaie, M. R. Akrami,
Volume 10, Issue 4 (1-2007)
Abstract
In order to determine the most important quantitative and qualitative characteristics of pomegranate fruit and its components for categorizing the genotypes, a study was carried out using 24 genotypes. In this study 28 quantitative and qualitative characteristics of pomegranate fruit and seed were evaluated. Analysis of variance showed that all of the characteristics in examined genotypes were significant, showing high variability in characteristics. Results of simple correlation analysis showed the existence of significant, positive and negative correlations among some important characteristics. Also, factor analysis showed that most of fruit juice, aril and seed characteristics constituted the main factors. Effective characteristics categorized in seven factors accounted for 89% of total variance. For each factor, eigen value more than 0.7 was considered significant. Cluster analysis was performed using these seven factors and genotypes, in the distance of 9, were divided into 5 main clusters. These groups were mainly distinct in fruit flavor and soft seededness. Furthermore, by using three main factors, genotypes positions determined in 3-plot analysis, which distinguished sweeter genotypes from sweet-sour and sour genotypes.
M .tousi Mojarrad, M. R. Ghannadha,
Volume 10, Issue 4 (1-2007)
Abstract
In order to evaluate grain yield potential and study dry matter remobilization to seed, 8 commercial bread wheat cultivars were examined in two separate experiments, using randomized complete block design with 4 replications in 2003. There were significant differences for most traits in normal and drought conditions. Sarday variety had the least grain yield in the two environments. There was considerable variation between genotypes for independent parameters of dry matter remobilization to different plant parts in the two environments. Factor analysis showed that seven factors accounted for 100 % of the data total variance in normal and drought conditions. Factor analysis also showed that the effective traits to components grain yield , height and phenological traits were important for the selection of appropriate genotypes. Calculation of drought resistance indices showed that four indices including mean productivity, geometric mean productivity, harmonic index, stress tolerance index were more effective than other indices for the selection of drought tolerant genotypes.
M. R. Jazayeri Nooshabadi, A. M . Rezai,
Volume 11, Issue 1 (4-2007)
Abstract
This research was conducted at Research Farm of Isfahan University of Technology to examine the relations among phonologic, morphologic and physiologic traits and their effects on grain yield of 20 oat (Avena sativa L.) genotypes. A randomized complete block design with three replications was used in each irrigation treatment (i.e. irrigation after 70±3, and 130±3 mm evaporation from class A pan). Grain yield had significant and positive correlations with harvest index and relative water content (RWC) in both irrigation treatments and with plant height at maturity in non-stress condition. Also, grain yield had significant and negative correlation with days to heading. Factor analyses resulted in three factors in each condition, which were named yield factor, phenologic factor and sink factor in both conditions. Based on stepwise regression, harvest index was the first character which entered the grain yield model in both conditions. At the next step, RWC and plant height entered the model, with harvest index explaining 66% and 64% of yield variation in stress and non-stress conditions, respectively. The results of path analysis in both conditions showed that harvest index had the highest positive direct effect on grain yield. The highest indirect negative effects through number of grain per panicle were obtained for the harvest index and panicle m-2 in stress and non-stress conditions, respectively.
V. Narjesi, H. Zeinal Khaneghah, A. Zali,
Volume 11, Issue 41 (10-2007)
Abstract
Thirty soybean cultivars from different maturity groups were evaluated in a randomized complete block design with four replications in Research Station of College of Agriculture at Tehran University in Karaj in 2004. The purpose was to determine genetic relationship of some important agronomic traits related to seed yield. Analysis of variance showed that there were significant differences among varieties for the traits under study, indicating the existence of genetic variation among varieties. Number of pods/plant, number of seeds/plant and 100-seed weight, all of which are considered yield components, had the highest genotypic correlation with seed yield. Using stepwise regression analysis, 85.6 % of seed yield variation was attributed to four traits, including harvest index, biological yield, protein precent and number of seeds /plant. Harvest index was more important for predicting seed yield compared to other traits based on standardized ßs. Results of path analysis showed that the harvest index and protein precentage had the highest and lowest direct and positive effect (p=0.536), (p=0.008), respectively. Therefore, harvest index may be considered as a selection criteria to improve seed yield in breeding programs. Results of factor analysis showed five independent factors accounted for 80.2 % of total variations in data. The first principle determined 28.2 % of total variation and was designated as phenologic factor.
M. Sabokdast , F. Khyalparast,
Volume 11, Issue 42 (1-2008)
Abstract
This research was conducted in order to determine the relationship between grain yield and yield components, using 30 common bean varieties in a randomized complete block design with four replications at the Agricultural filed faculty of Agriculture,Tehran university In this study 18 traits were assessed on 10 random plants from each plot. The result showed that there were significant differences among varieties in terms of trait under study, indicating the existence of genetic variation among varieties. Also results showed that the grain yield had a positive and significant genotypic correlation with number of seed/pod, pod weight, number of pod/plant, biological yield, days to flowering and maturity. Stepwise regression analysis showed that the maximum variation in grain yield could be attributed to the number pod/plant, number seed/plant, 100 seed weight and pod length. The results of path analysis showed that the highest direct effect, being positive, was related to number seed/plant and the lowest direct effect, which was related to number pod/plant. Factor analysis resulted in three factors that accounted for 78/7% of total variation. The first factor accounted for 38.39% of total variation and was designated as yield and yield component factor. This factor is comprised of pod weight, biological yield, grain yield and number of pod/plant traits.
B. Heidar, G.h. Saeidi, B.e. Sayed-Tabatabaei,
Volume 11, Issue 42 (1-2008)
Abstract
In this study, factor analysis was conducted to determine the factors which contributed to the variation of quantitative traits and path analysis was performed to find the direct and indirect effects of yield components on grain yield in bread wheat. A doubled haploid population of 157 lines of wheat (Triticum aestivum L.) was evaluated for agronomic and morphological traits, using a randomized complete block design with three replications in 2003 and 2004. The results of factor analysis based on maximum likelihood indicated five factors explaining 80.4% and 73.9% of total variation in 2003 and 2004, respectively. The first factor in 2003 had 30.5% contribution to the total variation, strongly influenced by the traits of pollination date, heading date, flag leaf length and days to maturity. This factor also indicated the negative relationship among the yield components and the importance of relationship between grain yield and some morphological traits. The first factor in 2004 was more affected by grain weight/spike, grains/spike and 1000-grain weight, thus it was named as grain yield factor. The second and third factors in 2003 were considered as plant height and grain yield and in 2004 as maturity and plant height, respectively. The results of path analysis showed that grains/spike had the most direct and positive effects on grain yield in 2003 (1.33) and 2004 (0.87). Because of the negative and high indirect effects of grains/spike via fertile spikes/m2 and 1000-grains weight on grain yield, the correlation coefficient between grain yield and grains/spike was very low. There was not much difference between the phenotypic and genetic direct effects of spike/m2 on grain yield, indicating that their relationship was less affected by environmental conditions. In general, the results showed that grains/spike and spikes/m2 can be more efficient compared to 1000- grains weight for increasing grain yield and can be used as selection indices in breeding programs. Also, according to the results of factor analysis, selection based on the fourth factor including biological yield, spike/m2 and grain yield as selection index can be effective to improve grain yield in breeding programs.
M. Vafakhah, G.h. Shojaei,
Volume 11, Issue 42 (1-2008)
Abstract
Continuous measurement of river discharge is a hard and expensive task in hydrology. To overcome this problem, the stage readings at hydrometric gauges are permanently taken and the discharge of any time at which the actual discharge is unavailable will be estimated through a relationship between discharge and stage. To study the stage-discharge relations and the capability of long-term data in establishing a permanent stage-discharge relationship, and also to determine the best time to measure the discharge of rivers, a study was conducted at the hydrometric station of the Zayandehrud regulatory dam using data from 1990 to 2003. The data were analyzed using simple regression analysis, the percentage of relative error and factor analysis. The results indicated that the best model to show the stage-discharge relation at the studied station is a power function model. Moreover, the model used for every year can only be used for that year. The results also showed that the most suitable times for the measurement of discharge are July, December and March.
A.m. Amini, A. Zeynal Hamadani, M. Ramazani,
Volume 12, Issue 43 (4-2008)
Abstract
The general goal of this research was to study the most effective Intra-Organizational factors in the success of the poultry-farm cooperatives in Tehran province. To this end, viewpoints of 151 managers and members of agricultural cooperatives were collected and studied. Survey study was performed to evaluate the effects of four major parameters including member awareness of the principles and philosophy, training programs of members, participation in cooperatives' affairs, and expertise and ability of managers on the success index of poultry cooperatives. The results showed that the degree of members' share of training and knowledge of cooperatives' principles was very low also participation in cooperative's affairs and the managers' expertise and skills were medium. Using multivariate regression analysis, the above-mentioned four factors showed significant effects on the success of the poultry cooperatives with coefficients of 1.87, 0.46, 0.57 and 1.04, respectively.
L. Yaghmaie, S. Soltani, M. Khodagholi,
Volume 12, Issue 44 (7-2008)
Abstract
In order to determine the effect of climatic factors on distribution of Artemisia aucheri and Artemisia sieberi in Isfahan province, thirty nine most important climatic elements in ecological condition of these species were selected. Using a factor analysis method, the most important factors in distribution of these species were determined. These factors which cover 92.3% of the total variance of spatial variation of the species include rainfall, temperature and shining with 41.91, 40.18 and 10.23 percent, respectively. Four most important vegetative types of Artemisia aucheri and Artemisia sieberi in Isfahan province are as follows: 1-Artemisia aucheri and Astragalus,2- Artemisia sieberi,3-Artemisia sieberi as a dominant species, and 4-Artemisia sieberi as a subdominant species. Then the effects of climatic factors on distribution of these species and the average elevation of distribution of these species were determined. The effect of factor analysis on these types shows that rainfall and temperature are the most influential factors in spatial distribution of Artemisia aucheri and Artemisia sieberi in Isfahan province, respectively. In general, Artemisia sieberi has a higher extended tolerance range than Artemisia aucheri, and could be mentioned as a nearly universal species.
M. Ramazani, H.a. Samizadeh Lahiji, H. Ebrahimi Koulabi, A. Kafi Ghasemi,
Volume 12, Issue 45 (10-2008)
Abstract
In order to study agronomic and morphological traits in maize hybrids in Hammedan, two early (108 and 301), three medium (604, 647 and TWC647) and two late maturing (704 and 711) hybrids were evaluated in a randomized complete block design with three replications in Agricultural and Natural Resources Research Station in Hammedan in 2005. 33 morphological and phonological traits were recorded from 10 plants randomly selected from two central rows of each plot. The maximum and minimum grain yield was obtained from SC647 and SC301, respectively. The grain yield had the highest correlation with dehusked ear weight. Factor analysis of data after varimax rotation identified four factors that accounted for 98.03% of total variance. The scatter plot of hybrids based on the two first factors (the seed yield factor and phenological structure factor) showed that SC704 had the maximum forage yield and the best physiological characteristics and SC647 had the maximum grain yield, cob diameter and number of seed in row.
M. Iravani, M. Solouki, A.m. Rezai, B.a. Siasar, S.a. Kohkan,
Volume 12, Issue 45 (10-2008)
Abstract
In order to investigate the diversity and relationship between agronomical traits with seed yield components in barley, twenty advanced barley lines were evaluated in a randomized complete block design with 3 replications at Research Center of Agriculture in Sistan in 2006. Each plot consisted of six rows spaced 20 cm apart and 5 meters long. In this research, 24 Agronomic traits were measured on five randomly selected plants in the central rows of each plot. Analysis of variance showed that there were significant differences among the lines for most of the traits. Line No.7 had the highest (406 grs/m2) and line No.5 had the lowest (309 grs/m2) seed yield. There were high correlation between seed yield and number of panicle/m2. Factor analysis results indicated that 7 independent factors explained 82 percent of the total variation. The first two factors, namely yield components and tillering capacity, explained 41 percent of the total variation. Therefore, it can be concluded that the traits are related to seed yield and tillering capacity, i.e., number of seed per main panicle. 1000 seed weight, number of seed per plant, number of days to physiological maturity and days to heading are the most important characteristics in selecting lines with high seed yield. Number of fertile tiller, total number of tillers and peduncle length were also next set of important traits. Number of days to emergence, nodule number and number of panicle per m2 were also important as selection criteria. Seed weight per plant, biological yield, awn length and the traits that were related to flag leaf had lower importance for selection of lines with high seed yield.
F. Amini, G. Saeidi, A. Arzani,
Volume 12, Issue 45 (10-2008)
Abstract
In order to investigate the relationship among seed yield and its components in safflower, path and factor analysis were conducted using the agronomic and morphological traits of 32 genotypes. Genotypes were evaluated on the Research Farm of Isfahan University of Technology, using a randomized complete block design with three replications. The correlation coefficients showed that number of seeds per capitula and number of capitula per plant had the highest positive correlation with both seed yield and seed yield per plant. The results of regression analysis showed that number of capitula per plant explained 43.6%, and along with seeds per capitula and plant height 60% of the phenotypic variations for seed yield. The regression analysis for seed yield per plant also revealed that seeds per capitula, capitula per plant and seed weight in order had more contributions to the variation of seed yield per plant and explained 81.2% of its variation. Path analysis showed that capitula per plant had the most direct positive effect on both seed yield and seed yield per plant however, this effect was decreased by the indirect and negative effect of seed weight. Results of factor analysis recognized 3 main factors which explained 81.81 % of total variation of the data. These factors were named the seed yield and its components, phenological traits and branching. In general, it can be concluded that seeds per capitula, capitula per plant and seed weight in order contributed more to the seed yield of safflower genotypes. In conclusion, these yield components can be used as selection criteria in breeding programs.
M.m Majidi, A Mirlohi,
Volume 12, Issue 46 (1-2009)
Abstract
This experiment was conducted to investigate the genetic diversity, identify traits explaining yield variation, recognize relationships between traits and classify accessions in a Iranian and forign germplasem of tall fescue. Forty six Iranian and foreign tall fescue accessions were surveyed for Phenological, morphological and agronomical characters in a randomized complete block design with three replications in field for 2 years. Significant differences were observed for all of the characters, indicating broad variation in this germplasm. Iranian accessions had a better performance for most of the traits in both years indicating their high potential for developing commercial varieties in breeding programs. Basis on stepwise regression analysis, crown width justified the majority of hay yield variation, followed by establishment rate, percentage of dry matter, height and number of stem. Hence, these characters could be used for selecting high yielding cultivars. Factor analysis revealed 4 factors which explained more than 80 percent of the total variation and confirmed the results of regression analysis. Using UPGMA method, cluster analysis revealed five groups. Accessions with similar country of origin or same ecological conditions were grouped in same cluster. Regarding the morphological characters the best accessions were identified to be used in the further breeding projects.
K Rabiei, M Khodambashi, A Rezaei,
Volume 12, Issue 46 (1-2009)
Abstract
Factor and principal component analyses are widely used in different sciences especially in agricultural science. To determine the factors that create variation between potato cultivars, in normal (non-stress) and water deficit (stress) conditions, two experiments were conducted in the form of randomized complete block design with three replications in summer 2002. Stepwise regression analysis showed that in normal conditions, stem length, number of stems/plant and leaflet width contributed significantly to yield. In stress condition, other than stem length and number of leaves/main stem, leaflet length also entered the model. As is evident, stem length had a detrimental effect on tuber yield in both stress and non-stress conditions. So, this trait could be used as an important criterion for the selection of high yielding genotypes. Principal component analysis revealed that number of stem, leaf length and leaf width were important traits creating variability between potato cultivars, especially number of stem that had high coefficients in the first principal for both environments. Factor analysis distinguished two factors in normal environment named leaf surface and structural attitude factors, and also two factors in stress environment called photosynthetic surface and structural attitude. Therefore, these factors should be intervened and attended to in breeding programs.
S Safae Chaykar, H Samie Zade, M Esfahani, B Rabiei,
Volume 13, Issue 48 (7-2009)
Abstract
In order to study the correlation of agronomic, morphologic and physiologic traits and their effects on grain yield of rice genotypes in two environments (favorable irrigation and water stress), 49 genotypes were evaluated using a completely randomized block design with 3 replications in two experimental conditions. All practices and conditions were the same for the two experiments with the exception of irrigation, where under stress conditions no irrigation was applied at tillering stage. Comparison of means showed significant differences between genotypes in each environment. Also, differences between yield and yield components of each genotype under two conditions were significant. The results of phenotypic correlations showed that the highest positive and significant correlation with grain yield belonged to number of panicle per plant (0.95) in irrigation conditions and to number of filled grains per panicle (0.92) in water stress conditions. Stepwise regression analysis for grain yield introduced number of panicle per plant, relative water content (RWC), flag leaf length and number of spikelet per panicle, respectively, as effective traits in grain yield in irrigation conditions, however, in stress conditions, number of filled grain per panicle, number of panicle per plant and relative water content were effective traits in yield. The results of path analysis showed that the number of panicle per plant had the highest positive and direct effect on grain yield in the two environments. Factor analysis introduced four factors in the two conditions named yield and crop production, phenologic, harvest index and plant shape and appearance quality of grains factors. Therefore, to select high yield and drought tolerant genotypes, we need to consider number of filled grain per panicle, number of panicle per plant and relative water content. In addition, traits such as panicle length, number of spikelet per panicle, flag leaf length and width that showed significant correlations with grain yield in stress conditions should also be considered important and second to the above mentioned traits.
S. Soltani , L. Yaghmaei , M. Khodagholi , R. Saboohi ,
Volume 14, Issue 54 (1-2011)
Abstract
The temporal and spatial vegetation dynamics is highly dependent on many different environmental and biophysical factors. Among these, climate is one of the most important factors that influence the growth and condition of vegetation. Of the abiotic factors affecting the geographic distribution of vegetation type, climate is probably the most important. Ecological research has traditionally aimed to generalize vegetation types that are assumed to be homogenous. Most of climatic classifications related to bioclimate are focused on limited climatic factors such as temperatue, precipitation and combination of them. As climate is a compound phenomena using limited factors cannot show the climate of a region, and as a result most climatic factors must be considered in bioclimatic classification. Therefore, a climatic study using various climatic factors could reveal the effective factors in distribution of vegetation. In order to determine bioclimatic zones in Chahar-Mahal & Bakhtiari province using multivariate statistical method, 71 climatic variables, which were more important in plant ecological conditions, were selected and evaluated by the factor analysis. The factor analysis revealed that the first three factors which explain %91.8 of total variance among the selected variables were temperature, precipitation, and radiation. According to results and using hierarchical cluster analysis in Ward’s method, bioclimatic classification in Chahar-Mahal province was carried out and 5 bioclimatic zones were found. In addition, Chahar-Mahal province was classified by 4 traditional climatic classification methods (Koppen, Gaussen, Emberger and De Martonne) and those classes were compared to climatic classes obtained by multivariate statistical method. The latter comparison was suggestive of the fact that multivariate statistical method provides a more appropriate classification in comparison to the traditional methods, specially because more dominant vegetation species could be defined for each of the newly described climatic classes. Furthermore, dominant species were determined for each climatic region.
K. Nosrati, M. Majdi,
Volume 21, Issue 4 (2-2018)
Abstract
The soil pollution especially in urban soils is projected to increase drastically and its effects on chemical cycles are yet to be known. Approaches to measure air and water quality are well established, but urban soil quality assessment has received little attention. Soil quality assessment can help as a way to better understand the pollution increase outcomes in urban environments and to establish approaches and integrated soil quality assessment protocols in urban planning and landscape management. Considering lack of information in urban soil quality of Iran, the objective of this study was to assess soil quality under urban land use effect using minimum data set in western part of Tehran. In view of this, 56 soil samples were collected in three land use types of agricultural, parks and urban landscapes, and vacant urban lots and 12 physicochemical properties were measured. The results of analysis of variance (one-way ANOVA) showed that under influence of the land use types, organic carbon, total nitrogen, lime, bulk density and sodium have significant differences. The factor analysis was used to select minimum data set and the results showed that two factors with eigenvalues more than one, explaining more than 68% of total variance, have the most loading factors on organic carbon and sodium. Finally, soil quality indicator (SQI) was determined and compared in different land use types. The results showed that SQI has significant difference in urban land use types and the least soil quality is related to vacant urban lots.
M. Jahan, B. Amiri,
Volume 22, Issue 3 (11-2018)
Abstract
Factor analysis is one of the multivariate statistical techniques that considers the interrelationships between apparently irrelevant variables and helps researchers to find the hidden reasons for the occurrence of an event. In order to evaluate the effects of different irrigation levels and humic acid foliar application and identify the factors affecting water use efficiencies of sesame (Sesamum indicum L.), maize (Zea mays L.) and common bean (Phaseolus vulgaris L.), a split plots experiment based on RCBD design with three replications was conducted during the 2014-15 growing season, at the Research Farm of Ferdowsi University of Mashhad, Iran. Irrigation levels (50 and 100% of water requirement) and foliar application and non-application of humic acid were assigned to main and sub plots, respectively. The results showed that in sesame, the highest seed yield and biological yield were obtained from 100% of water requirement and humic acid spraying treatment. In maize, humic acid spraying under condition of supplying 50% of water requirement increased seed weight per plant, plant height, and leaf area index and soil pH In bean, the highest seed weight per plant, plant height, leaf area index, crop growth rate and soil phosphorous content were observed in the treatment of 100% of water requirement and humic acid spraying. Factor analysis results also showed that in sesame, the variables of seed yield, biological yield, seed weight per plant, plant height, leaf area index, crop growth rate, soil phosphorous and water use efficiency were assigned to the first factor and the variables of soil nitrogen, soil pH and EC were assigned to the second one. In maize, seed yield was assigned in the same group with the variables of biological yield, leaf area index, crop growth rate, soil phosphorous, EC and pH and water use efficiency; in bean, this was with the variables of seed yield and water use efficiency. In general, the research results revealed that identifying the effective variables in each factor and those logical nominations according to Eco physiological knowledge can lead to the direct management of effective variables with regard to associated factor, thereby leading to water efficiency improvement.
H. Ghamarnia, F. Sasani, B. Yargholi,
Volume 23, Issue 1 (6-2019)
Abstract
Exploring the homogenous regions for site specific management is important, especially in the areas under different anthropogenic activities. This was investigated using multi-way analysis including Factor Analysis, Hierarchical Clustering Analysis and k means in the areas under long-term wastewater irrigation over a period of more than 40 years, in Shahre Rey, south of Tehran. By using Factor Analysis model, eight factors as eight geochemical groups were extracted to explain approximately 60% of the total variance related to 37 soil physicochemical properties. The most important groups included the nutrient elements (OM, OC and N), micronutrients (Mn and B), soil water adsorption capacity (Clay, Silt, Sand and CEC), salinity and osmotic pressure (EC, OP and TDS) and sodification (SAR and Na). The maximum values of Cophenet and Silhouette coefficients were equal to 0.77 and 0.83, respectively, dictating the selection of the average linkage approach in Hierarchical Clustering Analysis and three clusters in the k-average method with 19, 24 and 34 mapping units. The Thiessen Polygons method in GIS was applied to separate the geochemical groups in the form of mapping units. This output, which was, in fact, the combination of multi-way models and its visual representation in GIS under separated mapping units of study area, could present suitable management activities for the areas under each cluster.
