Multivariate Analysis of Soil-Vegetation Interrelationships in a South-Southern Secondary Forest of Nigeria

Abstract

Multivariate statistical techniques were employed to study soil-vegetation interrelationships in a secondary forest of South-Southern Nigeria. The grid system of vegetation sampling was used to randomly collect vegetation and soil data from fifteen quadrats of 10m x 10m. The result of principal components analysis identified seven basic sets of soil-vegetation variables that enhanced the interrelationships. Canonical correlation result indicated a positive association between organic matter and tree size, while the linear association between organic matter and tree density revealed an inverse relationship. The result of redundancy coefficient indicated that 18 percent of the variance in vegetation characteristics was accounted for by the variability in soil properties whereas, 81 percent of the variance in soil properties was accounted for by the variability in vegetation characteristics. The regression analyses on the other hand indicated that exchangeable sodium positively influenced tree species composition and richness; and that tree size as well as tree density exerted substantial influence on the contents of organic matter and total nitrogen of the soil. However, drawing inference from results of canonical correlation analysis and those of multiple regression analysis, it was concluded that soil and vegetation components of the secondary forest vegetation were mutually dependent and therefore exerted joint influences on each another.

Appraisal of arid land status: a holistic assessment pertains to bio-physical indicators and ecosystem values

“…Interpretation and identification of the significant variables on PCA bi-plot were carried out by the criterion provided by Legendre and Legendre (1998) and Iwara et al (2011). In this study, the cumulative percentage in PCA analysis conducted for gaining the weighting factors suggested that during all temporal events, the first four axes together accounted more than 80% variability (Table 5) and this proved the usefulness of this tool in this study (Wei-Giang and Bilquees 2008).…”

“…Similarly, the first four axes of PCA analysis carried out to characterize the temporal variations of land assessment parameters (RCS, revised SQI, HCS, and EMV) suggested 100% cumulative variability (Table 6), and usefulness of this tool was further proved by lack of any arch effect in PCA bi-plots (Figs 1 and 2). On each component, variable with loading ≥ 0.70 was identified as the significant variable (Iwara et al 2011;Mathur 2015).…”

“…This multi-variant technique was utilized for two objectives, i.e., (a) to find out weighting factor for various attributes with the help of variables like richness, Shannon and Weaver index, evenness, total herbage biomass, botanical species composition of climax species, RIV of climax species, carrying capacity, grazing intensity, density of unpalatable species, removal of plant parts, litter conditions, soil organic carbon, phosphorus, nitrogen concentrations, pH, electrical conductivity, and moisture content and (b) to characterize the temporal variation of land assessment parameters that includes RCS, revised SQI, HCS, and EMV. For the second objective, variables and their significant relationships were identified through the concept of component defining variables (CDV) which works through correlation coefficient (Iwara et al 2011). Student t test was applied to assess the significance of temporal factor for various quantified parameters (community and soil) and for calculated index (RCS, HCS, revised SQI, EMV, and for land status).…”

Dynamics of top-down factors with relation to ecological attributes of an endangered species Commiphora wightii

“…Squared cosines were used to link the variable with the corresponding axis and the greater the squared cosine, the greater the link with the corresponding axis. However, in order to determine basic soil, diversity (vegetation) and plant metabolites variables sustaining these interrelationships, the concept of component defining variables (CDV) which stipulates the selection and subsequent naming of variables with the highest component loading (correlation coefficient) as variables that provide the best relationships (Iwara et al, 2011;Mathur, 2012) was employed.…”

Diversidad microbiana y materia orgánica del suelo en sistemas silvopastoriles de la Región Chaqueña

“…(Hotelling, 1936) entre grupos de variables métricas, es decir, se analizaron las asociaciones entres conjuntos de variables. Este enfoque es ampliamente utilizado en el campo de las ciencias del suelo (Tan et al, 2003;Martin et al, 2005;Qi et al, 2009;Iwara et al, 2011;Silva Terra et al, 2014).…”