Canonical ordination analyses of Hellinger-transformed prevalence <i>Meloidogyne</i> spp. data on cultivated olives with respect to environmental variables.

<p>The response variables are the fitted explanatory variables from the four data set (i.e. climate, soil, topography and agronomic management) by forward selection procedure. A) Ordination plots of the redundancy analyses used to investigate the relationships between prevalence of <i>Meloidogyne</i> spp. and explanatory variables. Ecological predictors are represented on the plots as agronomic practices related with irrigation regimen (Irrigation) and cover vegetation on alley (Cover on alley), and soil loamy sand texture (Loamy sand texture) on olive orchards. The first canonical axis (RDA axis 1) explains 85.9% of species-variables relationships, and the second axis (RDA axis 2) explains 11.2%. B) Trend-surface analysis of the canonical axis 1 of the RDA analysis. Spatial structure variation related with agronomic practices including Irrigation (<i>P</i><0.001) and Cover on alley (<i>P</i><0.001) variables. C) Trend-surface analysis of the canonical axis 2 of the RDA analysis. Spatial structure variation related with soil properties including soil loamy sand texture (<i>P</i><0.05). For trend-surface analyses, fitted sampling site scores from RDA are plotted on geographical coordinates. Sampling sites with the same colours show similarly trends of relationships between Hellinger-transformed prevalence of <i>Meloidogyne</i> spp. data and explanatory variables of cultivated olives; symbol size indicates the level of similarly (i.e. small symbols for low similarity, and large symbols for high similarity. Interpretation of the spatial variation with respect to explanatory variables was performed using regression analyses of the two canonical axes on the environmental variables after normality tests.