The effect of the factor “crop” (wheat, switchgrass or miscanthus)
on Collembola community parameters was tested by
means of GLMs. Data were log-transformed prior to analysis
when necessary to ensure normal distribution and homogeneity
of variance, except for the data expressed as percentages
which were arcsin transformed. Significant differences between
means were tested at the 5% confidence level using the
Tukey HSD test. The BrayeCurtis index was calculated with
the free PAST software version 2.02 [18] to analyse faunistic
dissimilarity between sites.
All analyses, except when specified, were performed using
the STATISTICA software package (ver. 10.0, StatSoft, Tulsa,
StatSoft 1984e2013).
The effect of the factor “crop” (wheat, switchgrass or miscanthus)on Collembola community parameters was tested bymeans of GLMs. Data were log-transformed prior to analysiswhen necessary to ensure normal distribution and homogeneityof variance, except for the data expressed as percentageswhich were arcsin transformed. Significant differences betweenmeans were tested at the 5% confidence level using theTukey HSD test. The BrayeCurtis index was calculated withthe free PAST software version 2.02 [18] to analyse faunisticdissimilarity between sites.All analyses, except when specified, were performed usingthe STATISTICA software package (ver. 10.0, StatSoft, Tulsa,StatSoft 1984e2013).
การแปล กรุณารอสักครู่..
