Not sure I understand. What is the null hypothesis that you are testing for the component? If you mean hypothesis tests like the correlations are all zero, that is different. Or do you mean something like the component does not account for much variation (which means other components account for more).
Thank you for your answer. My study is related to stability analysis. So, using AMMI and GGE biplots we can assess the stability. As the biplots are based on PCA axes, will it be ok to interpret the biplot of the traits that showed non significant PCA value in ANOVA.