significance difference in more than 2 groups in r

I have precipitation data for five years. I need to know is there a significance difference between amount of precipitation? what test should I use? note that each year I have different samples for example year 2000 I only measured for 80 days but year 2001 for 300 days.

I saw in book that first they checked the normality with this code:

shapiro.test(residuals(aov(precipitation ~ years)))

what is aov and residuals? Is that a correct way to check the normality?