我测量了不同地点  同一植物的 花苞片长度

得到三组数据

m<-c(89.18,91.87,110.5,85.24,83.65,78.62,94.83,92.07,96.21,59.23,83.79,78.15,87.63,91.91,93.78,85.74,100.79,63.71,57.67,86.64,84.86,92.79,70.48,76.37,94.11,87.05,90.56,84.58,93.82)
n<-c(63.68,88.89,62.90,45.27,39.42,78.86,70.06,82.06,76.41,82.76,41.97,61.42,49.45,64.42,73.85,84.15,85.08,79.95,91.57,66.24,60.89,55.58,66.26,83.57,62.15,56.01,87.21,83.37,97.86,75.80)
q<-c(91.20,86.52,59.42,100.24,84.21,84.93,102.18,59.44,74.72,91.04,86.51,61.43,67.80,91.34,90.79,83.68,88.06,66.45,89.33,80.05,105.29,81.07,84.55,95.58,63.16,78.42,91.10,76.26,94.90,78.96,82.40)

经shapiro正态性检验 m组不是正态分布，其他两组是正态性分布

比较三组的平均值是否相等

用非参数检验的话 该用哪种呢？ 一般非参数检验当大于两个样本时该怎末办呢

请高手指教