Background: I work for a blueberry grower. The accumulation of heat units is important in understanding and predicting the development of the plant over a season.

We have 5 different weather stations across the farm. While there is no statistically significant difference (ANOVA) in the daily temperature or heat units when the sites are compared, there does appear to be a difference in the accumulation of heat units over time (see image). This translates into potentially different rates of development over the season.

What is my best approach to determine if the accumulation of heat units is significantly different (P<0.05) at different sites?