Hi All,

I have a dilemma that I have been trying to find the answers myself but cant seem to find anything to solve my problem.

I have conducted an experiment on fish to determine the effects of temperature on gut transit rates at different times 10 temporal points (1,2,3 hours eg). I have used an inert marker to detect these rates and quantified using an equation. My experimental design consisted of 6 replicates for each time point total of 60 fish at one temperature (10 and 20 Degrees) total 120 fish.

I initially intended to conduct a Tukey's test to determine the difference between gut sections at the different temperatures and the different time points, but I am unsure that the homogeneity of variance assumptions will not be met due to the uneven sample size. I could go down a parametric test route but unsure if there are any other options available to me. I also intended to conduct a multivariate general linear model but unsure if that is suitable as well.

I have also found the data to not be normally distributed and have natural log transformed the data which has caused the data to be more normally distributed.

Any suggestions would be greatly appreciated.