Yes, you can run a One-way ANOVA, but you risk increasing alpha errors. To avoid this, use a repeated-measures ANOVA or a Two-way ANOVA within group,  and ignore comparisons that are not relevant to your study.

Since the three parameters are measured on the same fish,  they are likely to be corelated. I advise you to use MANOVA insted.  Repeatation over time can be further handled by repeated measures analysis.

Thank you for your answers. 

Tukey post-hoc analyses were performed to determine which groups were different from the control group.  This test also adjusts the significance level since multiple ANOVAs were carried out. 

The issue we are having after running the mixed anova on SPSS is that the variance of the data is significant according to Levine's test. We are unsure as to the best method to use at this point to either transform the data or bootstrap. These are both analyses we are unfamiliar with. Any help or suggestions would be very much appreciated.

Kind regards

Hello!

If i understand correctly, you measure certain parameters on a continous scale multiple times on the same object.

Now IF your earlier measurement can alter or effect your future measurements, meaning the measurements are not unrelated to eachother, you have to go with repeated measures ANOVA (one-way), because it takes this very important consideration into account.

But IF your measurements are unrelated, you can use a regression model instead of ANOVA, since you are interested in a timescale. I would avoid multiple ANOVA tests on the same parameter.

The different variances are crucial, I would go with the regression model - as far as i understand measuring body lenght does not effect the growth of the fish. For later experiments i suggest to think forward and design them the way you want to analyse them.