CCA simply shows gradients that drive the composition of ecological communities. The PERMANOVA will show if there are significant differences in species composition before and after fishing was banned. To know if water quality causes species composition changes, you would need to perform regressions. However, statistical tests can never tell us WHY events occur :)

@all Your question is a common challenge in ecology and can be difficult to address definitively. However, there are several statistical approaches that you can use to explore the possible drivers of changes in fish communities in your study site.

One potential approach is to use multivariate analyses, such as RDA or CCA, to explore the relationship between fish communities and environmental factors, including water quality parameters and the temporal component (before and after the fishing ban). This approach can help identify which environmental factors are most strongly associated with changes in the fish community over time. To implement this, you would need to split your data into two groups based on the time period (before and after the fishing ban) and then perform the RDA or CCA analysis using those two groups as the explanatory variables. Additionally, you could perform a permutation test to evaluate the significance of the temporal component in explaining the changes in fish communities.

Another potential approach is to use structural equation modeling (SEM), which allows you to test hypotheses about causal relationships among multiple variables. In your case, you could use SEM to test the relative contributions of the fishing ban and water quality to changes in the fish community over time, while also considering potential interactions between these factors. SEM can help you to evaluate the strength and direction of causal relationships among variables, but it requires more data and a well-formulated hypothesis.

Finally, it may also be helpful to use additional univariate analyses to explore the effects of individual environmental factors on the fish community, such as regression or correlation analyses. These analyses can help you to identify which environmental factors are most strongly associated with changes in the fish community and provide additional evidence to support or refute potential hypotheses.

In summary, RDA/CCA, SEM, and univariate analyses can all be useful approaches for exploring the potential drivers of changes in fish communities in your study site. It is important to choose the appropriate statistical method based on your study design, data structure, and research question. I hope this helps!