Hello,

I have collected the numbers of migrating fish (round 25 species) for the last 3,5 years every day. I have also collected environmental data like water temperature, water flow, moon phase and more (15 different factors) every day.

Now I want to find out, which of the environmental data affects the number of migrating individuals for each species significant. Normally I would go for a simple correlation and regression modell, but the problem here is, that the suggested correlations are not linear. Another problem is, that the environmental data among each other are not always independent. So for example, the I have high numbers of mirgrating fish of species A, when the water temperature is between 10 and 15°C for the first time in spring. When I have the same water temperature in autumn, the are no migrating species.

Another species (let's call it species B) is only migrating in high numbers in early march. For the rest of the year there have been almost no individuals of species B.

Which statistical process could I use here?

So at the end I want to have a result like this: The migration of species A affected significantly by water temperature, day of the year and watwer flow. The moon phase doesn't affect the migration significantly. The numbers of migrating indiviuduals are highest when the water temperature is between 10 and 15°C between day 76 an 98 of the year and when the water flow is between 30 and 50 m³/s.

Thank you very much!

Best wishes,

Tobias