[mike-selmo]

This is actually a really interesting question. I imagine the complexity of the model to really accurately predict this would be practically unattainable currently.

You are introducing a new element to an environment without knowing how it would react to it. You could attempt to predict based on predators or food sources from its original environment and compare them to closest matches (if any) to the new environment.

This would make so many wild assumptions though that minor errors could result in cascading effects in a direction the model never predicted.

Doing some googling and found that there is the I = PAT equation which is just for human influence and is based on an extremely simplistic model. I think the nuances and details needed for an accurate ecosystem model would be several orders of magnitude more complex.

A model that could connect events like reintroducing wolves to Yellowstone with stopping erosion and other similar effects would be pretty incredible.