 So, one of the reasons why ecological forecasting and decision support go hand in hand is that decisions are all about the future and ecological forecasts are designed to predict the future. That's why ecological forecasting is so important and has many opportunities to be included as part of decision and management. They help us to predict information that's critically important for us to be able to manage our environmental systems more effectively. I think it's really useful to understand the components of a decision so we can understand how scientific information fits into them. Decisions have three major components to it. The first is uncertain scientific facts and information or ecological forecasts. The second are individual or stakeholder values. What do we care about? What are our goals and objectives? And then the third are constraints. These can be environmental constraints, social constraints, and oftentimes with environmental management decisions, policy or regulatory constraints. When we talk about decision support, we're not only talking about tools like ecological forecasts and scenarios and various modeling approaches that allow us to integrate science into decision making, we're also talking about processes. These are processes that allow us to work together, scientists and decision makers to really understand the information that needs to be brought to bear to make better decisions. That means that when we think about ecological forecasts, I think it's really important to figure out ways that we can co-produce ecological forecasts. That means we want to figure out ways to work with stakeholders so that scientists and stakeholders are in the same room helping to think through and define the problems and make sure that we're producing forecasts that are relevant and useful. Which means that we're really talking about a model where we're not just trying to push out information once we have a forecast that we think is relevant and useful and hope that it's useful for other people and managers. We're also not trying to design forecasts so that they're only responsive to a very narrow set of decisions and that we're defining them based only on stakeholder decision needs. Ideally, what we're doing is we're pushing the frontiers and developing forecasts that might be useful for a range of different decisions and applications. We're also working with stakeholders to understand the decisions and objectives that they care about. So as we're designing forecasts, we're not just making interesting predictions, we're making predictions that relate to specific types of management decisions that are relevant, and we're also thinking about how we make projections that allow us to test different interventions, different management options, and understand what might happen in the future. That means one of the most powerful ways that ecological forecasts can be used as a laboratory world, an experimental world where we can do anything we want to it and understand what will happen. This is what Hollings talks about in his book on adaptive management, is that we can make inferences, we can understand what would happen, and we can use models as a way of developing large scale experiments where it would be unethical to perturb various ecosystems, and instead we can perturb the model and understand what may happen as a result of our actions. So we've talked up to this point about some of the components and opportunities for ecological forecasts, but I think it's also useful to understand how we break down and think about making really complex, hard decisions like environmental management decisions. And so one of the acronyms that might be useful to remember is PROACT. What this stands for is problem, objective, alternatives, consequences, and trade-offs. If you're using Mike Dietz's ecological forecasting book, you'll see this in the decision support chapter. In that chapter he has a conceptual model that breaks down the PROACT steps. One of the things that he highlights is that in addition to alternatives, we're oftentimes thinking about scenarios so that we can understand what might happen in the future so we can constrain the boundary conditions that we need for our models so that we can perturb them based off of the same boundary conditions. In addition, when we talk about consequences, I'm thinking about consequences for the full range of objectives that we care about. However, in ecological forecasting, oftentimes the ecological forecaster is most interested in what consequences we can forecast that relate most specifically to those particular models or ecological systems that they care about. One of the things that's important to know about the PROACT approach is that it's not a linear process. Ideally, this employs adaptive management. So even though you may see conceptual models that really lay it out as problem, objective, alternatives, consequences, and trade-offs, if you're doing this approach, you would ideally be thinking about adaptive monitoring, improving your forecast intuitively over time, and how you use that information to feed in to understand whether or not the management actions you took are actually working, and if they're not what actions you might want to take and how you might be able to use the models to help you assess different approaches. So let's break down each of the PROACT steps. A lot of this is described in much more detail in a book called Structured Decision Making by Robin Gregory and for those that are interested in taking a deep dive into how a decision analyst thinks about constructing decision models that use ecological forecasts, I recommend you look at that reference to really understand the various components and how you can improve your ecological forecast for those decision models.