Chapter 2 Preamble
This is a book about Ecological Modelling. Since we are talking about a field that links Models and Ecology, we need first to think about what ecology is, and what models are. And so that is where we begin.
Ecology is the science that studies the relations between all living beings, namely how they influence each other and how they are influenced by the abiotic factors of the surrounding environment. To be able to describe those relationships there is a natural requirement to quantify the nature and strength of these relationships, and hence the need for quantitative methods becomes obvious.
Models are representations of reality. This is a fundamental aspect that one should never forget, they approximate reality, but they are not real. All models are wrong, but some are useful, is perhaps one of the most famous quotes by G.P. Box, and it is a fundamental aspect that I constantly return to, and that I tend to impose upon my students. Because unfortunately, just like artists, statisticians have the bad habit of falling in love with their models. Unlike in the celebrity world, in science there are usual worst consequences than a broken heart or an empty wallet.
Actually, as I think about this book I realize that the “all models are wrong” is in itself wrong. There are in fact many true models, and we will use them extensively in this book. Those are the models that we simulate inside a computer. And those are perhaps some of the best models, since they are both real and extremely useful. By simulating a reality we can control all the aspects of a small world, and hence we can evaluate methods performance, we can illustrate theoretical results and we can show that things do what they say on the tin, or not!
I refer to Ecological Modeling as the art of being able to conceptualize the ecological world as a set of relations, and to be able to translate those relations into equations. Then, to be able to make these equations functions of parameters that hopefully have a meaning, and then, if lucky, to be able to obtain data that contain information about the relevant parameters of the model. To think about where in the data is information about a given aspect of a system is a crucial step. Perhaps even harder, is to perceive what are the filters that information went through, via a sampling process. And then to be able to introduce observation models to undo those filters and allow reliable inferences to be based on noisy data. Not all models do that, and most models in this book do not separate observation and process model. Those models are arguably the more useful, but you need to understand simpler models before going into those more sophisticated and hence much harder to implement models. By estimating the parameters of the models one is then able to make some informed statements about useful descriptions of the natural world, like relating the weight of a fish with the temperature of the water it lives in or the number of eggs in a nest to the nearest source of food or body of water. Beyond the scope of this book, if one is really really lucky, those models and the results they provide might then be used by others to make management decisions that will ultimately be useful to the ecosystem under study.