A Short Introduction to Modeling Populations and Species
Models in Population Ecology
Population ecology is concerned with understanding how populations of plants, animals and other organisms change over time, and from one place to another, and how these populations interact with their environment. This understanding may be used to make forecasts of a population's size, distribution or other properties, to estimate the chances that a population will increase or decrease, or to estimate the number of individuals that may be harvested while ensuring a high probability that similar harvests will be available in the future. Thus, the focus of any given study in population ecology may be motivated by very practical considerations in fields as diverse as fisheries harvest regulation, wildlife management, pest control in agricultural landscapes, water quality monitoring, forest harvest planning, disease control strategies in natural populations, or the protection and management of a threatened species.
Population ecology, as mentioned above, is concerned with changes in the abundance of organisms over time and over space. Abundance and how it changes can be described by words such as "abundant" or "rare", and "fast" or "slow", but population ecology is fundamentally a quantitative science. To make population ecology useful in practice, we need to use quantitative methods that allow us to forecast a population's future, and express the results numerically.
Frequently, the need to make forecasts leads to the development of models. A model is a mathematical description of the population. A model may be as simple as an equation with just one variable, or as complex as a computer algorithm with thousands of lines. One of the more difficult decisions in building models (and one of the most frequent mistakes) concerns the complexity of the model appropriate for a given situation, i.e., how much detail about the ecology of the species to add to the model.
Simple models are easier to understand, and more likely to give insights that are applicable in a wide range of situations. They also have more simplistic assumptions, and lack realism when applied to specific cases. Thus they cannot be used to make reliable forecasts in practical situations. Including more details makes a model more realistic, and easier to apply to specific cases. However in most practical cases, available data are limited and permit only the simplest models. More complex models require more data to make reliable forecasts. Attempts to include more details than can be justified by the quality of the available data may result in decreased predictive power and understanding.
The question of the appropriate level of complexity (i.e., the trade-off between realism and functionality) depends on:
- characteristics of the species under study (e.g., its ecology),
- what we know of the species (the availability of data), and
- what we want to know or predict (the questions addressed).
Even when detailed data are available, general questions require simpler models than more specific ones. For example, models intended to generalize the effect of one factor (such as variation in growth rate) on a population's future may include less detail than those intended to forecast the long-term persistence of a specific species, which in turn, may include less detail than those intended to predict next year's distribution of breeding pairs within a local population.
The purpose of writing a model is to abstract our knowledge of the dynamics of a population. It serves to enhance our understanding of a problem, to explicitly state our assumptions, and to identify what data are missing and what data are most important. If the data required for building the model are plentiful, and if our understanding of the dynamics of a population are sound, we may use the model to make forecasts of a population's size or behavior.
The above introduction to modeling is extracted from the textbook Applied Population Ecology, which discusses various types of models with different degrees of complexity. Most realistic models of populations and metapopulations incorporate factors such as age structure, density dependence, environmental variation and spatial structure. Such models are often too complicated to solve analytically, but they can easily be implemented as computer programs.
See:
Avoiding Common Mistakes in
Population Modeling
Bird Modeling Studies at Applied
Biomathematics
