Understanding the multitude of factors and their interactions involved with animal movement is complex at best. However, without some attempt at building such a model it is difficult to fully understand movement patterns, let alone test hypotheses that will lead to new insights.
Many types of models have been developed. Overall, model selection depends on the questions to be addressed, temporal and spatial scales of concern, and data available.
General Movement Models
A recent paper by Nathan et al. (2008, see Figure 1) summarizes the components best incorporated into movement models by asking the following questions.
- Why move? Is there a physiological or psychological state that requires moving?
- How to move? What is required? Walking, flying, or swimming?
- When (initiation and cessation) and where (direction and position) to move? What is a species ability to orient in space and time?
- What are external restrictions to movement? Resources, landscapes, predators, etc.
These factors have been examined in four types of models:
- Cognitive processes
- Optimality theories
- Comparisons with random
Spatial Movement Models
Models become more complex when they incorporate a spatial component. These would address the following types of data.
- Observations of occurrence. Where did the animal move? What was the probability of movement or survival? Presence/absence? Abundance?
- Independent geographic locations. Point locations, repeated sampling of the same individual, and sample movements over a long time period
- Serial location. Movement tracks, frequent resights, serial set of observations, and an estimated sets of observations.
Resource Selection Spatial Movement Models
Another approach would be to examine resource selection among individuals in a spatial array. This generally involves incorporation of GIS into models and the following components:
- Sampling at the individual or population level
- Estimating resource use or nonuse
- Quantifying resource availability
- Relating resource use to availability
Utilization distributions are models used in resource selection analyses that summarize resource use as a continuous and probabilistic process (probability density function) whereas other resource selection analyses techniques depend upon correct classification of habitat at exact locations. These models require the ability to appropriately define home ranges.
- Discrete habitat categories (e.g., forest, grassland, management unit)
- Continuous habitat attributes (e.g., % canopy cover, distance to human activity)
- Combination of the two
Home Range Models
Every data set and organism is unique, thus choice of a home range model needs to be based on a comprehensive evaluation of movement and to place it in a broader context (ecology, physiology…of your species). There are many types of home range analyses that can be chosen based on the questions to be addressed and the data available.
Edited by Matthew Johnson (USGS Forest and Rangeland Ecosystem Science Center, firstname.lastname@example.org) and Dylan Kesler (University of Missouri, email@example.com)
- Boulet, M. and D.R. Norris. 2006. The past and present of migratory connectivity. Ornithological Monographs61:1-13.
- Nathan et al. 2008. A movement ecology paradigm for unifying organismal movement research. Proceedings of the National Academy of Sciences of the United States of America 105:19052-19059.
- Ramenofsky and J.C. Wingfield. 2007. Regulation of migration. Bioscience 57: 135-143.
- White, G.C. & Garrott, R.A. 1990. Analysis of wildlife radio-tracking data. Academic Press, San Diego, California, USA.