Inter-location Measures

From Anatrack Ranges User Guide
Revision as of 11:50, 10 November 2014 by Admin (Talk | contribs)

Jump to: navigation, search

This allows you to calculate simple measures between locations or groups of locations. You need to load a location file, choose a measure, select between which locations you want that to be calculated and then press the ‘Run location analysis’ button. The analysis will run – producing a summary of its progress and output the results to a file. What happens after the run is finished depends on the options you have selected.

If you select first to last or a location interval other than 1 then the output file will be displayed in the Statistics window.

For other options, after running the analysis a plot for the first range will appear in a separate window, and the user interface will switch to Input & Graphics and display the location file with an extra column added containing the newly calculated attribute. To show later ranges in the plot, select them in the Ranges table. The currently selected location in the Locations table is displayed in red in the plot. Running subsequent analyses on the same file will result in multiple plot windows which can be viewed at the same time. This helps to discover how and when a juvenile disperses or an animal migrates, whether the distance moved is constant all day etc.

Headings

Results are given as headings in 360 degrees, between a user-defined number of locations.

Distances

Results are given in metres either between locations, or from a focal site (e.g. a nest) to the location. When focal site to location is selected for distances the dispersal detection option is enabled.

Times

Note that this option is only enabled if your input file has location qualifying variables defining times for each location (these must have labels containing combinations of YY, MO, DD, HH, MI, SS). Results are given in minutes.

Speeds

Note that this option is only enabled if your input file has location qualifying variables defining times for each location (these must have labels containing combinations of YY, MO, DD, HH, MI, SS). Results are given in metres per hour.

Between Which Locations

span first to last

Calculate the chosen metric between the first and last location of each range. The value calculated is the sum of the result between each consecutive location in the range (with the exception of headings, which just calculates the heading between the first and last locations).

location interval

Allows you to input the location interval between which the metric is calculated. Entering 1 will result in calculations being made between all locations. For values greater than 1 the result will be the sum of the calculation between each consecutive location, e.g. if locations were taken at hourly intervals, a location interval of 24 could be used to estimate the daily distance moved.

focal site to location

Calculate the chosen metric between the focal site and each location. This option is not offered for times and speeds because the focal site doesn’t have an associated time value.

When focal site to location is selected for distances the Dispersal detection option is enabled.

Dispersal Detection

This option is only available when distances and focal site to location are selected, and when your input file has location qualifying variables defining times for each location (these must have labels containing combinations of YY, MO, DD, HH, MI, SS). This latter requirement is because the question you are asking is "when did dispersal occur?".

In this options two attribute columns will be added to the loc file, the first contains the distance from the site to the location, and the second whether the individual is classed as having dispersed yet (0=no, 1=yes). This enables easy splitting of files into before and after dispersal using modify, or estimation of home range areas before and after dispersal by using the Make Selections button in Location Analysis.

Minimum dispersal distance (m)

Set minimum distance to recognise dispersal. It is important to set a value such as 10 m for birds with initial locations in a nest, to prevent premature triggering when they leave the nest. Alternatively, if you would not classify a bird within 1000 m of the nest as a disperser, set this distance (see Walls & Kenward 1995, 1998).

Alpha for dispersal detection

none

Dispersal is classified as having occurred when the distance from the focal site is exceeded.

1%, 5% or 10%

The stochastic dispersal detection starts with the first 3 locations, and estimates their arithmetic mean centre (Ac), and their confidence limit for distance from the Ac, using the relevant probability level. It then estimates the arithmetic mean for the next n locations in the data set, where n is usually 3 (but see below), and defines the "dispersal" vector (i.e. average direction of dispersal) between these. Dispersal is recognised if the orthogonal distances of the n locations along this vector are all outside the confidence limit for the first N (i.e. 3). If not, N increments by 1, and the process repeats through the set of locations.

The use of a dispersal vector means that dispersal will not be recognised erroneously if an animal makes consecutive excursions in different directions. However, dispersal may be recognised late if the animal makes an initial movement in one direction, and then departs in another direction. The estimation is only rigorous with N+n locations where n > 3. Missing values in the dataset (recorded as -9, -9), are assumed to represent occasions when the animal was sought but not found, and was outside the confidence limit for the current N. Dispersal is, therefore, recognised with a minimum of 4 (3+1) locations if coordinates for the last location are -9, -9. It is advisable to set a minimum distance as well as a probability level. This will prevent premature detection of dispersal (e.g. when a bird leaves a nest after several records there, instead of leaving the natal area).

After running Dispersal detection, the screen will display a distance/time plot with a vertical red line just after the appropriate location.

Add Stats To Loc File

If you select the Add stats to loc file option in Output Files the new location file with the added attribute column will be saved with a new name. This enables the attribute to be used as a location qualifying variables in subsequent analyses.