Difference between revisions of "Range Use Plots"

From Anatrack Ranges User Guide
Jump to: navigation, search
(Created page with "== Introduction == In Ranges you are able to load utilisation plots and Input & Graphics#Incremental area plots|incremental area anal...")
 
 
Line 5: Line 5:
 
== Utilisation plots ==  
 
== Utilisation plots ==  
  
Examination of utilisation plots (Ford & Krumme 1979) provides a method of deciding on the percentage of locations that define a core range. In Ranges utilisation plots display the area of estimated home range cores at 5% intervals from 20-100%. A more objective technique for excluding outlying locations to estimate a range core is available in [[Clusters|cluster analysis]], but is appropriate only for multinuclear analyses. The inspection technique is more appropriate for peeled polygons such as [[Convex Polygons|MCPs]], which are mononuclear, and not at all appropriate for ellipses because of their intrinsically smooth distribution.  
+
Examination of utilisation plots ([[Bibliography|Ford & Krumme 1979]]) provides a method of deciding on the percentage of locations that define a core range. In Ranges utilisation plots display the area of estimated home range cores at 5% intervals from 20-100%. A more objective technique for excluding outlying locations to estimate a range core is available in [[Clusters|cluster analysis]], but is appropriate only for multinuclear analyses. The inspection technique is more appropriate for peeled polygons such as [[Convex Polygons|MCPs]], which are mononuclear, and not at all appropriate for ellipses because of their intrinsically smooth distribution.  
  
 
If you have individually selected ranges displayed in the map display then just the values for those ranges will be displayed in the plots. If you choose the map display option ''display all'' then the mean values will also be displayed.
 
If you have individually selected ranges displayed in the map display then just the values for those ranges will be displayed in the plots. If you choose the map display option ''display all'' then the mean values will also be displayed.

Latest revision as of 16:38, 19 November 2014

Introduction

In Ranges you are able to load utilisation plots and incremental area analysis plots created in location analyses. In both cases, plots can be for single ranges, or to show means and limiting values for combinations of ranges. Ranges will automatically load these plots in the main window after a run has been conducted, and if on conducting the run, you specify a permanent filename you will be able to load the files into Ranges later.

Utilisation plots

Examination of utilisation plots (Ford & Krumme 1979) provides a method of deciding on the percentage of locations that define a core range. In Ranges utilisation plots display the area of estimated home range cores at 5% intervals from 20-100%. A more objective technique for excluding outlying locations to estimate a range core is available in cluster analysis, but is appropriate only for multinuclear analyses. The inspection technique is more appropriate for peeled polygons such as MCPs, which are mononuclear, and not at all appropriate for ellipses because of their intrinsically smooth distribution.

If you have individually selected ranges displayed in the map display then just the values for those ranges will be displayed in the plots. If you choose the map display option display all then the mean values will also be displayed.

Individual ranges

If there are a few locations far from the range centre, the slope of the plot is initially steep, but becomes shallower when only the core locations remain. This slope discontinuity, if present, is a useful indicator of how many locations constitute the core. Discontinuity tends to be most marked in cluster analysis, which is particularly effective for eliminating outliers. Discontinuities can also be seen in contours plotted to contain a proportion of the locations (especially for Harmonic Mean contours), but not in the smooth distributions from ellipses or contours based solely on location density.

The most appropriate use of inspection is to estimate a single proportion of core locations to be used across all the ranges, for example as the largest percentage of locations that include the core for 95% of the ranges. This will produce areas slightly less than the true core for some ranges, but variation in area is much less than if a small number of outermost core locations are excluded than if they are included. When selecting a single core value, ranges with no discontinuity (and typically a relatively small area that includes all the locations) are treated as 100% core (i.e. lacking excursive locations). In cluster analysis, 85% cores seem to exclude excursive locations effectively for a number of species. Moreover, the distribution of cores estimated subjectively in this way gave very similar results to those obtained by objective outlier exclusion (Hodder et al. 1998).

All ranges

This displays the means and data values for utilisation distributions across all ranges.

Incremental area analysis plots

With this option, you can examine how the range area changes as successive locations are added, starting with the first 3 locations. In a pilot study, this helps to decide how many locations are needed to define each range. Areas are shown as the percentage of the area that uses all the locations.

Note that all cores, and outer edges based on density estimates (ellipses and contours) may decrease in size after an initial increase. This happens when an increase in density of core locations reduces the influence of outliers. In this case, the number of locations required is that at which range areas tend to become stable.

If you have individually selected ranges displayed in the map display then just the values for those ranges will be displayed in the plots. If you choose the map display option display all ranges then the mean values will also be displayed.

Individual ranges

This gives a good impression of when most achieve stability only if all have similar numbers of locations, so it can be best to exclude ranges with relatively few locations (the reason for there being few locations may also indicate an atypical range).

All ranges

In this case, a line connects the mean area for all the ranges, and all of the data values are also shown.