I'm handling communal roost of endangered Crested Ibis. The raw data indicate a strong difference between direction distribution of roost departure and arrival. I'm thinking to divide eight sections of the 360° and compare the direction distribution in each section between departure and arrival. But I'm not satisfied in this method. I'm puzzled whether there are better method to quantitatively compare the differences? Thank you for any suggestions.
From anecdotal observations I can confirm that some social birds and bats take different departure and arrival flight paths.
Presumably you have two-dimensional cardinal degree directions of Ibis departures and arrivals (per individual bird/day/season?), Obviously, using degrees as continuous or categorical variables in SDM makes no sense. Have you tried sine/cosine of the angle (degrees) between departure and arrival direction? The vertical spatial and the time dimension (day time) may also have to be considered. On the latter, depending on your data type and quantity, you may select a specific slot in absolute time or correct for daylight hours (in case the Ibis flight pass or departure/arrival depends on light (direction).
have fun
Article Reduced Dependence of Crested Ibis on Winter-Flooded Rice Fi...
"The angular Q-Q plot is proposed by Fisher (1993, Section 5.2) as an alternative graphical means of comparing the distributional shapes of two samples; but not any differences in their central directions, as both samples are centred around their median directions in the construction of the plot. The R function TwoSampleQQ below
[http://circstatinr.st-andrews.ac.uk/resources/Chap7-RCommands.txt]
produces such a plot" (A.Pewsey, M.Neuhäuser and G.D.Ruxton, 2013, Circular Statistics in R, Oxford University Press)
http://circstatinr.st-andrews.ac.uk/resources/Chap7-RCommands.txt
This free link provides graphical tool for azimuth data with adjustable bin sizes for directional (line) data and an easy interface that displays the vector mean. Why not use 10 or 5 degree bins?
http://snet.visiblegeology.com/snet.html
Dear colleagues, thank you so much for your comments. I tried but failed to explore my data. Actuary I recorded number of departure and arrival birds regarding eight 45-degree directions for each day. I can compare the difference between departure and arrival using Chi-square test for each day. But how to assess the overrall differences for a whole season? Could you kindly suggest any refferences for me? Many thanks. Dongping
It is difficult to understand why this is a problem for you. Circular, histogram statistical software packages are designed for these problems. They compile statistics for you with respect to you temporal (time-based) or spatial (polygon or map) parameters and input values. If your data re organized in a spreadsheet by day, month, or year, then the vector mean for each population should suffice. If you compare the vector mean, and percentile range for each population you have the azimuthal variation between the populations. Is this what you're looking for?
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Climate Science
- Climate Change