A technique for evaluation of spatial data using asymmetrical weighted overlap values

• Published in: Animal behaviour Vol. 48; no. 6; pp. 1285 - 1292
• Main Authors: Smith, Andrew T., Dobson, F.Stephen
• Format: Journal Article
• Language: English
• Published: Kent Elsevier Ltd 01.12.1994; Elsevier; Harcourt Brace Jovanovich Ltd
• Subjects: Animal populations; Animal, plant and microbial ecology; Biological and medical sciences; Fundamental and applied biological sciences. Psychology; General aspects. Techniques; Methods and techniques (sampling, tagging, trapping, modelling...); Weights & measures; Data analysis; Analysis method; Spatial distribution; Methodology; Software; Mathematical method; Overlap
• ISSN: 0003-3472; 1095-8282
• DOI: 10.1006/anbe.1994.1364

Behavioural analyses often require documentation of spatial overlap of individuals or groups. Most often these data have been presented as unweighted overlaps, calculated as the percentage of home range area that is shared between neighbours. This metric does not take into consideration the frequency of use of space in the area of overlap, which often is less than in the core of a home range. A method for calculating asymmetrical weighted overlap values between neighbouring individuals or groups is described using a program written for SAS. This method yields a null expectation of the likelihood of interaction between neighbours, by taking into consideration their respective frequency of use of space in the area of overlap. A comparative analysis of a sample data set is presented using both unweighted and weighted overlap values. The results indicate that the two methods would generate different null expectations of the frequency of social interactions. The asymmetrical weighted overlap method is a valuable approach for the characterization of the texture of space use by animals and the generation of expectations for interaction frequency between neighbours.