Measures of spatial pattern for counts

A - Papers appearing in refereed journals

Perry, J. N. 1998. Measures of spatial pattern for counts. Ecology. 79 (3), pp. 1008-1017.

AuthorsPerry, J. N.

SADIE (Spatial Analysis by Distance Indices) is a new methodology to detect and measure the degree of nonrandomness in the two-dimensional spatial patterns of populations. It applies the same principles to data in the form of maps as to data in the form of counts at specified locations, but with different techniques. This paper considers data in the form of counts such as occur commonly in ecology. For such data the method has an advantage over traditional approaches that measure only statistical variance heterogeneity, because all the spatial information in the sample is used. Two indices and associated tests are reviewed, one based on the total distance of the sample from a completely regular arrangement, the other from a completely crowded arrangement. A new diagnostic plot is presented to aid interpretation. Results from some artificial data are studied to survey the properties of both indices for defined patterns of clustering. Indices based on the distance to regularity are powerful at detecting aggregation when several clusters are present; those based on the distance to crowding have the power to detect aggregation only when a single cluster is present. Methods are presented to estimate the typical cluster size and intercluster distance, suitable for data from sample units in the form of a contiguous grid. Examples are given for cyst-nematode field data and plant virus disease.

Year of Publication1998
Journal citation79 (3), pp. 1008-1017
Digital Object Identifier (DOI)doi:10.1890/0012-9658(1998)079[1008:MOSPFC]2.0.CO;2
Open accessPublished as non-open access
Funder project or code102
Project: 041001
Project: 041587

Permalink -

Restricted files

Publisher's version

Under embargo indefinitely

10 total views
0 total downloads
2 views this month
0 downloads this month