contributed by Karen Colson and Jason Karl
Description and Uses
Similarity index is a comparison of the current vegetation (in terms of kinds, proportions, and amounts) on an ecological site to what the site is capable of producing at its reference state. According to the BLM Ecological Site Inventory technical reference, “A similarity index determines how closely the current plant community resembles either the potential natural community or some other reference community.” Similarity index is expressed as percentage of the reference community that is currently on an ecological site. These percentages can, in some cases, be construed as successional stages (see table below).
Successional status inferred from similarity index on an ecological site.
Reproduced from BLM Inventory and Monitoring Technical Reference 1734-7 – Ecological Site Inventory
|Similarity Index||Successional Status|
|77-100%||potential natural community|
To calculate the similarity index, the weight (in pounds per acre) of each species (or species) group is recorded for a site. The allowable production of a species or species group cannot exceed that of the reference community. The similarity index is then determined by dividing the total allowable production across all species by the total production for the reference community. The BLM Ecological Site Inventory technical reference describes procedures for calculating similarity index based on composition as well.
While similarity index is typically used to measure departure from reference state for ecological sites, it can be used to define departure from any vegetation community or desired state that can be defined in terms of its composition, proportions, and amount of vegetation.
Advantages and Limitations
The advantage of similarity index is that it is relatively quick and easy to calculate and provides a good, single number index of how close or far to a management target a site is. One significant limitation of this approach, however, is that it reduces many aspects of ecological functioning of a site down to a single number and may obscure important changes taking place. Consider, for example, Table 9 from Habich’s (2001) Ecological Site Inventory manual.
Table reproduced from Habich (2001), p47
In this example, the overall similarity index for the mesquite-short grass community is 265/665=39.8%. However, there is much more mesquite production in this site that would be expected at reference state. The similarity index methods, however, limits the observed mesquite production to the maximum expected for the reference state. In this example, important changes may have taken place (or be taking place) that have caused production of mesquite to be so much higher than expected.
Another limitation is that comparisons to a reference state may not always be appropriate. In this example above, loss of native grasses (supported by the low observed production of bluegramma) and increases in mesquite abundance may have pushed this site past an ecological threshold to a new state. If this is the case, then comparing the current site to the ecological site’s reference state may no longer be a valid comparison because the reference state may no longer be achievable without significant investment of restoration.
- Habich, E.F. 2001. Ecological site inventory, Technical reference 1734-7. Bureau of Land Management. Denver, Colorado. BLM/ST/ST-01/003+1734. 112 pp. Available online at: http://www.blm.gov/nstc/library/pdf/1734-7.pdf.
- NRCS. 2003. National Range and Pasture Handbook. Natural Resources Conservation Service, Grazing Lands Technology Institute. http://www.glti.nrcs.usda.gov/technical/publications/nrph.html.
Technical and Application References
Similarity index is one method for judging the condition of rangelands against a reference state. Other methods include:
These methods do not necessarily judge departure from a reference community in the same way that similarity index does, but they provide alternative methods for judging condition according to a set definition.
Conceptually, similarity indices have been developed and applied to many different aspects of ecology apart from rangeland management. Examples include similarity of diets between two species, or similarity in habitat configurations between populations in high-quality vs. low-quality habitats.