Wednesday, November 7, 2007
270-3

Using Nested Quadrats to Measure Pasture Plant Communities.

Julien Winter, Carignan Gaëtane, Ralph Martin, McLean Nancy, and Alan Fredeen. PO Box 550, Nova Scotia Agricultural College, Nova Scotia Agricultural College, Truro, NS B2N 5E3, Canada

The occurrence of plant species in pastures is highly scale-dependent. Plant species occur over wide range population densities from ubiquitous plants that form the sward “matrix”, to uncommon plants that are sparsely distributed. To compare two pastures, we should try to detect plants around 50% of the time in randomly dropped quadrats. For matrix species this could require 10 cm2-sized quadrats, common species 10 m2 quadrats, and uncommon species 100 m2 quadrats (Critchley et al., 1998). The usual methods for recording pasture plants include cover estimated in square quadrats of around 1 m2, or point quadrats that records the plants that intersect points along a line. These methods are adequate for measuring the structure of the sward matrix, but they operate at very limited spatial scale. Nested quadrats including sizes from 1, 10, 100, 1000 cm2, and 1, 10, 100 m2 have been used to measure species-area curves: the increase in the number of species with increasing area searched. We used blocks and transects of nests to measure both species-area curves and the frequency of individual species. Nests that were in contact with each other were added together to get areas of 200, 400, and 800 m2. Transects of 1 m2 and 10 m2 sized-nests were used for greater sensitivity at those scales, and to analyze for aggregated plant distributions. The population density of individual species was estimated by calculating a species abundance index. This involved ranking the quadrats from 7 to 1 with increasing size, then finding the average ranking at which a specie was observed. Species with a high SAI occurred in smaller quadrats. This allowed us to compare pastures for species composition. The SAI was converted to plant population density, if the plants were not strongly aggregated.