QUADRAT VARIANCE MODELS

Quantitative Ecology – Spring 2008

The interpretation of the spatial pattern of continuously distributed species is influenced by the size and shape of the quadrat used to sample the population. There are several methods by which the effect of sampling unit size on variance can be evaluated using the sampling method of belt transects.

The blocked-quadrat variance (BQV) method calculates variance for different quadrat sizes by blocking or combining the areas of abutting quadrats on the belt transect. The largest block is determined by the size or length of the belt.

The paired-quadrat variance (PQV) method calculates variance for different quadrat sizes with the added effect of spacing among quadrants and blocks overlap one another.

The two-term local quadrat variance (TTLQV) method is similar to the BQV method but avoids the limitations of blocking by a power of 2 as in the BQV blocking scheme. However, as block size increases some quadrats may not be used in this method. Calculated variances are plotted for each block size or spacing regime to interpret whether the species of interest is randomly, uniformly, or aggregately distributed.

MATERIALS AND METHODS

In this experiment, we will test the effect of quadrat size on variance in continuously distributed populations. Two belt-transects, each with 30 abutting quadrats of 0.’5 m² areas will be used to simultaneously sample two plant species.

(1). Choose 2 species to study. Carefully collect 1 specimen of each species comparison and for later identification.

(2). Run a 50 m tape to designate the transect line and place a 0.5 m² square quadrat at the beginning of the transect line.

(3). Count the number of individuals of one of the two selected species that are contained within the perimeter of the quadrat. If any portion of the plant lies inside the inner edge of the quadrat include it in the count. Record the data on the data sheets provided and then repeat the count with the second species that was selected.

(4). Carefully, flip the quadrat over while holding the side furthest from the start firmly in place such that the adjacent 0.5 m² area on the transect can be sampled.

(5). Repeat steps 3 and 4 until data for 30 quadrats have been collected.

(6) Move tape indicating the transect to a new location that does not overlap the area sampled with the first transect and repeat steps 3 and 4 until 30 quadrats are called.

QUADRAT VARIANCE MODELS

Quantitative Ecology – Spring 2008

Analysis:

(1). For each of the three quadrat-variance methods (BQV, PQV, and TTLQV):

(a) Construct schemes for pooling quadrate;

(b) Calculate variances for the different block sizes; and

(c) Generate variance plots with variance on the Y-axis and block size or spacing on the X-axis.

(2). Estimate the average distance (the number of quadrate) between clumps of individuals for each species along the belt transects.

(3). Pool data for other species from other workers and perform an ANOVA using NCSS that compares the average distance between clumps among the different species. If the null hypothesis that all mean distances between clumps are equal is rejected, perform a Student- Newman|-Keuls test to determine which mean distances are statistically significantly different.

(4). Report all results in a clear, well-organized format and briefly interpret the