2_-_Attachment_1_Vegetation_Sampling.pdf

vegetation sampling for desert tortoise permanent plots (2023)

ATTACHMENT 1. Vegetation Sampling Methodology for Permanent Trend Plots for the Desert Tortoise (Gopherus agassizii)

Prepared by Peter Rowlands in 1979, revised subsequently (K.H. Berry)

Background

Sampling of arid and semi-arid vegetation types, as any other, is first accomplished by the delineation of whatever homogeneous stands of vegetation occur in the area (Mueller-Dombois and Ellenberg 1974). In the past, researchers have used different methods, depending upon the nature of the terrain and vegetation sampled.

McCleary and Wagner (1972) compared belt transects 1 m wide by 100 m long with 100-m line intercepts along the midline of these transects. They found that line intercepts were quite accurate in proportionately measuring the importance of the most common desert plant species as compared with results gained from their belt transects.

Unfortunately, they did not use the improved line intercept technique defined by Strong (1966) and therefore could not estimate absolute density of individual species. In addition, no line intercept technique can provide the researcher with absolute cover measurements. McCleary and Wagner’s results were also lacking in that their belt transects were too small to reveal true density of overstory elements (e.g., Yucca brevifolia). A minimum of 0.1 ha is recommended by Rowlands (1978) to accurately establish density and cover of overstory species in the arid and semi-arid Southwest.

Several authors, notably Wells (1961) and Webb and Wilshire (1979), have utilized the point-quarter technique (Curtis 1959) to measure density of desert plants. However, this method was developed to measure importance of forest trees in Midwestern deciduous forests. It can be highly inaccurate in vegetation where species form aggregates as in desert communities where many plant species (e.g., Yucca, Larrea, Coleogyne, Lycium, many species of perennial grasses) form clones and individuals are difficult to discern (Ludwig et al. 1979). Rowlands (unpubl.) has found that the point-quarter method underestimated the total number of stems by about 66% in a simulated single-species plant population whose members were arranged in a Poisson-Poisson distribution. Based on species-area curve analysis, Rowlands (1978) determined that at least 200 m2 was required to accurately sample upland Mojave Desert vegetation and that the overstory was accurately sampled by plots which acre 0.1-1.0 ha or more. The minimum sampling area increased as density of overstory plants, usually Yucca brevifolia, Juniperus spp., and Pinus monophylla, decreased.

Belt Transects

The vegetation on each plot will be sampled using two methods. Within each homogeneous plant assemblage, one permanent belt transect will be established.

Additional belt transects will be established only if, in the studied opinion of the researcher, there is more than one plant assemblage on the study site. The perennial vegetation will be sampled using a 2 x 100-m belt transect subdivided in to fifty 2 x 2-m quadrats. This belt transects will be permanently marked at the initial and ending corners with four foot sections of reinforcing bar. Annuals will be sampled using twenty-five

Berry-vegetation sampling for desert tortoise permanent plots (2004)

Figure 1: Diagram of the vegetation sampling procedure.

2 meters x Steel stakes

Perennial quadrats

Annual subquadrats

20 x 50-cm subquadrats placed in alternating corners of every other perennial quadrat, or about 4 m apart. The long axis of each of these subquadrats will be parallel with the cross borders of the belt transect (Figure 1). It is anticipated that the annual subquadrats will be sampled yearly. The location of the annual subquadrats should be alternated from year to year.

Perennial Vegetation.- Information gathered on each perennial plant species within the belt transect includes cover, density, volume, and frequency for each species. The estimated density is simply the number of plants counted within the belt transect (200 m2). These densities are expressed as numbers of plants per ha (10,000 m2 ) for each species by multiplying by 50. A plant is counted if the root area within the quadrat is greater or equal than 50 percent of the total root area of that plant. If the plant in question is a “tight clone”, that is, several to many stems arise from the ground at the base of the plant and the canopies associated with the individual stems overlap (Figure 2), include the plant of 50% of the stems are inside. Perennial bunch grasses will have to be evaluated by this method along with some mall shrubs such as Coleogyne ramosissima.

In such tight clones, all of these stems probably represent one individual plant. Where

Figure 2.

more “open clones are concerned, for example, Larrea (Vasek 1979; Sternberg 1976), and distances between clone members of the same individual plant are relatively great and canopies of members do not overlap, densities will be stem densities, not necessarily densities of individual plants. However, in such cases, clone members function ecologically as individuals since vascular connections between members break down after a period of time (Vasek 1979). Therefore, members of loose clones, where canopies

Plot Boundary

Include Shrub

Do Not Include Shrub

Include “Tight Clone”

Do Not Include “Tight Clone” of clone members do not overlap, should be treated as individual plants for data gathering purposes.

To obtain an estimate of the canopy volume for each species, measure in cm the major (greatest) and minor (smallest) diameters of the canopy and the height of each plant that was counted during the density estimation. The volume of each plant is estimated using the calculations for an ellipsoid:

V= πd1d2h where V = volume, π = 3.1416, d1 = minor diameter, d2 = major diameter, and h= height of the plant. Values are converted to volume per ha by multiplying by 50.

Frequency is then estimated by determining how many of the 2 x 2-m quadrats are occupied by each plant species, dividing this value by the total number of quadrats sampled (50), and multiplying by 100 to get percent.

Cover for each species is estimated by using the area of the polygon (ellipse) encompassed by the shrub, i.e., (π x d1 x d2). Cover for species A per ha (cm2/ha) is calculated by summing the cover and multiplying by 50. Cover of all species per ha is then calculated.

Relative values for density, cover and frequency will be obtained by summing the values for each species and dividing each species value by the totals. Figure 3 shows the format to display this information and gives computational formulas. Although not discussed above, comparisons may be made between 2 x 2-m quadrats or groups of 2 x 2-m quadrats. Such information is useful to examine the heterogeneity of the study area.

Figure 3. Data display for perennial plant information, not including the summary on cover, which should be an additional column.

Plant species

Density1

( /ha) Relative2 density

Volume3 (cm3 /ha)

Relative4 volume

Frequency5

Relative 6 frequency

A1 A2 Ai Total 100 100 100

1. Density = number of plants of species i counted in the belt transect x 50 to yield number of plants per ha.

2. Relative Density = Density of species i divided by Total density of all species combined x 100.

3. Volume = canopy volume for each plant species in the belt transect.

4. Relative Volume = Volume of species i divided by Total volume of all species combined x 100

5. Frequency = number of 2 x 2-m quadrats in which a plant species is found x 2.

6. Relative Frequency = Frequency of species i divided by Total frequency of all species combined x 100.

Annual Vegetation.-- Density estimates of annual species may not be particularly useful and are quite difficult to obtain. Therefore, an estimate of density will not be made in this work. Estimates will be made of annual cover, biomass, and frequency.

Cover will be estimated by placing over each of the 25 subquadrats a 1 x 1-cm grid and ocularly estimating the number of grid squares covered by each species. This number multiplied by 10 is the cover in cm2 per m2. This will be done for each of the 25 subquadrats and averaged over the 25 subquadrats.

Estimates of the biomass of each species will be made by clipping the above ground portion of each annual plant in all 25 subquadrats, drying the material in the air for several days, and weighing the material. Biomass will be expressed as g per m2 by multiplying by 10 and averaging over the 25 subquadrats.

Frequencies of each species will be estimated in a manner similar to the procedure for the perennial vegetation. Presence will be noted for each of the twenty-five 20 x 50cm subquadrats. This number will then be divided by the total number of subquadrats sampled (25) and multiplied by 100 to get percent. (This is equivalent to dividing by 4).

Relative values for each species will be computed for cover, biomass, and frequency. Figure 4 shows the format for display of this information as well as appropriate formulas to obtain values for the table.

Figure 4. Data display for annual plant information

Plant Species

Cover1

(cm2/m2) Relative2

Cover

Biomass3

(g/m2 ) Relative4

Biomass

Frequency5

Relative6

Frequency

Non-grass7

Density ( /m2 )

Relative8

Density

A1 A2 Ai Total 100 100 100 100

1. Cover = average cover on 25 subquadrats for each species

2. Relative Cover = Cover of species i divided by Total cover of all species combined x 100.

3. Biomass = dry weight biomass of each plant species averaged over the 25 subquadrats.

4. Relative Biomass = Biomass of species i divided by Total biomass of all species combined x 100.

5. Frequency = number of 20 x 50-cm subquadrats in which a plant species is found x 4.

6. Relative Frequency = Frequency of species i divided by Total frequency of all species combined x 100.

7. Non-grass Density = number of non-grass annual plants of species i counted in all subquadrats multiplied by 10 to yield number of plants per m2. This is then averaged over the 25 subquadrats.

8. Relative Density = Density of species i divided by Total density of all species combined x 100.

Literature Cited

Curtis, J.T. 1959. The Vegetation of Wisconsin. U. Wisconsin Press, Madison. 657pp.

Ludwig, J.A., M. Ortiz, and J.C. Owens. 1979. An evaluation of the plotless wandering quarter method for estimating the density of aggregated organisms. (Abstract).

Bull. Ecol. Soc. Am. 60:138.

McCleary, J.A., and K.A. Wagner. 1972 A study of three xeric plant communities.

Southwestern Natur. 17: 169-180.

Mueller-Dombois, D., and H. Ellenberg. 1974. Aims and methods of vegetation ecology. John Wiley & Sons. 547pp.

Rowlands, P.G. 1978. The vegetation dynamics of the Joshua tree (Yucca brevifolia

Engelm.). Ph.D. Dissertation. Univ. of California Riverside. 192pp.

Strong, C.W. 1966. An improved method of obtaining density from line transect data.

Ecology 47: 311-313.

Webb, R.H. and H.G. Wilshire. 1979. Recovery of soils and vegetation in a Mojave

Desert ghost town. Submitted to J. of Arid Lands

Wells, P.V. 1961. Succession in desert vegetation on streets of a Nevada ghost town.

Science 134: 670-671.