=, =t'

" " ~'

~"

4 ~ ,-4-

~' 4" 4~

~,

.4 ,d'

A' 4·' 4~ 4:

4' A' ,Al ~"

A) jJ)

Al A\'

• • ..

.. .)

•

'ft

'. \~.

Exercise 11

Populationand Community Structure:

Quadrat SampHngTechniques

A number of techniques are available for obtaining quantitative information about the composition and structure of populations and communities. The most widely applicable technique, however, is that of sampling with quadrats or plots of standard size. Quadrat sampling techniques can be adapted for almost all plant communities, and for communities of sessile or sedentary animals as well.

PROCEDURE

The details of quadrat sampling. including the size, shape, number. and arrangement of the sample plots, must be determined for the particular type of community being sampled and on the basis of the type of information desired. A good discussion of the considerations involved in the design of quadrat sampling procedures is given by Goldsmith et al. (1986).

SIze and Shape of Plots

Plot size should be based on the size and density of the plants being sampled. Plots should be large enough to contain significant numbers of individuals. but small enough tbat plants can be separated, counted, and measured without confusion leading to duplication or omission of individuals. For plant communities, suggested plot sizes are 0.1-1.0 square meter for herbaceous vegetation, 10--20 square meters for shrubs or saplings up to about 3 meters in height, and 100 square meters for forest trees,

Plot shape influences the ease of laying out plots and may affect the efficiency of sampling. In communities of low profile, circular plots can be laid out very easily by the use of a center pole and a freely rotating radius line. With a radius line consisting of a tape measure, circular plots of any desired size can be defined quickly. With regard to efficiency, some studies suggest that elongate rectangular plots furnish a more accurate analysis of the composition of a stand of vegetation than an equal number of square plots of the same area, especially when the long axis of the plot is oriented parallel to environmental gradients in the area being sampled. Radii and rectangular dimensions for plots of various sizes and shapes are given in table 11.1.

Number and Arrangement of Plots

The number of plots sampled, as a minimum, must be sufficient to turn up the bulk of the species present. This number may be determined in a semiobjective manner by plotting a species-area curve as the sampling is being carried out. This curve consists of the cumulative species total plotted against the number of samples taken. A species-area curve usually rises sharply at first, since the first samples reveal many new species, but eventually levels off. indicating that additional samples reveal few new species. An adequate sample number usually should

Colinvaux (1986): 378-382

Krebs (1984): 157-167,438--446 Smith (.1980): 587-588,666---670. Smith (1986); 202,204-207

69

____ ~~ ___ .r _
l'
" I
I i
'Iii,
~ i. 70 Population and Community Structure: Quadrat Sampling Techniques
r'
.-
, Table 11.1
I,:
~ :' Dimensions for Sampling Plots of Various Areas and Shapes
I',
I"
! ~ ,
"
i:L Circular Square Rectangular Rectangular Rectangular
"t
I, Area (Radius) (Side) 0:2 Sides) 0:5 Sides) 0: 1 a Sides)
: ~I:
:;:
'I; 0.56 1.00 0.71 X 1.41 0.44 X 2.20 0.32 X 3.16
I,'
':1
,'1 2 0.80 1.41 1.00 X 2.00 0.63 x 3.16 0.45 x 4.47
'II
::; 3 0.98 1.73
,ij 1.22 X 2.44 0.78 X 3.86 0.55 X 5.48
ii,
~:! 4 1.13 2.00 1.41 X 2.82 0.89 x 4.45 0.63 X 6.32
~~ ~
,,: 5 1.26 2.24 1.58 X 3.16 1 . .00 X 5.00 0.71 X 7.07
iij
'j 10 1.78 3.16 2.24 x 4.47 1.41 X 7.07 1.00 X 10.00
,j 20 2.52 4.47 3.16 X 6.32 2.00 X 10.00 1.41 X 14.14
"
i!
:1 30 3.09 5.48 3.94 x 7.88 2.45 x 12.25 1.73 X 17.32
: ~
40 3.57 6.32 4.47 X 8.94 2,.83 X 14.15 2.00 x 20.00
50 3.99 7.07 5.00 X 10.00 3.16 X 15.81 2.24 X 22.36
100 5.64 10.00 7.07 X 14.14 4.47 X 22.36 3.16 X 31.62
200 7.98 14.14 10.00 X 20 .. 00 6.32 X 31.62 4.47 X 44.72
300 9.77 17.32 12.25 X 24.50 7.74 x 38.70 5048 X 54.77
400 11.28 20.00 14.14 X 28.28 8.94 X 44.70 6.12 X 63.24
500 12.62 22.36 15.81 X 31.62 10.00 X 50.00 7.07 X 70.71
1000 17.84 31.62 22.36 X 44,72 14.14 X 70.71 10.00 X 100.00 be well into the latter portion of theeurve. Adequacy of sampling may be evaluated further by cruising the stand after sampling has been completed and noting the number and general abundance of species not encountered in the sample plots.

Sampling locations within a stand may be chosen systematically or randomly. A systematic sampling procedure requires an even spacing of plots throughout the stand. To achieve this pattern, plots may be located at uniform intervals along evenly spaced transects through the stand. A systematic arrangement of plots is relatively easy to achieve and is thought by some workers to give a closer approxima tion of the true composition of stands of vegetation (Greig-Smith 1983). However, for statistical analyses of sampling data to be valid, the plot locations must be selected randomly. One of the easiest ways to locate plots randomly in the field is to establish a baseline along one side of the sampling area. Pairs of random numbers thencan be selected as coordinates of sampling locations, one number indicating the distance along the baseline, the other distance into the area. Awbrey (I977) describes a technique for locating random points in an area of almost any size. This tech.nique uses a pocket calcula tor to convert rectangular-grid coordinates to polar coordinates, which can be located quickly with respect to a base point with the aid of a compass and measuring tape.

Sampling P,rocadure and Data Analyals

Once sample plots have been laid out, several types of quantitative data can be obtained for the speciespresent.

For many kinds of plants and animals, the number of individuals is the most important item of data. For some plants, and even some animals, an arbitrary definition of what constitutes a single individual may be needed. Criteria for inclusion or exclusion of individuals occurring on the edge of the plot must be defined. For example, plants with rooted bases lying more than halfway inside the boundary may be counted and measured as if they lay completely inside, and plants lying more than halfway outside may be completely excluded. The area covered by individuals often is another important data item, especially for plants. For large woody plants, the diameter or circumference of the trunk may be measured and the basal area (cross-sectional area of the trunk) obtained from table A.7 or table A.8 (see Appendix). For smaller plants, the diameter of the crown foliage may be measured and the areal

e ~

• • • • .,

.. .-

• ..

~

Population and Community Structure: Quadrat Sampling Techniques

71

coverage of the crown obtained from table A.S. Individual basal-area or coverage values may be recorded according to quadrat number in a notebook or data sheet, where they will serve to indicate both the number and size of the_ individuals of each species encountered. In some cases, still other quantitative values, such as biomass, number of tillers (plants such as grasses), or numbers of flowers and fruits may be desirable.

In many cases, the measurement of cover values for individual organisms is impossible or impractical. Cover can be recorded, however, as estimates of the percent of the plot covered by each species. Usually this is done by assigning cover-class estimates (Goldsmith et al. 1986). Several scales for ranking cover have been suggested; one of the most common is the following:

Cover Class Range o,f Percent Cover Midpoint
0-1% 0.5%
2 1-5% 3.0%
3 5-25% 15.0%
4 ~!5-50% 37.5%
5 SO-75% 62.5%
6 i'5-100% 87.5% Considering the difficulties of obtaining accurate measurements of the diameters of irregular canopies of many plants, cover estimates obtained in this manner often are as accurate as more time-consuming direct measurement.

In summarizing quadrat data, density, dominance, and frequency values can be determined for each species.

Density refers to the number of individuals per unit area, dominance to the basal area or crown coverage per unit area, and frequency to the fraction of sample plots containing the species. For a particular species, these values may be expressed in an absolute form, or in a relative form that shows the percentage that the species value is of the total for all species. Relative values for density, dominance, and frequency may be combined into a single Importance Value, which reflects these three somewhat different measures of the importance of the species in the community. These various values are calculated with the following equations:

density =

number of individuals area sampled

density fo:r a species X 100

total density for all species

dominance = total of basal area or area'! coverage values area sampled

relative density =

relative dominance =

dominance for a species X 100

total dominance for all species

number of plots in which species occurs

total number of plots sampled

frequency value for a species X 100

total of frequency values for all species

relative density + relative dominance + relative frequency

frequency =:

relative frequency =

importance value =

Since two of the above relative values, rather than all three, may sometimes be used to obtain an Importance Value, some ecologists prefer to calculate the Importance Value as the average instead of the total of the available values.

Summarized data can be recorded in table 11.2. Quadrat sampling techniques may be used in connection with studies of ecological problems such as succession, zonation, or aspection. Studies also may be carried out to compare results obtained with quadrats of different size or shape. Results obtained with quadrat sampling also may be compared with results obtained with other sampling techniques.

~". '

"

I

IT

Table 11.2

Summary of Community Analysis by the Quadrat-Sampling Technique

Date Locality Stand (Number or Type) _

Observer Name Quadrat Size Number of Quadrats _

Density Relative Dominance Relative Frequency Relative Importance
Species ( ) Density ( ) Dominance Frequency Value




















Total ?J ~ ~

"" .,

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ""'1 '"'1

'"" ~

~ ~ ~ .,

~ ~

" ,

~

!!it-. ~ ..

Population and Community Structure: Quadrat Sampling Techniques

73

CLASS ACTIVITIES

Quadrat sampling can be carried out in a field location selected by your instructor, or by use of the map of a desert plant community-a rich alluvial fan environment in Borrego Valley, San Diego County, Califomia-vincludedat the back of this manual. For map sampling, plot coordinates can be obtained from a random-numbers table (Appendix, table A.I) and located by reference to the meter scales along the sides of the mapped plot. Clear plastic discs, squares, or rectangles of different sizes can be used to represent quadrats.

Comparisons can be made of the community analysis obtained with different procedures; some possible com-

parisons include:

-Equal numbers of identical quadrats located randomly versus systematically

-Equal numbers of quadrats of the same size, but different shape

-Quadrats of different size, but adjusted in number to sample the same area

In addition, the two subareas of the map---the wash and non-wash areas=-also can be compared.

DISCUSSION

Describe conditions under which species importance rankings based on density, dominance, and frequency values may give very different results. What is the value of converting these absolute values to relative values? What is the reason for combining the three relative values into a single Importance Value? Between what limits may the Importance Value range?

SELECT'ED BIBLIOGRAPHY

Awbrey, F. T. 1977. Locating random points in the field. Journal of Range Management 157-158.

Cain, S. A., and G. M. DeOliveira Castro. 1959. Manual of vegetation analysis. Harper & Row, New York. New York, USA.

Goldsmith. F. B., C. M. Harrison, and A. J. Morton. 1986. Description and analysis of vegetation. Pages 437-524 in P. D. Moore and S. B. Chapman, editors. Methods in plant ecology. Blackwell Scientific Publications, Oxford, England.

Good, R. E., and N. F. Good. 1971. Vegetation of a Minnesota prairie and a comparison of methods. American Midland Naturalist 85:228-231.

Grieg-Smith, P. 1983. Quantitative plant ecology. 3d ed. Blackwell Scientific Publications, Oxford, England. Jackson, M. T., and R. O. Petty. 1971. An assessment of various synthetic indices in a transi tional old-growth forest.

American Midland Naturalist 86:13-27.

Kershaw, K. A .• and J. H. Looney .. 1985. Quantitative and dynamic ecology. 3d ed. Edward Arnold, London, England. Kuchler, A. W., and I. S. Zonneveld. 1988. Vegetation mapping. Kluwer Academic Publishers, Hinham. Massachusetts, USA.

Lang, G. E., D. H. Knight, and D. A. Anderson. 1971. Sampling the density of tree species in a species-rich tropical forest. Forest Science 17;395--400.

Lindsay, A. A., J. D. Barton, Jr., and S. R. Miles. 1958. Field efficiencies of forest sampling methods. Ecology

39:428-444 .

Payandeb, B. 1970. Relative efficiency of two-dimensional systematic sampling. Forest Science 16:271-276. Phillips, E. A. 1959. Methods of vegetation study. Henry Holt and Co., New York, New York, USA.

Rice, E. L. 1967. A statistical method for determining quadrat size and adequacy of sampling. Ecology 48: 1 047- 1049.