A Quantitative Method for the Description and Classification of Range Vegetation.

(1)

A Quantitative Method for the Description

and Classif ication of Range Vegetation’

CHARLES E. POULTON AND E. W. TISDALEZ Professors of Range Management, Oregon State College and University of Idaho, respectively

Infroducfion

Numerous contributions by

ecologists of an earlier period as well as those of our contempo- raries have enhanced the understanding of range vegetation, but few of these studies have provided the detail which is desir- able for the evaluation of site

potential in modern manage-

ment. The notion appears still prevalent that the comparatively

low production of range lands

makes detailed attention to their

ecology uneconomic. This atti-

tude fosters the study of com-

plexes rather than the basic

units, or individual ecosystems (Tansley, 1935) that comprise them. Data thus assembled are

confounded by unrecognized site

differences and defy understanding from the outset, particularly

in regard to interpretation of

plant succession or range condi-

tion and trend (Hanson, 1950;

Dyksterhuis, 1958).

Growing awareness of the

need for a more detailed and

clearer understanding of the

range resource makes it important that we consider how ecology can contribute to the devel-

opment of improved guides for

the evaluation of range sites re- gardless of current condition. In spite of the amount of work di- rected at the problem of range site quality and its component

factors, general agreement has

not been reached on such ques- tions as: how best to gain an understanding of site, what kinds of information are required, how much emphasis should be placed on vegetation and how much on soils and climate. Granted differences in intensity and details

of study, it would seem that

f

sound application of ecological

principles should result in an

overall approach of broad application and usefulness.

Such ecological understanding is essential if we aim to. judge the potential and limitations of range lands and to classify them

for management and improve-

ment purposes. This is particularly true in areas of variable

topography and soils such as

characterize much of the western range. Under these condi- tions valid extrapolation of results from one area to another becomes more difficult than on

comparatively uniform areas.

Even on areas of apparent uniformity, unrecognized diversity frequently exists and creates a

danger of “lumping” vegeta-

tional units that are different.

Until individual ecosystems are

recognized in research and in

management, ecology will not

make its potential contribution

to practical resource manage-

ment.

It is the purpose of this paper to present a research approach

which has proved satisfactory

for developing an understanding of range sites and hence for clas- sifying range lands. This method seeks to define the basic units of vegetation and to determine their relation to soils and other factors of the environment. The advantages of a combined study

of vegetation and soils are

stressed. Work of the authors in

the sagebrush-grass region of

Oregon and Idaho is used as the basis for the discussion.3 Pre-

liminary results from this in-

vestigation suggest the kinds of

information needed to classify

non-arable lands and to differ-

13

ment and appraisal purposes. Review of Liferafure

Early ecological research in

the western United States em-

phasized the recognition of

major vegetational types. The

sagebrush-grass community was

separated by vegetational and in some cases by gross soil characteristics from the bunchgrass and salt-desert shrub types (Weaver, 1917; Clements, 1920; Shantz and Zon, 1924). As in the case of many other major communities,

subsequent investigations have

not yet provided an adequate

analysis of the various units evident in this extensive complex- to say nothing of explanations

for their existence (Billings,

1951).

Most of the early studies were either non-quantitative or lacked

adequate geographical scope to

enable correlation of data from area to area within the region. During this early period much attention centered on the value of individual plant species as indicators of land potential and use (Shantz, 1911; Kearney, et al., 1914; Clements, 1920; Kelley,

ITechnical Paper No. 1287 of the 0 r e go n Agricultural Experiment Station.

2The authors are indebted to many persons for help in the development and. use of the approach and meth- ods described. Particular thanks are due to R. E. Eckert of the Agri- cultural Research Service and to M. A. Fosberg and M. Hironaka of the University of Idaho who were directly involved in the develop- ment and refinement of the meth- ods, to L. D. Calvin of Oregon State College for statistical guidance and help in the early development stages and to R. Daubenmire of Washington State University and D. N. Hyder of the Agricultural Re- search Service for review and val- uable criticism of the manuscript. 3The specific project in which this

(2)

1922; Aldous and Shantz, 1924; Shantz and Piemeisel, 1924).

Continuing interest in the re-

lation between vegetation and

site factors is evident in numerous, more recent studies. Some of these were concerned with the relationships between vegetation

zones and great soil groups

(Dunnewald, 1930; Thorp, 1931;

Martin and Fletcher, 1943; Spils- bury and Tisdale, 1944). These studies served to focus interest on vegetation-soil relationships and to utilize knowledge of soil development and profile characteristics. Increased impetus was given this approach as more soil

scientists became interested

(Gardner and Retzer, 1949;

Olson, 1952; Wieslander and

Storie, 1952). More detailed studies of regional vegetation have been reported by a number of

workers (Daubenmire, 1942; Tis-

dale, 1947; Billings, 1949). These studies considered both seral and

permanent vegetation and pro-

vided more knowledge concern-

ing the plant communities.

Other research has been concerned with the correlation between individual or small groups of soil factors and plant species

or communities in attempts to

clarify the sit.e concept. A pio- neer study in this field was that

of Hanson and Whitman (1938)

on grassland vegetation in North

Dakota. Other studies of this

general type include those of

Billings (1950) ; McMinn (1952) ; and Gates, et. al. (1956).

Progress in the combined

study of vegetation and soils has been made possible by the clari- fication of concepts for studying vegetation, reflecting in part the influence of European ecologists

(Hanson, 1950, and Becking,

1957). Equally important has

been the increased knowledge

available regarding the morphology, genesis, and composition of range soils. The combination of these two disciples in the interpretation of range site is illus- trated by the published work of

Anderson and Fly (1955) and

Anderson (1956). Increased

awareness of the values of con-

currently studying vegetation

and soils has set the stage for advancing the usefulness of ecol-

ogy in range improvement and

management.

Methods

The procedures adopted by the authors are designed to effect an

integrated vegetation-soil ap-

proach to the understanding of

range site. Basically this in-

volves a study of the individual

ecosystems comprising the veg-

etation and concomitant inter-

pretation of both vegetation and

soil. The interpretive unit is

taken as the habitat-type. It is the basic environment unit of the landscape for classification and management purposes. The habitat-type is an abstract unit that may be defined as the collective area which is capable of support-

ing the same relatively homo-

geneous climax plant association

(Daubenmire, 1952). Thus the

habitat-type denotes a specific

ecosystem-the fundamental

(3)

unit of effective environment or site. It is hypothesized, therefore,

that habitat-types may be ex-

pected to differ not only in their climax vegetation and soils, but also in their seral vegetations

which develop as a result of

management treatments or dis-

turbances such as grazing, fire, insects, artificial revegetation, etc.

The initial objective of this method is, therefore, to define

and characterize each habitat-

type of a given region in quantitative terms. The methodology used is guided by two major considerations. First, it is essential that the geographic variation and patterns of the plant cover and soils be measured ade-

quately. This demands the in-

tensive study of many locations thus creating two major prob-

lems in research: (1) the re-

quirement of abnormally high

budgets for travel, and ’ (2) the balancing of precision of individual measurements against the need for an adequate sample in terms of numbers of stands ex-

amined. Secondly, the methods

used must attempt to sample essentially all features of the vegetation and soil since, at the outset, it is impossible to predict relevant or limiting factors or to identify correctly all diagnostic characteristics of each habitat- type.

A combination of methods was developed which appears to provide an adequate sample at each location and which is sufficiently rapid to enable the study of

many stands. In the soils in-

vestigation, the approach in-

volves a detailed study and description of the profile at each site and collection of adequate samples for laboratory analyses. These two areas of investigation, vegetation and soils, have been

worked into a systematic pro-

cedure for the description and

classification of habitat-types in range areas. These procedures consist of the following essential steps.

Reconnaissance

Make a thorough reconnais-

sance of the general area selected for study to obtain a preliminary idea of the nature and variability of the vegetation and soils. The area included in one study can be of any size, but preferably it should be fairly extensive so that the full range of variation in the vegetation and soils typical of the ecological region may be encountered. In the

course of the reconnaissance,

preliminary information is

sought regarding the kinds of

permanent vegetation present in the area and the seral communi-

F--- 30 FEI

1: 2 II 109 8 7 6 5 4 3

_!

REFERENCE CORNEW Fl

f

ties which may have been pro-

duced by such influences as

heavy grazing, fire, or insect epidemics. Advantages are real- ized in gathering this information by systematic note-taking.

Marginal-punch, hand-sort cards

with spaces provided for perti-

nent vegetation and soils data

have been used by the authors

with considerable success. The

use of these cards facilitates reference to the data and greatly

speeds summarization and com-

parison, particularly when the

data have been obtained over an extensive area during a period of several years.

I i,j,

ORIGIN

7 E

MACROPLOT

FIGURE 2. The macroplot and its subsampling design.

w

z

3

25’

JX2FOOT DBSERVATIO

PLOT

46’- _---.

36’- _--.

3l’- _-w-s

26’- B-w

21’- ---.

16’0 ---.

TRANSECT

8 BELT

(4)

Quantitative Study Selection of Sfudy Loclafions Selection of relatively undisturbed sites for intensive study is the next major step. It devel-

ops naturally from the recon-

naissance phase which yields hypotheses as to the habitat-types comprising the study area. This selection of study locations is subjective to a considerable extent; and it involves the biases as well as the training, experience, and judgment of the workers involved. The objective is to sample the natural units of the vegetation as they have devel-

oped through evolution, plant

migration, succession, and com-

petitive sorting of the species. The approach adopted by the authors has been to set up a few readily observable criteria based

upon the reconnaissance data

and to consider all locations

which satisfy these criteria-re- gardless of how suitable they in-

itially may seem for detailed

study. In the sagebrush-grass region, the authors have followed the hypothesis that the permanent vegetation of the area con- sisted of an overstory of shrubs, usually species of Artemisia, and

a moderately dense understory

of perennial herbs, chiefly

grasses. Any area meeting these simple criteria must be considered as a possible study location. Rejection of a stand for intensive study must be based on lack of

sufficient area for adequate

sampling or on some abnormal- ity of vegetation or soil (usually

related to disturbance) which

can be substantiated from the

reconnaissance data. Even where potential study locations are re- jected, the reconnaissance notes for them are retained and sum-

marized. Such sites may later

prove valuable in the study of succession or because of the light they throw on the effect of specific factors in the ecosystem.

They are not uncommonly

picked up later for intensive

study as a better understanding

FIGURE 3. Determining shrub density on the four-foot belt transect.

of the vegetation is developed. If numerous stands represent- ing essentially equivalent, permanent vegetation are available, selection of study sites could be

by random drawing. Relatively

uniform geographical distribu-

tion of study sites over a region is also an important factor to be considered, and thus restricted

randominzation may be desir-

able. In the experience of the authors, it is usually feasible to study virtually all available stands of near-climax sagebrush- grass vegetation. In these studies, five stands represent the minimum group used to describe and characterize a habitat-type; and it is usually found that little

new information is discovered

by studying more than ten samples of each.

Each location is examined

carefully for homogeneity of

vegetation-based on species

composition, relative dominance, and vigor-and stratified if this

appears justified. Many stands

will not require stratification because of the diversity of the vegetation, but the investigator must continually watch for vegetation patterns with in stands. The authors proceed on the hy-

pothesis that if a repetitive vegetation pattern can be seen in the species composition, relative dominance, or relative vigor, this pattern has causes which can be determined only if each variant is represented by its own sample. This is particularly true where the vegetation pattern coincides with observable soil and relief characteristics. Stratification is required in Figure 1, for example, because of consistent differences between the vegetation in the foreground and that in the background. The height and density of big sagebrush (Artemisia trident&a) is sharply decreased in the foreground. The dominant grasses are bottlebrush sitanion (Situnion hystrix) and sandberg bluegrass (Pou secundu). In the

background, in contrast, blue

bunch wheat grass, (Agropyron

spicutum) and thurber needle-

grass, (Stipu thurberiunu) are

dominant beneath the taller

shrubs. The difference in the

foreground is interpreted as

being primarily due to a fundamental site difference associated with the following obvious soil

characteristics: more bare

(5)

friable soil surface in the foreground. This stratification tech- nique, where applicable, defines the population to be sampled.

Vegetation Data

If a stand is stratified, a

sample plot is established in each of the subdivisions. Each variant is sampled by a macroplot 50 by 100 feet in size, within which four 50-foot transects are located

in restricted random fashion.

Along these transects, 4-foot

belts and l- by 2-foot observation plots are located (Figure 2).

A list of all identifiable species is prepared for the macroplot to provide constancy data. Plants whose identification is in doubt are given a number and collected for future study. Dif- ferences in the period of seasonal

development plus the ephemeral

nature of many spring-flowering species in sagebrush-grass vegetation make it impossible to obtain a complete species list at any one time. At least two observations are needed during one growing season.

The foliage intercept of shrub species is measured along each transect line marking the reference side of the four-foot belt.

The density (number per unit

area) of shrub species is re-

corded by size classes for each

four-foot belt (Figure 3). The

mean maximum height of ma-

ture shrubs is determined as a site index by measuring the tall- est mature plant of each shrub

species in each five-foot com-

partment of the belt.

The basal area percent and frequency of all herbs are determined on the l- by a-foot observation plots (Figure 4). Data in percent basal area for all herbaceous, vascular plants, percent

ground cover for mosses and

lichens, litter, stones and gravel, and percent bare ground are determined with the aid of estimate rings equivalent to two and five percent of the observation plot area. Basal area rather than foliage cover of the herbaceous

species was adopted as a standard of measurement in order to

minimize year to year differ-

ences due to climatic variation

and/or grazing use and season

to season differences due to vegetation growth. In certain cir-

cumstances a foliage area or

crown-spread cover method may

be satisfactorily used (Dauben- mire, 1959).

A set of “ground rules” must be prepared to insure uniformity

of estimate. For example, the

presence of lobes and discon-

tinuities in shrub crowns, irregu- larities in the bases of bunch- grasses, the special case of mat- formers, the basal area of rosette

plants, single-stemmed grasses

and annuals all require rules to insure consistency of estimate. Since these rules are determined largely by the characteristics of the plants themselves, the rules will vary with the vegetation being studied. Careful documen- tation of these rules is a basic requirement of this method.

Yield determinations are not

made as a part of the regular

sampling procedure. After the

ecological interpretation is made,

two or more representative ex-

amples of each habitat-type are

selected for a replicated plot

study of herbage yield. These

yield studies are designed to ex- tend over a period of at least five years to provide a measure of variability among years as well as among habitat-types.

Soils Data

The soil at each macroplot location is examined from pits and

described by genetic horizons

following standard procedures in the Soil Survey Manual (U. S. Dept. Agric., 1951). This exam- ination is made at an intensity sufficient to allow correlation of the soil at each macroplot location at Phase of Soil Type level. Following preliminary t e s t in g along the perimeter of the macroplot, one or more pits are dug at points just outside the macroplot boundary. The soil descrip- tions are made or checked by

qualified soils personnel. A

sample from each soil horizon is also collected for laboratory analysis. Bulk density, perma-

nent wilting percentage, and

moisture equivalent are deter-

mined for representative profiles to enable calculation of important soil moisture characteristics for each habitat-type.

The procedure adopted for laboratory analysis is to make certain determinations such as pH, salt content, organic matter, per

FIGURE 4. The area estimation.

(6)

cent nitrogen, and carbon: nitrogen ratio on all sites. Additional chemical analyses, including ex- change capacity, base saturation percentage, and exchangeable ions, are made only for selected profiles, usually one or two from each habitat-type. Further chemical analyses are made only if their importance is suggested by the data from these selected sites.

Supplementary Data and Information

The location of each site is re- corded in detail and the general site denoted by a metal fence post or other prominent and fire proof marker. A designated cor- ner of the macroplot, the ends of each transect, and the location of the soil pits are also

marked with smaller metal

stakes to make possible re-sampling of the same area.

value in relation to specific sites or habitat-types unless related to

factors such as seasonal soil

moisture supply or effect of temperature on growth period. Soil

moisture and temperature data

collected throughout several sea-

sons on one example of each

habitat-type would constitute

important environmental infor-

mation now lacking-information possibly of greater value

than conventional climatic rec-

ords.

Information regarding the past history of the site is obtained wherever possible. Frequently it is possible to learn much about

the grazing history from the

owner or land administrator, but

information regarding fires and

disturbances other than those of recent date usually has to be de- duced largely from the vegetation itself.

Photographs, both mono-

chrome and color, are taken of the general site and of details of

the vegetation and soil. All

photos are taken as photopoints with permanently ‘.marked locations. One of the most useful types of photograph obtained to

illustrate the habitat-type has

been a color photograph showing the exposed soil profile and a sectional view of the vegetation immediately above it.

Table 1. A Simplified Example of a Species Association Table Adapted From Unpublished Oregon Data (Eckerf, 19571.1 Actual Dominance Data Are Not Used Because the Purpose Is Only fo Illusfrafe fhe Tab- ular Layout.

Species Dominance Ratings by Habitat-Types __________. Artr/Agsp Artr/Fsd Arar/Feid Arar/Agsp

The physiography and other

gross features of the habitat are

noted in detail according to

standard descriptive terms and

criteria. For example, the fol-

lowing observations have been

found useful: slope percent, di-

rection of slope, position on

slope, classification of land form and macro- and micro-relief, ele- vation, and nature of soil form- ing material.

Climatic data are obtained

where possible, but few weather records are available for stations

in or immediately adjacent to

areas of undisturbed sagebrush- grass vegetation. Interpolation of climatic data from sites even a few miles away is likely to lead to error in regions of complex

topography. Even where data on

precipitation and temperature are available, these have limited

Poa secunda

Artemisia tridentata Agropyron spicatum

Festuca idahoensis Lappula redowskii Phlox diffusa

Eriogonum sphaerocephalum Lithospermum ruderale Poa cusickii

Collinsia parviflora Erigeron linearis Artemisia arbuscula Chrysothamnus viscidiflorus Phlox hoodii

Haplopappus stenophyllus Trifolium gymnocarpon Tetradymia canescens S ymphoricarpos rotundifolius Ribes cereum

Lithophragma bulbiferum Leptodactylon pungens Chrysothamnus nauseosus Aster scopulorum

Chaenactis douglasii Eriastrum filifolium Footnote

*D=important, s=subordinate, +=&en present but very subordinate lunpublished Ph.D. thesis.

Time Requirements

The time required to sample vegetation and soils by the procedures outlined varies with the

habitat-type and with the dis-

tance between study locations.

On the average, about three man days are required to complete the field work on one location and travel to the next if only one macroplot sample per location is involved. An efficient crew consists of an experienced staff member and one student

helper. Two such teams can

work together to advantage, especially where one concentrates

on the vegetation sampling and

the other on soils. The total number of locations that can be sampled during a given season depends largely on the length of time during which the vegetation remains green and in suitable condition for study. In the

sagebrush-grass region this pe-

riod is usually limited to about

D* D D D

D D

S

D

S”

D

S

+* +

:

D

S

+ S

S +

S S

+ +

D D

S +

S S

(7)

six weeks. Soil studies can be continued beyond this period if necessary, but there are definite advantages to making the study of vegetation and soils at the same time.

Interpretation of Data

Following compilation of the

field and laboratory results, the stand data are grouped on the basis of similarities in vegetational, soil, and other characteristics. Effective analysis of this

kind of information requires a

logical sequence of steps in the development of association tables designed to show the similarities and differences in the vegetation of the stands sampled (Table 1).

These tables are developed by

arranging the species and the

data concerning them in groups of similar distribution and dominance. This ecological arrange- ment is the basis for interpretation of the data and for showing the natural groupings with their similarities and differences. It is essential to consider all characteristics of the vegetation sampled in this tabular analysis

and interpretation. Mere pres-

ence of a given species in two groups of stands, for instance,

may have no significance; but

the fact that the frequency or basal area is consistently greater in one group of stands than in another may constitute the dif- ferentiating criterion. The prep- aration of these association tables is an essential, yet not in- frequently overlooked, first step in the analysis and interpretd- tion of community ecology data. These tables enable one to see the degree to which the multiple-factor criteria of the abstract classification units are dis- tinct or tend to intergrade and thus to represent a continuum.

Following the vegetational interpretation, a somewhat similar

approach is used for the soil

characteristics. When the soils

interpretation is completed at

Series or Phase of Soil Type *

level, the soils and vegetation

classifications are c a r e f u 11 y cross-checked by preparing vege-

tation/soils association tables

(Table 2) . Similarities and differences in tentative groupings are then re-examined in the light of both kinds of information together with any accessory data available on climate, past history, etc., which may throw light on the ecological interpretation.

Statistical analysis to determine the reliability of the hy-

pothesized differences and sim-

ilarities among the descriptive c r i t e r i a of the habitat-types should follow the development This does not imply universal

and careful appraisal of the vari- coincidence of units of vegetation and soil. Many contributing

ous association tables. Unfor-

reasons for disparity may be

identified: classification stand-

tunately,

ards for vegetation and soil may differ; the interpretation of the

many workers have

classification units of soil and/or

attempted to short-circuit this

vegetation may not sufficiently

approach their natural units;

step by jumping directly to bio- obscure vegetational differences

resulting from unrecognized or

uninterpretable successional

status may obscure the t r u e

vegetation-soil relationship; fac-

tor compensation may result in

apparently identical vegetation’s occurring on two or more tax- onomic soil units; or significant

vegetational differences m a y

occur on two soils which are not

readily distinguishable on the

basis of their morphology and

other obvious characteristics. In spite of these sources of “con- fusion” and uncertainty, the im-

portant consideration is that

when done concurrently the soils interpretation adds a major criterion to the vegetation interpretation and vice versa. When in- consistencies appear in the relationships, the need for re-ex- amination of both vegetation and soils interpretations is indicated -often with mutually beneficial results.

Table 2. An Example of a Vege- tation -Soil Relafionship Table Taken From Unpublished Ore- gon Data (Poulfon, 1955).1

Habitat-Type

Soil Type

Pose/ Artr/ Agsp/

ERO Agsp Pose Bakeoven very

stony loam x Uncorrelated

Brown Lithosol x Ritzville silt loam X

Ellisford silt loam X

Uncorrelated Chernozemic

Lithosol X

Walla Walla silt

loam X

Shaniko silt loam X

1 Unpublished Ph.D. thesis.

metric and mathematical manip- ulation as though it were the key

Before investigations of this to, rather than a tool in, ecologi-

kind can be considered complete, cal interpretation. In the opinion and experience of the writers,

the interpretations should be

statistical tests are inappropriate up to this point. Until the association tables are developed,

checked in the field to see if they on_ does not know what one

will withstand objective re-ex-

should compare or what hy-

potheses to test-it is mere prob- ing in the dark. As a minimum,

amination. Actual use of the in- succeeding

formation in range resource in- statistical analysis should determine the degree of variability and central tenden-

ventory and in condition and

cies within groups of classification criteria. The confidence one

can have in differences amoq

trend appraisal is, of course, the

habitat-types should be deter-

mined as a check on the descrip-

best check against its validity. tive criteria isolated. In addition to the obvious characterization

(8)

can, however, be used as a field test of the validity of the ecological interpretations and hypothe- s i z e d vegetation-soil relation-

ships. The earlier data from

“potential study locations” as

well as from additional new

stands are often useful in this respect. An additional test of the

validity and usefulness of the

habitat-type concept is the extent to which each can be recognized by others on a purely ob- servational basis using criteria

developed from the intensive

study. In the tests of this nature made to date by the authors, it has proved possible to recognize habitat-types not only in rela-

tively undisturbed areas, but

also to a considerable extent in

disturbed vegetation. F u 11 e r

recognition in disturbed areas

must, in some cases, await in- v e s t i g a t i 0 n of successional patterns in each habitat-type and further studies of vegetation-soil

relationships. Preliminary tests

indicate that these same meth-

ods, with modification, m a y

prove suitable for sampling seral vegetation.

As the interpretation of these

synecological studies is com-

pleted, they make a significant

and essential contribution to

autecology. The association table also provides the orientation for etiologic ecology through experi-

mentation by making possible

the recognition and statement of

hypotheses regarding the inter-

relationships among individual

species and between the species

and measurable characteristics

of its environment.

Associations a m o n g species

and between species-environ-

ment characteristics may be

tested biometrically from the

data available. Such attempts

(Gates, et al., 1956) are more likely to be successful where the

habitat-types have been previ-

ously and accurately determined.

Each habitat-type may be ex-

pected to differ not only in its climax vegetation and soils, but also in its reaction to various

treatments or disturbances such as grazing, fire, insects, artificial

revegetation, etc. Thus the

study of individual species-envi-

ronment relationships will tend

to be confounded if habitat-type differences a r e unrecognized. Species tend to have different ecologies, different phenotypic expression, in different environ-

ments. For example, the same

species may respond differently to the same kind of treatment on different habitat-types. Needle-

and-thread grass (Stipa comata

Trin. & Rupr.), for instance, is a major component of undisturbed vegetation in the big sagebrush/ needle and thread grass association on sandy soils in the Pacific Northwest; but it occupies a successional position in other habitat-types on finer soils, even

under similar climatic condi-

tions. In terms of grazing management this species acts as a decreaser in a sandy-soil habitat- type and an increaser on finer- textured soil. Attempts to study its autecology or environmental requirements in the field with- out recognizing this habitat-type

difference would likely prove

futile.

The authors are convinced that research of the type outlined in this paper provides a way to tie

together the findings of plant

ecology and soil science and thus to obtain their full contribution towards a better understanding of the range resource-a degree of understanding ess tial to in-

tensive management. While de-

!

signed primarily fo non-for-

ested range lands, recent experience has indicated that this approach can prove equally useful on other types of non-arable lands and for ot er land management interests.

P

Numerous ecological studies

have shown that native vegetation in a relatively undisturbed condition is an excellent index of land potential and a sound basis for classification of g r a z i n g

lands. More recently, soils investigations have shown similar

value-especially when corre-

lated with intensive studies of

the vegetation to provide for

soils interpretation in land management. Few studies are available where a combined investi-

gation and interpretation of

vegetation and soils has been

made with sufficient intensity

and correlated over a sufficiently large area to define the individual ecosystems (habitat-types) on a combined vegetation and soils basis. Such natural units constitute the logical basis for range

management and improvement

and for the generalization of

other research results.

A method is described which consists of extensive reconnaissance followed by i n t e n s i v e study of the vegetation and soils in numerous stands of relatively

undisturbed vegetation. Use is

made of a large macroplot sub- sampled by line and belt transects and by small observation plots. The concomitant study of vegetation and soils is an essential part of the procedure. The methods described appear to be sufficiently intensive to accom- plish the objective and to permit the sampling of a large number of stands. Analysis of the plant data in association tables and a

comparable treatment of

soils information enables

the the recognition of natural groupings among the stands. The reliabil-

ity of these groupings can be

tested by the degree of associ-

ation between plant and soil

data, by statistical analysis, and by application of the results in the field.

The method appears useful on non-arable lands in general, as well as for other multiple-use interests besides range management.

LITERATURE CITED

ALDOUS, A. E. AND H. L. SHANTZ.

(9)

ANDERSON, E. WILLIAM. 1956. Some soil-plant relationships in eastern Oregon. Jour. Range Mangt. 9: 171- 175.

ANDERSON! K. L. AND C. L. FLY. 1955. Vegetation-soils relationships in Flint Hills bluestem pastures. Jour. Range Mangt. 8: 163-169.

BECKING, R. W. 1957. The Zurich- Montpellier school of phytoso- ciology. Bot. Review 23: 411-488. BILLINGS, W. D. 1949. The shadscale

vegetation zone of Nevada and eastern California in relation to climate and soils. Am. Midl. Nat. 42: 87-109.

1950. Vegetation a n d plant growth as affected by chemically altered rocks in the western Great Basin. Ecol. 31:62- 74.

1951. Vegetation zonation in the great Basin of western North America. In: Les bases ecologiques de la regeneration de la vegetation des zones arides. Union Internationale des Sciences Biologiques Serie B. No. 9: 101-122. CLEMENTS, F. E. 1920. Plant indica-

tors, the relation of plant communities to process and practice. Carnegie Inst. Washington Publ. 290. 388 pp.

DAUBENMIRE, R. F. 1942. An ecological study of the vegetation of southeastern Washington and ad- pgcent Idaho. Ecol. Monog. 12: 53-

.

1952. Forest vegetation of northern Idaho and ad j ace n t Washington and its bearing on concepts of vegetation classification. Ecol. Monog. 22: 301-330.

1959. A canopy coverage

method of vegetational analysis. Northwest Sci. 33: 43-64.

DUNNEWALD, T. J. 1930. Grass and timber soils distribution in the Big Horn Mountains. Jour. Am. Sot. Agron. 22: 577-586.

DYKSTERHUIS, E. J. 1958. Ecological principles in range evolution. Bot. Rev. 24: 253-272.

GARDNER, R. A. AND J. L. RETZER. 1949. Interpretative soil classification: timber, range, and water- sheds. Soil Sci. 67: 151-157. GATES, D. H., L. A. STODDART, AND C.

W. COOK. 1956. Soil as a factor in- fluencing plant distribution on salt-deserts of Utah. Ecol. Monog. 26: 155-175.

HANSON, H. C. AND W. WHITMAN. 1938. Characteristics of m a j o r grassland types in western North Dakota. Ecol. Monog. 8: 57-114.

1950. E c o 1 o gy of the grassland. II. Bot. Rev. 16: 283-360. KEARNEY, T. H., L. J. BRIGGS, H. L. SHANTZ. J. W. MCLANE. AND R. L. PIEMEI~EL. 1914. Indicator significance of vegetation in Tooele Valley, Utah. Jour. Agric. Res. 1: 365-417.

KELLEY, A. P. 1922. Plant indicators of soil types. Soil Sci. 13:411-424. MARTIN, W. P. AND J. E. FLETCHER.

1943. Vertical zonation of great soil groups on Mt. Graham, Ari- zona. Univ. Ariz. Tech. Bull. 99. MCMINN, ROBERT G. 1952. The role

of soil drought in the distribution of vegetation in the northern Rocky Mountains. Ecol. 33: 1-15. OLSON, 0. C. 1952. The soil profile

as an aid to range management. Jour. Range Mangt. 5: 124-128.

SHANTZ, H. L. 1911. Natural vegetation as an indicator of the capa- bilities of land for crop production in the Great Plains area. U.S.D.A. Bur. Plant Ind. Bull. 201.

AND R. L. PIEMEISEL. 1924. Indicator significance of the natural vegetation of the south- western desert region. Jour. Agric. Res. 28: 721-801.

AND R. ZON. 1924. Natural vegetation, grassland and desert shrub. In: Atlas of Am. Agric. The physical basis of agriculture. U.S.D.A. 29 pp., illus.

SPILSBURY, R. H. AND E. W. TISDALE. 1944. Soil-plant relationships and vertical zonation in the southern interior of British Columbia. Sci. Agr. 24: 395-436.

TANSLEY, A. G. 1935. The use and abuse of vegetational concepts and terms. Ecol. 16: 284-307.

THORP, J. 1931. The effects of vegetation and climate upon soil profiles northern and northwestern Wyoming. Soil Sci. 32: 283-297. TISDALE, E. W. 1947. The grasslands

of the southern interior of British Columbia. Ecol. 28: 346-382.

U. S. DEPT. AGRIC., SOIL SURVEY STAFF. 1951. Soil Survey Manual. U.S.D.A., Agric. Res. Adm. Hand- book Number 18. 503 pp.

WEAVER, J. E. 1917. A study of the vegetation of southeastern Wash- ington and adjacent Idaho. Univ. of Nebraska. The University Studies 17: 1-114.

WIESLANDER, A. E. AND R. E. STORIE. 1952. The vegetation soil survey in California and its use in the management of wild lands for yield of timber, forage, and water. Jour. Forestry 50: 521-526.