Trend evaluation is customarily associated 
with condition evaluation. However, a very im- 
portant distinction must be noted. Trend is 
evaluated on a specific site, a site which has its 
own unique enironment. Condition guides, on 
the other hand, are based upon sampling many 
sites from which numerical data are statisti- 
cally evaluated. Thus, condition guides are 
mathematical abstractions representing only 
averages of site characteristics and modal en- 
vironment potentials—they are not a true mea- 
sure of actual condition or potential on a spe- 
cific site. The use of condition guides in trend 
interpretation must be limited to suggesting 
site potential and approximating range condi- 
tion as a guide for assuming that upward trend 
is possible or that downward trend has oc- 
curred. The guides should not be used as a 
measure of trend because they cannot measure 
a specific site. Since each site may be consid- 
ered as some point in a continuum gradient, 
condition guides based upon such a gradient 
will more accurately evaluate specific sites. 
Management practices can and are classified 
as a means for discussing their merits, limita- 
tions, and applicability in vegetation manipula- 
tion. Plant communities may be classified in 
any manner that will meet our objectives, 
whether they be for land management, demon- 
stration of continuum gradients, maps of simi- 
lar site potentials, or descriptions of modal en- 
vironments. In land management, plant com- 
munities should probably be grouped according 
to similarities and limitations in management 
practices, productivity, and numerical simi- 
lairty of important characteristics. A single 
classification may not meet all these criteria. 
For example, the numbers of three-step “hits” 
and percent composition of decreasers and in- 
creasers may be similar for two classification 
groups, such as Agropyron on steep slopes and 
Artemisia tridentata/Agropyron on moderate 
slopes. Both can be evaluated with the same 
condition guide and could be grouped together; 
however, managment is different due to the 
presence or absence of Artemisia. 
Mapping also requires grouping of plant 
communities for efficiency and simplicity. Since 
mapping objectives in land management are di- 
rectly related to resource management, plant 
communities should be classified according to 
significant differences in management charac- 
teristics. For example, Agropyron/Festuca on 
moderate slopes may be divided into shallow 
and deep soil groups to reflect differences in 
forage production, revegetation potential, and 
sensitivity to grazing. In range mapping, each 
map unit must be rated for condition. In this 
way, the land manager obtains a _ picture 
(map) and an inventory of areas differing sig- 
nificantly in management requirements and 
212 
limitations. However, he must understand that 
map units and inventory lists are largely gen- 
eralized approximations; they may not repre- 
sent real biological entities or actual site po- 
tentials and characteristics. 
Sampling will often be influenced by the cli- 
max continuum philosophy. All studies evalu- 
ating plant community distribution or species 
distribution should be designed to sample con- 
tinuum gradients. If gradients occur, this sam- 
pling will tend to more realistically evaluate 
the situation as it occurs in the field. If gra- 
dients do not occur, the sampling will clearly 
indicate grouping or clustering of similar plant 
communities. Sampling limited to a few thou- 
sand acres will often indicate grouping or clus- 
tering. Samples of vegetation over hundreds of 
miles in all directions are required for good 
gradient expression. 
Interpretation of research findings is influ- 
enced by this philosophy. Study results often 
are applicable only to certain portions of a con- 
tinuum gradient. These results generally will 
be less and less applicable as distance from the 
study site increases and as different plant com- 
munities are encountered. Suggestions of gross 
similarities in vegetation, such as a “juniper 
zone,’ tend to be misleading because they 
imply some kind of uniformity in vegetation 
and environment. Such broad distributions 
often reflect the ecological amplitude of a spe- 
cies. Most species, such as juniper, occur over 
wide areas encompassing part of a State or 
several States. Such breadth of amplitude en- 
compasses great variability in environment 
and is not a suitable basis for plant community 
classification that is useful in vegetation man- 
agement. 
AN EXAMPLE OF CONTINUUM IN 
CLIMAX 
Variability of community groups.—The 5 
million-acre Blue Mountain landmass in cen- 
tral and northeastern Oregon was evaluated by 
intensive ecological reconnaissance. Using the 
climax continuum philosophy, 450 sample lo- 
cations were selected. Only locations free or 
nearly free of disturbance by grazing or log- 
ging were selected. Herbaceous vegetation was 
sampled using the modified three-step method 
(Parker and Harris 1959). Study objectives 
were to develop range condition guides and to 
develop or to improve vegetation management 
guides. One method used to accomplish these 
objectives was classification of plant communi- 
ties based upon characteristics important in 
management. Table 1 shows the mean and 
standard error for hits and for composition of 
three species and two groups of species, de- 
creasers and increasers, for five nonforest plant 
ee 
