78 
aquatilis, and a dense ground cover of the moss, 
Scorpidium scorpioides; and (2) a dry type, 
characterized by much standing dead plant 
material, which gives the type a grey-brown 
coloration, and having a somewhat rubbly, hum- 
mocky surface of mixed sedges, small shrubs and 
various herbs. In between these two extremes is 
an array of species occurrence as shown in both 
tables that makes classification into separate 
community types extremely difficult. Stands 
that fit the more or less wet category are 
somewhat easier to lump together than are 
stands that fit the more or less dry category, 
which shows high variation, and, at present, it 
seems that the vegetational and floristic shifts 
are far too subtle in this area of low variation in 
relief, plant size and life form to develop a 
rational and useful classification scheme. Suffi- 
cient departures from smooth normal curves can 
be seen in the frequency arrays of Table 2 to 
suggest that at least one more major environ- 
mental gradient may be operative in these pat- 
terns. The one-dimensional ordination does 
reveal some informative patterns of species and 
community distributions. 
In Table 1 the general species’order is that of 
a gradual shift from those of wet stands to those 
of dry stands, moving from the top of the 
column toward the bottom. Characteristic 
species toward the wet (top) are an unidentified 
alga, Scorpidium  scorpioides, Pedicularis 
sudetica, Eriophorum angustifolium, and Carex 
aquatilis. To this group, as the alga and Scorpidi- 
um decreased, were added Polygonum vivi- 
parum, Salix reticulata, Salix arctica, Equisetum 
variegatum, and Carex membranacea. Largely 
confined to the center of the array were 
Cardamine pratensis, Carex marina, Melandrium 
apetalum, Dupontia fisheri, Carex atrofusca, 
Pedicularis lanata/arctica (not distinguished con- 
sistently in the field), Eriophorum vaginatum, 
Carex bigelowii, and Eutrema edwardsi/. Toward 
the dry end of the species list, stands were 
characterized by Oryas integrifolia, Carex 
misandra, C. bigelowii, Arctagrostis /atifolia, 
Draba spp., Papaver macounii, Chrysanthemum 
integrifolium, Saxifraga hirculus and S. op- 
positifolia. Additional species of the sand-silt 
surfaced high-center polygons and the dry pat- 
terned plain are those lowest in the list. When 
species presence is plotted against the stand 
ordination (Table 1), although gaps are expected 
and do occur in a one-dimensional array, a 
fairly gradual change is seen progressing from 
left to right. The progression of vegetational 
groups is from one group of species that domi- 
nate and characterize the wet centers of low- 
center polygons, the flat areas of ridged wet 
plains, the smooth wet plains, and the areas of 
seasonally high water immediately around lakes; 
through various intermediate mixtures of species 
that characterize the smooth plains, the ridges of 
low-center polygons, and the dry smooth plain; 
to the groups that dominate and characterize the 
centers of high-center polygons, the patterned 
plains, and ridges in wet ridged plains. Stands 1 
and 28 had no species in common; no one 
species was found in all stands, and only 
Eriophorum angustifolium, Polygonum vivipar- 
um, and Salix reticulata had presence values 
above 75%. 
When the frequency values for the species of 
the 14 sampled stands are arrayed on the same 
basic ordination (Table 2), although smooth 
normal curves are rare, most species show fairly 
consistent patterns, with stands having high 
values being close together and the stands of low 
value arrayed in one or both directions away 
from these. Only Carex aquatilis and Scorpidi- 
um scorpioides show similar patterns of fre- 
quency distribution. Twenty-two of the total 59 
species are found with percent frequency at or 
above 50% in one stand or another, yielding a 
fairly high number of important species, as 
compared with species that are rare within or 
between stands. Diversity, both in numbers of 
species and in number of individuals (to the 
extent that frequency here reflects density), is 
highest in the stands toward the dry end of the 
array. There are sufficient departures from 
normal curves to suggest strongly the need for 
additional frequency sampling and development 
of a two-dimensional array. 
Acknowledgment 
We thank David F. Murray and Barbara M. 
Murray for assistance with the site selections and 
plant identifications. The Prudhoe Bay Environ- 
mental Subcommittee and the State of Alaska 
assisted in funding these studies through the 
Tundra Biome Center, University of Alaska. 
