Document Sample
					Biol. Rev. (1968), p p . 481-529

                  BY W. D. BILLINGS AND H. A. MOONEY
         Duke University, Durham, North Carolina, and Stanford University,
                           Stanford, California, U.S.A.

                                       (Received I 5 May I 968)
        I. Introduction      . . . . . . .                          .     .     .    .     .    481
       11. Geographic extent and general characteristicsof arctic and alpine vegetation         484
             (I) Present and past distributions of tundras          .     .     .    .     .    484
                                     . . .
             (2) Environmental characteristics                      .     .     .    .     .    485
                                     . . .
             (3) Vegetational characteristics                       .     .     .    .     .    489
      111. Adaptations of plants to arctic and alpine environments        .     .      ,   .
                                       . .
             ( I ) Life forms and general morphology                .     .     .    .      .
             (2) Physiological ecology of the life-cycle in arctic and alpine vascular plants
                                       . .
                     Seed dormancy and germination                  .     .     .    .     .
                                   . . . .
                     Seedling establishment                         .     .     .    .     .
                                       . .
                     Chlorophylls and other pigments                .     .     .    .     .
                                       . .
                     Photosynthesis and respiration                 .     .     .    .     .
                     Effects of water availability and use          .     .     .    .     .
                                   . . . .
                     Annual cycle of growth                         .     .     .    .     .
      IV. Primary productivity . . . . . .                          .     .     .    .     .
       V. Summary      .     .     .     .    .     .     .    .     .     .     .    .     .   522
      VI. References.        . . . . . . . . . . . .                                            524

                                         I. INTRODUCTION
  Among the earth's terrestrial environments, none has less biologically usable heat
or has fewer kinds of adapted plants than the tundras and barrens above and beyond
the alpine and arctic timberlines. Here, plant phenotypes are environmentally selected
by a climatic severity unknown in the mild, moist tropical environments where vascular
plants originated and remain in such variety today. Nor does so severe a selection
operate in temperate regions where warm summers help to compensate for cold winters
by supplying a season of tropical heat to the metabolic activities of plants emerging
from dormancy. Only a few kinds of phenotypes have passed this low-temperature
screening successfully and have added their gene-enzyme systems to the floras and
vegetations of polar and alpine regions. It is the purpose of this review to bring together
in brief form the principal facts and theories concerning plant adaptations to these
cold environments.
   I n short space, we cannot improve on the detailed coverage of certain aspects of
plant adaptation to arctic or alpine conditions provided by Holm (1922), Schroeter
(1926), Sarensen (1941), Pisek (1960), Tikhomirov (1963), Bliss (1962b), Tranquillini
(1964), and others. Here, we shall attempt to bring such information up to date and
to discuss the problems of physiological adaptation to low-temperature environments.
482                      W. D. BILLINGS H. A. MOONEY
   In attacking this problem, it is appropriate to ask ‘What can we learn from studying
arctic and alpine plants that cannot be learned from other kinds of plants?’ Before
answering this question, we must define our terms.
   We shall define an arctic plant as one growing beyond the arctic timberline or
belonging to a species whose main distribution is beyond this timberline. An alpine
plant has the same relationships with the alpine timberline. Many arctic or alpine
species occur also in meadows or open areas below timberline. Because of this fact and
because timberlines are often broken, indistinct, or even lacking, timberline is only
a rough boundary between the severity of a tundra environment and the relative
protection of a forest or a subalpine meadow.
   Timberline is a relatively reliable guide in the Northern Hemisphere but it is much
less so in the southern part of the world. For example, in the alpine regions of New
Zealand the Nothofagus timberline is at too low an elevation to indicate the real
boundary between alpine and subalpine conditions. The situation is carried to an
extreme in such places as the western slopes of the Andes in central Chile where there
is no montane forest, thus no timberline, and the lower edge of the alpine zone is left
solely to the ecologist’s judgement. In the last analysis, what constitutes arctic or
alpine conditions always is a matter of judgement. However, the effect of a forest on
microclimate rules out the radiation conditions of an open tundra, and thus the
timberline boundary has some reality. (See Brockmann-Jerosch (1919) for a more
complete discussion of timberline and its climatic relationships.)
   In both arctic and alpine situations there are environmental gradients, genetic
clines, and vegetational continua which cut across any sort of arbitrary boundary.
It is thus impossible to delineate arctic conditions from subarctic, alpine from sub-
alpine, in an absolute manner. Arctic and alpine conditions are primarily a matter of
degree. However, such gradients and clines are usually steepest at timberline; thus,
timberline provides us with an approximate and reasonably acceptable lower limit to
arctic and alpine conditions. Unfortunately, this may not be true for much of the
Southern Hemisphere for a variety of environmental and genetic reasons which result
in a low timberline or none at all. While we shall be concerned primarily with true
arctic and alpine plants in this review, we shall not exclude information derived from
subarctic, subalpine, or subantarctic plants if this can be helpful in understanding the
adaptive ecology of plants of cold regions.
   While some plant species are endemic to the Arctic, and a great many more are
endemic to the alpine regions of one mountain range or another, a relatively large
group of common tundra species are widespread and occur in both arctic and alpine
locations. Such taxa are commonly called ‘arctic-alpine’ species. They provide impor-
tant links between the arctic flora and the numerous alpine floras; an understanding
of their adaptive mechanisms can provide answers to questions concerning evolution
within and migrations between these similar, but different, low-temperature environ-
ments. Among the more widespread of these arctic-alpine species are Trisetum
spicatum,* Oxyria digyna and Silene acaulis, which are almost cosmopolitan in arctic
and alpine regions of the Northern Hemisphere. Trisetum spicatum and a few other
       *   Nomenclature follows Polunin (1959) in so far as the species are included by him.
                      The ecology of arctic and alpine plants                         483
species of arctic-alpine vascular plants also occur in the mountains of the cold-
temperate parts of the Southern Hemisphere. They are, thus, ' bipolar arctic-alpine '.
A number of species of lichens and mosses are also bipolar arctic-alpine.
   Terrestrial plants of arctic and alpine regions are mainly flowering plants (Angio-
sperms), bryophytes, and lichens; ferns are also represented but with fewer species.
Almost all the Angiosperms are herbaceous perennials or very low shrubs ; annuals are
very rare. The perennial herbs are of four principal life-forms: cushion or polster
plants, rosette plants, leafy-stemmed plants, and grass-form plants.
   All tundra plants, woody or herbaceous, show rapid shoot growth after melting of the
snow-cover in spring or early summer. Such extremely rapid growth, in a matter of
a week or two, is one of the unique characteristics of such vegetation. The energy and
materials for this fast shoot growth are supplied by carbohydrates and lipids stored in
roots, rhizomes, or bulbs. Tundra plants also have the ability to metabolize and to
reproduce at low growing-season temperatures which are not far above the freezing
mark. Under such conditions, sexual reproduction is often replaced by apomixis,
vivipary, or various kinds of vegetative reproduction. All polar and alpine Angio-
sperms, except for the few small and delicate annuals, have these characteristics in
common. Additionally, many species, especially those of exposed, windy habitats, also
can withstand extremely low temperatures and desiccation during the dormant season.
In an already severe regional environment, windy ridge crests which are snow-free
in winter provide one extreme in local severity, while the late-melting snowbank is the
cause of an environmental severity of a quite different sort-the very short growing
season. Angiosperms adapted to either extreme are relatively few; some lichens are
well-adapted to the wind-swept environments and a few moss species are abundant
in some but not all late snowbank sites. Really extreme situations of both types are
without any plants.
   The uniqueness of polar and alpine plants lies in the fact that they are the only plants
adapted to metabolizing, growing, and reproducing at low temperatures. Many other
kinds of plants from the forests, grasslands, and cold deserts of the middle latitudes
can tolerate and survive extremely low temperatures during the dormant season but
require higher temperatures for growth and development than do the plants of arctic
and alpine environments. In subarctic or subalpine environments, such higher tempera-
ture requirements are met by higher daytime temperatures since subalpine nights are
often colder than alpine nights because of cold-air drainage. Similarly, subarctic
nights often are colder than those of the Arctic because the sun goes below the horizon
for several hours in contrast to the continuous daylight of higher latitudes. Higher
daytime temperatures are closely allied with photosynthetic processes and thus it is
daytime temperature that marks the real boundary between true arctic or alpine
tundra and subarctic or subalpine meadows.
   T o answer our earlier question: by studying arctic and alpine plants we can hope
to learn how this relative handful of species in the world's flora has succeeded not
only in surviving low temperatures during dormancy but in manufacturing relativeIy
large amounts of food at low temperatures in very short periods of time. We can get
additional dividends by applying the question to alpine plants in particular. Alpine
484                                 AND H. A. MOONEY
                       W. D. BILLINGS
plants are subject not only to the rigours of low temperatures day and night but also
to the other environmental stresses inherent in high altitudes: low partial pressures of
oxygen and carbon dioxide, strong winds, and intense solar radiation including ultra-
violet. The answers we get to these questions of low-temperature adaptation of
populations and ecosystems will be worth the quest not only scientifically but also
from the economic standpoint of food production in cold climates.

                       (I)  Present and past distribution o tundras
   Because of the moderating influence of the Arctic Ocean, tundra vegetation occurs
as far north as Peary Land on the northern tip of Greenland, reaching its limit with
Saxifrage oppositifoliaat KapMorris Jesup in latitude 83" 39' N. (Holmen, 1957). South
of this extreme, tundra extends around the seaward edges of the Greenland ice-cap
and in a circumpolar band across northern Eurasia. Wherever mountain ranges enter
the Arctic from the south, as in Alaska, Scandinavia, and the U.S.S.R., arctic tundras
merge almost imperceptibly with alpine tundras and mountain meadows. Farther
south, Northern Hemisphere types of alpine tundra and fell-fields exist in many
mountain ranges of the middle latitudes and reach almost to the tropics in Nepal and
south-western China.
   Along the equator, in the high mountains of Africa and South America, are alpine
regions very different from those farther north in environments, floras, and vegetation.
The small alpine areas of the Southern Hemisphere in the southern Andes, New
Zealand, and Australia superficially resemble those of the North in environments and
in the life forms of the plants but, except for a few bipolar species, have a flora of quite
different derivation. Subantarctic tundra occurs on some of the oceanic islands south
of 55' S. and, sparsely, even on the shores of Palmer Peninsula in Antarctica to 64' S.
   Good (1964) provides some figures on the approximate area of land covered by
arctic or alpine vegetation. By far the largest amount is in the Northern Hemisphere:
almost 9.1 million square miles compared to less than 0.5 million square miles in the
Southern Hemisphere. In the Northern Hemisphere about 60 yo of this vegetation is
north of 60" N. and can be considered as arctic or subarctic. Northern Hemisphere
alpine areas are relatively small, isolated, and floristically diverse as compared with the
circumboreal belt of arctic tundra. Conversely, in the Southern Hemisphere almost
all of the tundra vegetation is alpine, with less than 0.1yobeing south of 60" S. In the
Arctic, the land is peripheral to the Arctic Ocean, which moderates the climate and,
in summer allows thawing and temperatures high enough for plant growth. In the
Antarctic the high continent is central and the ocean peripheral; the result is ice-
covered land with little or no chance for plant establishment.
   The present distribution of tundra ecosystems with the greater area occurring at
high latitudes in a circumpolar zone has been characteristic only of the warm inter-
glacials of the Pleistocene and perhaps also of the later Tertiary. During much of the
Pleistocene, tundra climates and biota have migrated back and forth with the cycles of
                       The ecology of arctic and alpine plants                          485
continental and alpine glaciations. While little is known of the distribution of tundra
during full-glacial times, continental ice-sheets and alpine glaciers covered much of
the tundra’s present range. In North America and Europe, except for the unglaciated
parts of Alaska and nunataks in the Scandinavian mountains and in the Canadian
Cordillera, arctic plants could survive only south of the ice-front. Alpine plants grew
on the lower mountain slopes, in the lowlands, and probably in mountain ranges such
as the southern Appalachians, which are completely forest-covered today. I n many
places there must have been considerable mingling of the arctic and alpine floras. The
resultant tundra in middle-latitude North America may have been only a relatively
narrow zone between ice and forest. However, because much of Eurasia was both
unglaciated and very cold, full-glacial tundra there could have covered a very large
area, particularly in Russia and Siberia. An unusual situation existed in north-western
Alaska, which was unglaciated and separated at full-glacial from the rest of the North
American continent by 2000 miles of ice. Here, a large sample of North American
tundra vegetation remained in the Arctic during the height of glaciation and for many
thousands of years had better migrational connexions with the Eurasian tundra than
with the remnants of the American tundra, which had retreated far to the south. The
effects of this isolation are still evident in tundra floras and species structure in western
and north-western North America.

                            (2)  Environmental characteristics
   Arctic and alpine tundra ecosystems may be described briefly as treeless areas beyond
timberline characterized by cool or cold summers and occupied by low herbaceous or
shrubby vegetation, often with extensive mats of lichens or mosses, Beyond these
generalities, there are many variations due principally to the uneven distribution of
snow and meltwater, permafrost, soil parent materials, and topography. Table I is an
attempt at a rough quantification of the general environmental characteristics of an
arctic tundra station (Barrow, Alaska) and an alpine tundra location (Niwot Ridge,
Colorado) compared with those of a cold temperate forest of pine. It should be
emphasized that long-term environmental data from tundra locations are rare because
of the physical difficulties of operating instruments in such places. While much of the
information in Table I is short-term and therefore approximate, there is value in the
data and they speak for themselves in telling of the severity of tundra environments.
   In regard to summer solar radiation, the daily totals at Barrow, Alaska, are only
slightly lower than those at the two montane stations. However, since the arctic
photoperiod is continuous, the intensity or rate at Barrow is only about half that in
Colorado and Wyoming. The greater altitude of these latter stations also contributes
to their higher rates of radiation and also contributes to the higher amounts of U.V.
(Caldwell, 1968). Air and soil temperatures are generally lower in the Arctic but
wind-chill on the alpine ridge keeps the summer maximum temperatures low; this is
a common phenomenon in the mountains.
   Precipitation is much lower at Barrow than in the two middle-latitude stations even
though these stations themselves are dry as compared to many mountain regions.
Fcr example, Sonnblick at 3106 m. in the Hohe Tauern of Austria has an annual mean
486                           W. D. BILLINGS H. A. MOONEY
             Table     I.    Comparative environmental characteristics of an arctic,
                              an alpine and a temperate forest ecosystem
      component                   Arctic tundra*             Alpine tundra?         Temperate forestf
Solar Radiation   ( I 965)    Source A                   Source B                 Source A, Laramie,
  Highest daily total         760 langleys               780 langleys             704 langleys
   (June, July)
  Av. July daily total        426 langleys               497 langleys             5 14langleys
  Av. July intensity          0.30 cal./cm.-* min.-'     0.56 cal./cm.-* min.-'   0 5 8 cal./cm.-B rnin.-'
  Quality                     Low in short wave-         High in all wavelengths High in all wavelengths
                               lengths, particularly      particularly short u v.
                               short U.V. (2950-
                               3 1.50 A.1
  Max. photoperiod            84 days                    15 hr.                   15 hr.
Temperature                   Source A                   Source C                 Source C
 (+ I m., "C.)
     Annual mean              - 12-4                      -3'3                       8.3
     January mean             - 26.7                     - 12.8                   -  1.7
     July mean                   3'9                        8.3                     20.6
     Absolute max.              25.6                       18.3                     37'2
     Absolute min.            -48.9                        36.6                   -33.8
Soil ( - 15 cm., "C.)         Source D                   Source C                 Source C
  Annual mean                 - 6.2                      - 1'7                       8.3
  January mean                - 14.5                       -                        -
  July mean                      2'5                                                  -
  Absolute max.                  2'5                       13.3                       31-1
  Absolute min.               - 15'5                     - 20'0                    - 10'0
Precipitation (mm.)           Source A                   Source C                 Source C
  Annual mean                  107                        634                      533
  Highest monthly               71                        203                      203
   (wettest month)
  Lowest monthly                  0                         6                         0
   (driest month)
Wind (km/hr.)                 Source E                   Source C                 Source C
  Annual mean                   19'3                       29.6                     10.3 .
Max. water stress in          Low (- 4 to    - 5 bars)   Rel. low (- 6 to   -8    Rel. high but no data
 plants (Source F)                                        bars)                    (probably -25 bars or
 Metabolic gases (Source G)
   COa fmg./lJ              057                             036                        044
   Oa (partial pressure   160                             I00                       I22
    in mm.)
 Soil frost activity     Source D: much,              Source C: some, active, Source C: none in
                          active                       on small scale; large,   growing season
                                                       fossil stone nets
 Depth of soil thaw           Source F: 20-100 cm.    Source H: 30 cm to       Source H: no perma-
                               depending on site;      bedrock; permafrost      frost
                               permafrost universally rare
Vegetation                    Source D : Carex-       Source C: mosaic of      Source C: open forest
                               Dupontia-Eriophorum     alpine vegetation types of Pinus ponderosa
                               tundra                  with Kobresia meadow
                                                       tundra as climatic
  SOURCES:    A, Climatological Data, National Summary, US.Weather Bureau, 1965; Clark & Marr
(1966); C, Marr (1966); D, Brown & Johnson (1965); E, Watson (1959); F, J. Dennis & T. F. Rochow
(1967), personal commun.; G, W. D. Billings; H., Marr (1961).
  * Barrow, Alaska (altitude 7 m., latitude 71' 20' N.).
  t Niwot Ridge, Colorado (altitude, 3749 m.; latitude 40"N.).
  1 Bummer's Gulch, Colo. (altitude 2195 m., latitude 40' N.).
                      The ecology of arctic and alpine plants                       487
of 1643 mm., M t Washington at 1917m. in the White Mts of New Hampshire receives
 1784 mm., and though quantitative data are lacking, there are many alpine locations
that receive much more precipitation. Even though Barrow could be considered a
'middle Arctic' station in the sense of Polunin (1951), its precipitation is as typically
low as that of a 'high Arctic' station such as Isachsen at 78" 47' N. on Ellef Ringnes
 Island which has an annual mean of 99 mm. or Eureka at 80" N. on Ellesmere with
only 67 mm. per year. As Savile (1964) says of Hazen Camp (81"49' N., Ellesmere),
'aridity is far more important than temperature as a limiting factor in plant growth
[at Hazen] '. If Barrow were hilly rather than flat and poorly drained, aridity would be
an important factor in plant growth there. As it is, water from annual soil thaw above
permafrost provides abundant soil moisture (Brown & Johnson, 1965).
    Low precipitation in the high Arctic often results in such weak soil leaching that
salts accumulate on or below the soil surface in amounts such as to restrict all plant
growth except that of an occasional Puccinellia. Savile calls such areas 'true desert'.
Apparently, Stocker (1963) would agree, but goes even farther in his description of
most of the high Arctic as 'cold-desert'. Such deserts are rare in alpine situations but
occur on the eastern slopes of the high Andes in Argentina, for example.
   Alpine sites are windier than most arctic locations and, of course, have lower con-
centrations of metabolic gases due to lower atmospheric pressures. Soil frost activity
and permafrost are much more important environmental factors in the Arctic than
they are in most alpine sites.
   In these arctic and alpine environments the plants are small, close to the ground,
and often widely separated, with bare soil or rock in between. Compared with the
situation in a forest, the modification of the climate by vegetation is minimal and the
physical environment is dominant. Radiation and wind have easy access to the soil
and rock surfaces and to most of the above-ground plant tissues as well. I n such open
windy places, the effects of real microenvironment are pronounced. Even a few centi-
metres difference in microtopography makes a considerable difference in soil tempera-
ture, depth of thaw, wind effects, snow drifting, and resultant protection to twigs,
buds, and leaves. Such microtopography may be a small rock, a peat hummock, or
polygon rim caused by soil frost action-or it may be a small depression in which
snow accumulates. Whatever the microtopographic pattern, the results are an uneven
distribution of snow and consequently unevenness in protection from wind-blast, in
the distribution of soil moisture, reception of solar radiation, and in soil temperature.
As we have shown (1959), some of these microtopographic features change through
time and peat hummocks may eventually rise so high as to be wind-blasted in dry
winters so that soil frost-thrusting may push through the scars at the tops of the hum-
mocks and the whole bog may be base-levelled into soil polygons.
   I n an open vegetation of isolated cushion plants, solar radiation reaches the soil in
relatively great amounts even in the Arctic. On the other hand, in sedge or grass turf
communities, the sod insulates against heat flow with resultant slowing down of soil
thawing and freezing. T h e soil and lower air-temperature situations are thus quite
different between an open fell-field and a moist meadow.
   Local temperature gradients are controlled by the distribution of soil moisture,
488                          W. D. BILLINGS H. A. MOONEY
which is dependent to a large extent on snowdrift pattern which is the result of inter-
action between wind and topography (see diagram in Fig. I ) , These gradients are
sharper in alpine places but the effects are also visible in the Arctic. The drier the
summer, the steeper are the local moisture gradients because they relate directly to
position of late snowbanks; this is particularly characteristic of the alpine zone of the
Sierra Nevada of California. With the possible exception of the effects of certain types

         SNOW COVER                                                                IE
                                                                      TOPOGRAPHIC S T

                                 snow free

                                                                  upper slope


                         lee exposure                            windward exposure

                                             WIND EXPOSURE
        Fig.   I.   Alpine vegetational pattern along topographic and moisture gradients in the
                    Beartooth Mts., Wyoming. Redrawn from Johnson & Billings (1962).

of underlying rocks such as dolomite, limestone, or serpentine, the distribution of soil
moisture controls most local environmental and vegetational gradients both in alpine
and arctic regions. All of the plants are adapted to low temperatures but some are
better adapted to drought and extreme temperatures; these are on the ridges and open
fell-fields, where they have less control over their environment but where the sun’s
rays hitting dry soils provide warmer microenvironments in a shallow layer close to
the ground.
   While isolated tundra plants may not have much influence on the local climate, the
microclimate within the crown of the plant itself may be considerably different from
that in the open air between plants. This ‘phytomicroclimate’ is influenced by leaf
size, shape, colour, pubescence, and arrangement and the temperature and heat balance
of the plant itself. The hollow stems of wet meadow plants even enclose a phyto-
microclimate where temperatures far exceed those in the open (Billings & Godfrey,
 1967). The heating effects in plant tissues are most pronounced on bright sunny, cool
days, and are almost negligible in cloudy or misty weather; the phenomenon is more
                       The ecology of arctic and alpine plants                         489
prevalent under alpine conditions. Some of the heat energy is convected into the air
within the plant canopy and this air, protected from the wind a few centimetres above,
becomes heated. The plant is immersed in a warmer microenvironment than might be
imagined by an observer subjected to the chilling breeze.
   Arctic environmental characteristics are described by Bliss (1956),Serensen ( 1 9 4 1 ) ,
and in considerable detail and scope in the great compendium on Cape Thompson
edited by Wilimovsky & Wolfe (1966).Alpine environmental data are available in
Bliss (1956,1966),                  ,
                   Marr ( 1 9 6 1 ) Hedberg (1964),Turner (1958),Gates & Janke (1966),
and Mark ( 1 9 6 5 ~ ) .
   Since so much of the alpine or arctic landscape is characterized by scattered plants,
such plants have only a very local effect on the modification of the microenvironment.
The plants themselves are at the mercy of the structure and fluctuations of the physical
environment. Also plant-to-plant interactions which result in competition or ‘inter-
ference’ are less marked than they are in the forests and meadows of milder climates.
In this respect, the vegetations of arctic and alpine areas resemble those of deserts.
As with desert plants, tundra and fell-field plants are ill-equipped to utilize the milder
environments at the limits of their ranges where better-adapted competitors deny
them the needed light and water. Yet even in the most open fell-fields, cushion plants
such as Silene acaulis are invaded and eventually out-competed by other tundra
species (Griggs, 1956). Except for the work of Griggs and a few others, however,
little is known about interactions between plants in arctic or alpine regions.

                              ( 3 ) Vegetational characteristics
   Beyond saying that arctic and alpine vegetation consists of low perennial herbs and
dwarf shrubs, it is impossible to generalize about its composition and structure. In the
Arctic and in the Northern Hemisphere mountains, tundra vegetation is a complex
mosaic of communities arrangeable into types or along vegetational gradients. The
controlling environmental factors are topographic exposure and distribution of snow
and meltwater superimposed upon geological substratum patterns. One of the
commonest gradients is that from cushion plants such as Silene acaulis and Saxijraga
oppositijolia on exposed ridges and rocks to dry meadows and heaths at midslope,
thence across snowbanks to the dense peaty turf of bogs and wet meadows dominated
by grasses such as Deschampsia caespitosa and various species of Poa and by the sedges
Carex and Eriophorum. But there are innumerable variations on this theme.
   Once one leaves the Arctic and comes south into the middle-latitude mountain
ranges of the Northern Hemisphere, the floristic variety increases somewhat and the
vegetational mosaics and gradients become more complex. Still, however, most alpine
vegetation resembles that of the Arctic both floristically and in structure. However,
on entering the tropical mountains, the vegetation becomes much more complex
floristically and often is marked by the presence of columnar or arborescent life-forms.
The tropical alpine environment is under diurnal control rather than annual; there are
great differences between day and night heat budgets and temperatures. There is
relatively little resemblance to the Arctic here in either environment or vegetation.
   Southern Hemisphere alpine vegetation does show the typical arctic-alpine gradients
    32                                                                     Biol. Rev. 43
490                   W. D. BILLINGS H. A. MOONEY
from ridge to bog but floristically has little resemblance to the vegetation of the North
except for a few grasses, sedges, lichens, and mosses held in common. Very dwarf
shrubs are abundant.
   It is not possible to mention all the valuable papers which have been written on
tundra vegetation. Excellent descriptions of typical vegetational patterns, composition,
and structure, however, can be found in the following publications.
   Arctic. Britton (1957), Bocher (1963), Hanson (1953), Johnson, Viereck, Johnson &
Melchior (1966), Larsen (1965), Polunin ( 1 9 4 9 Raup (1965).
   Eurasian Alpine. Dahl(1956), Gjaerevoll(1956), Jenny-Lips (1948), Braun-Blanquet
& Jenny (1926), McVean & Ratcliffe (1962), Braun-Blanquet (1948), Sukachev (1965),
and many papers in the series Beitrage zur geobotanischen Landesaufnahme der
Schweiz published by the Schweizerischen Naturforschenden Gesellschaft in Bern.
   North American Alpine. Bliss (1963), Johnson & Billings (1962), Marr (1961).
   Tropical Alpine. Hedberg (1964), Troll (1957), Weberbauer (191I).
   Southern Hemisphere Alpine. Billings & Mark (1961), Costin (1967), Ward &
Dimitri (1966).
   Subantarctic. Taylor (1955).

                          (I) Life forms and general morphology
   The most obvious adaptations to severe tundra or alpine environments are reduc-
tion in plant height and a tendency toward an herbaceous habit. I n exposed places,
even small trees are blasted by wind and blowing snows so that ‘tree’ growth is pos-
sible only as a twisted shrubby ‘Krummholz’ near the ground. Snow collects and packs
hard in the dense branches of the Krummholz and thus protects the evergreen leaves
against death by desiccation in the dry winter air when the soil is cold or frozen. Snow
also insulates the leaves against temperature extremes: low in winter and high in the
bright sun of early spring before the soil has thawed.
   Exposure to the rigours of winter is not the only factor limiting upward and
northern distribution of trees. As Hustich (1948), Daubenmire (1954), and others
have pointed out, summer temperature strongly affects cambial growth of that season,
and also seed production. Hustich correctly stresses the sensitivity of trees at timber-
line to the smallest changes in climate-particularly in the temperature of the warmest
month. Boysen-Jensen (1932) suggests that timberline occurs where the amount of dry
matter produced each summer is consumed entirely in respiration and in the pro-
duction of new leaves; there is nothing left over for wood. One might say that the use
of carbohydrate for the permanent cellulose of aerial wood is an unattainable luxury
when compared to the efficient re-use of sugars, starches, and lipids by herbaceous
plants and prostrate shrubs in the same cold environment. These digestible substances
can be stored underground and used the following season in leaf production and stem
elongation. On the other hand, radial growth of trees appears to be dependent mainly
upon carbohydrates produced during the growing season in which that growth occurs
(Hustich, 1948). If that growing season is cold, enzyme systems leading to sugar and
                      The ecology of arctic and alpine plants                        491
starch seem to have priority over those to cellulose. This could explain the sensitivity
of radial growth of trees at timberline to even small year-to-year differences in tempera-
ture during the summer. If favourable summer temperature years are few, woody stem
volume will increase very slowly; if favourable temperature years never occur, the
production of even a dwarfed tree trunk is impossible.
   In the same sense, it would seem as if the metabolic expense of producing a whole
new crop of leaves each year is exorbitant for a tree under timberline conditions (Bliss,
 1966). Moreover, as Strain (1966) found with Populus tremuloides below timberline
in the Sierra Nevada of California, there is always the danger of young leaves of
deciduous species being severely damaged by late frosts. This may be a common
weakness of deciduous species in subalpine or subarctic environments since we have
noticed that evergreen species in such places break bud dormancy somewhat later
than do the deciduous species. Evergreen trees can afford the apparent waste of these
days of uncertain weather early in the growing season since their older leaves are
already in photosynthetic operation; deciduous trees must take the chance of pro-
ducing leaves as early as possible.
   Even though evergreen leaves save the expense of producing all new photosynthetic
tissue each year, there are some disadvantages. Principal among these is the prevalence
of winter-killing of exposed leaves due to desiccation by dry winds at times when soil
moisture is cold or frozen, Snow blast also takes its toll. Under these conditions, the
Krummholz mat at the base of the tree has advantages over the battered and one-
sided ‘flag-form’ crown where leaf life is relatively short. Under the most extreme
situations a crown extending above the snow is impossible and only the mat-like
Krummholz represents the species.
   Timberline in equatorial mountains is rather difficult to define (Troll, 1957).
Columnar or sparsely branched small ‘trees’ of genera not usually tree-like are
scattered through shrubby or tussocky alpine slopes above the true forest. The leaves
and inflorescences of these columnar plants are usually covered with thick coats of
woolly hairs; Troll (1958) callsthis life form ‘Wollkerzengewachse’, woolcandle form.
Hedberg (1964) uses the term ‘giant rosette plants’. There has been parallel evolution
of these forms on the high East African volcanoes and in the northern Andes. In
Africa the principal genera represented by giant rosette plants are Senecio and Lobelia,
while in South America the comparable genera are Espeletia, Lupinus, and Puya. Very
similar in form is the silver-sword, Argyroxiphium sandwicense, on the high Hawaiian
volcanoes. These equatorial alpine environments have no winter and summer seasons
but are characterized by strong incoming radiation by day and strong outgoing
radiation at night all through the year. The winters of the middle-latitude mountains
are unknown here but frost occurs almost every night.
   The ecology of these equatorial alpine locations in Africa (‘ afroalpine’) has been
studied by Hedberg (1964). The principal advantages of the columnar giant rosette
plants (Senecio and Lobelia) apparently are concerned with the nightly protection of
the buds, which are never dormant. The large woolly leaves curve inward at night,
covering the growing tip and protecting it against frost. Additionally, Lobelia keniensis
contains water in the rosette ; this water according to Hedberg apparently is secreted by
492                    W. D. BILLINGS H. A. MOONEY
the plant and covers the shoot apex and leaf primordia to the depth of ca. 10 cm.
Every night the uppermost water freezes but the lower water never freezes and its heat
protects the growing tip of the shoot. I n addition to these large ‘night buds’, the
sheathing of the stem by the old hairy leaves and the hairy bracts on the inflorescence
not only may act as insulation against loss of heat during the night but they also reflect
the intense sunlight during the day. While this work of Hedberg’s pertains to only
the ‘afroalpine’ rosette plants, it is probably true to a large extent for the South
American and Hawaiian giant rosette plants also.
   Above and beyond timberline outside of the tropics, the vegetation is dominated
by herbaceous plants and low or prostrate shrubs. Again, the same advantages in a
cold environment pertain to the herbaceous life forms-but to a considerably lesser
extent over low shrubs than over trees. The more severe the environment, the more
likely the shrubs are to be prostrate or restricted to snow accumulation areas. The
amount of new wood necessary each year is very small because the support function is
not needed in prostrate shrubs. As a result, such shrubs are almost as well adapted to
severe conditions as herbs, and some species occur almost to the limit of plant growth
in the Arctic; Salix arctica and Cassiope tetragona both extend north of latitude 83Qin
northern Greenland (Holmen, 1957). I n such places, some shrub species are likely to
reflect site conditions by their form, each species being very selective of habitat.
Cassiope tetragona (non-prostrate) is of this type and is restricted to snow-covered
areas with soils rich in humus. On the other hand, Salix arctica is prostrate, which
allows it to grow in wind-swept places free of snow in the winter; but it also occurs in
many other sites including those covered with as much as 4-5 m. of snow (Holmen,
   The same type of dwarf shrub adaptation has arisen independently in the sub-
antarctic alpine flora of New Zealand, and is related there (as in the north) to increasing
severity of the environment. For example, A. F. Mark (personal comm.) has found that
species of dwarf shrubs (‘suffrutescent ’) increase in floristic percentage at the expense
of species of ordinary shrubs (‘woody’) and of herbaceous species with increase in
elevation from ‘lower alpine’ to ‘high alpine’ (as defined by Wardle, 1964). These
data appear in Table 2.
   One of the advantages of a prostrate shrub or suffrutescent plant in a severe environ-
ment is that of permanence; its yearly maintenance can be low and its reproduction
need occur only at long intervals. Raup (1965) found a number of shrubs of Salix
arctica in north-east Greenland to be more than 60 years old, some were older than
 IOO years, and one was about 236 years old. Almost all of them showed great irregu-
larities of growth through time and also the ability to remain alive and grow using only
one side of the stem. The long-lived prostrate shrub form apparently provides a
certain amount of flexibility in meeting temporal exigencies in the environment, and
in this respect may be the equal of the well-adapted herbaceous perennial. This is
 particularly true if the shrub is evergreen, for the same advantages exist in evergreen
tundra shrubs that exist in evergreen trees at timberline: it is not necessary to spend
 food reserves on a wholly new photosynthetic apparatus each year. Hadley & Bliss
 (1964) found that evergreen tundra shrubs on Mt Washington had lower respiratory
                           The ecology of arctic and alpine plants                                493
and photosynthetic rates than the deciduous shrub Vaccinium uliginosum or various
herbaceous species. Pisek & Knapp (1959)also found that alpine evergreen shrubs and
timberline evergreen trees have lower respiratory rates than deciduous ones in the
same environments. T h e evergreens can operate at these lower rates since they do not
have to produce a wholly new photosynthetic apparatus each year. Hadley & Bliss
(1964)also suggest that the older leaves may act as winter food storage organs since
lipids and proteins are mobilized and translocated from old to new leaves during the
growing season.

              Table   2.   Percentages of shrubs, dwarf shrubs, and herbs in the
                                  alpine Jlora of New Zealand
                     (Data from A. F. Mark. Alpine zones after Wardle (1964).)
                                                        occurring in
                                             Species     both ‘ lower      Species
                                           confined to   alpine’ and     confined to    All ‘alpine’
                                         ‘lower alpine’ ‘high alpine’   ‘high alpine’     species
    Total no. of species. . .               358            161              79            598
    % ‘woody’ (ordinary shrubs)              21’2            7‘5              .
                                                                             63            15.6
    yo ‘ suffrutescent’ (dwarf shrubs)       13.1           37-9            392            23.2
    % ‘herbaceous’                           65.6           54’7            547            61.2

   As well adapted as some shrub species are, the greater number of higher plant
species beyond timberline are herbaceous. Almost all of these herbs are perennials
with large underground root or stem storage systems; few are biennial, and fewer still
are annual. Of the ninety-six species of higher plants in the flora of Peary Land,
North Greenland (Holmen, 1957), only one, Koenigia islandica, is an annual. Similarly,
even in an alpine flora, that of the Beartooth Plateau (Johnson & Billings, 1962),only
three species are annual out of a vascular flora of 191 species. Apparently, in most
arctic or alpine locations, annual species constitute only 1-2yoof the flora. However,
alpine floras in relatively dry regions tend to have somewhat higher percentages of
annuals. For example, in Howell’s (1951)       alpine flora of 108 taxa in the southern
Sierra Nevada of California, six are annual. Went (1953)suggests that at least forty
species of annuals occur at elevations of greater than 3000 m. in the southern Sierra.
Perusal of his table I indicates that perhaps twenty-seven of these species occur as high
as timberline at 3300 m. or even higher. There is no doubt that the annual life form is
more common in the Sierran alpine zone than in that of most other mountain
ranges. This could be partially due to the warm dry microenvironments during the
summer in this great mountain range which emerges above a Mediterranean-type
climate. Additionally, most of the Sierran annual species are not arctic-alpine but are
in genera common in the nearby deserts and may represent upward migration and
evolution from desert to dry alpine conditions.
   T h e typical herbaceous plant of arctic and alpine environments is an herbaceous
perennial with a relatively large root and/or rhizome system. These perennials are of
three main types : graminoid, leafy dicot, and cushion dicot. A few ferns and lily-like
plants also occur.
494                     W. D. BILLINGS H. A. MOONEY
  I n a typical alpine situation, dicots generally have a deep primary root system with
shoots proliferating near the soil surface (Rauh, 1940; Daubenmire, 1941). A few
species have rhizomes or bulbs which are utilized in food storage and a few reproduce
vegetatively by horizontal rhizomes or even roots. Most alpine dicots in western North
America reproduce by seeds rather than by rhizomes. As shown in Fig. 2, western
American Oxyriu digynu plants (south of the Canadian Rockies) lack rhizomes while

    Fig. 2. Typict.. plants of Oxyriu digynu grown in the same environment from different seed
    sources and showing the extensive root system (from Mooney & Billings, 1961).A, Beartooth
    Mts, Wyoming, seed source; non-rhizomatous.B, Sagavanirktok River, Alaska, seed source ;

those in the rest of its arctic-alpine circumboreal range have reproductive rhizomes
(Mooney & Billings, 1961).At the risk of over-generalizing, it appears that spreading
rhizomes may be more typical of arctic perennials than of alpine ones. Whether
characterized by rhizomes or not, the underground portions of most arctic or alpine
perennial dicots have much greater dry weight than the shoots (2-6 times greater) and
act as carbohydrate storage organs (Mooney & Billings, 1960; Scott & Billings, 1964;
Aleksandrova, 1958).This also seems to hold true for alpine herbfields in New Zealand,
but less so for alpine cushion plant communities there (Bliss & Mark, 1965). T h e
New Zealand alpine cushion plants are essentially dwarf evergreen shrubs-and like
Bliss’s M t Washington alpine shrubs, most of the storage is in the form of lipids in the
leaves rather than in roots. I n Celmisiu viscosu the leaves contain from I 5 yo to 20%
lipids (Bliss & Mark)-and burn easily. Apparently, grasses and sedges have smaller
root:shoot ratios than do dicots (Scott & Billings, 1964); the reasons for this are not
                      The ecology of arctic and alpine plants                          495
known. Rhizomes and stolons are more frequent in these graminoid plants than in
dicots, particularly in the Arctic.
   In addition to the large food storage capacity, the herbaceous perennial, whether
dicot or monocot, has another adaptation which makes the most of a short cold
growing season: the pre-formed shoot and flower bud. This has been illustrated for
almost all of the plants of north-east Greenland by Ssrensen (1941).Hodgson (1966)
reports it for Alaskan grasses also but it is absent in most temperate zone grasses.
We, also, have found it to be almost universal in alpine plants as well. The flower-bud
primordia are often initiated early in the growing season of the year before flowering
and are usually well developed by the time the perennating bud is formed late in the
summer and winter dormancy comes on. In some species, the first indication of floral
primordia is two seasons ahead of flowering (Ssrensen). The advantages are obvious:
shoot elongation and flowering can and do occur very quickly after snow-release and
temperature increase. As a corollary, flowering is thus also dependent upon environ-
mental conditions during floral initiation in the year preceding actual flowering. Most
of these flower buds overwinter quite close to the soil surface and in snow-free areas
are subjected to intense cold. In Braya humilis, the fruiting inflorescence itself regularly
overwinters and completes development the following summer in spite of growing in
exposed sites almost without snow and subject to bitter cold (Ssrensen, 1941).
   Mark (1965b) has found pre-formed flower buds in the narrow-leaved snow tussock,
Chionochloa rigida, in New Zealand. Apparently, pre-formed flower buds are quite
common in New Zealand alpine plants (Mark, personal comm.), which indicates
parallel evolution of this adaptation in both Northern and Southern Hemispheres.
   The disadvantages of the annual life form in a cold short growing season lie princi-
pally in the necessity to complete the whole life-cycle from germination to seed pro-
duction. This demands an ability to germinate and carry on metabolism at high rates
at low temperatures. In fact, although data do not seem to be available, it might be
postulated that annual plants in an arctic or alpine environment should have higher
photosynthesis and respiration rates than perennial plants from the same locations at
the same temperatures. Annual plants do better under desert conditions because even
though the growing season is as short there as that in the tundra, temperatures are
much higher and growth and reproduction consequently more rapid. Whether in the
desert with its short period of water availability or in the alpine zone with its short
period of growing temperatures, annuals are of necessity small and delicate. Only in
non-stress locations (such as the moist subtropics and tropics) is it possible to afford
the metabolic luxury of a large annual plant. Koenigiu islundica, certainly one of the
most widespread of arctic-alpine annuals, and often the only annual species present in
a tundra region, is extremely small in size in both vegetative and reproductive organs.
Dahl (1963) believes that Koenigia is confined to wet sites because of the necessity
of keeping the plants' temperatures below the lethal limit of 45" C. by evaporation.
This seems quite reasonable and undoubtedly is important in allowing the species to
exist in geographic areas otherwise too warm for it. However, a wet environment might
have other advantages for an annual in an environment with cold nights: the thermal
stability of the surrounding water might allow warmer night temperatures near the
496                    W. D. BILLINGS H. A. MOONEY
plants, thus keeping respiration and growth rates relatively high for longer periods
of time. This could be an advantage to an annual plant faced with a short growing
season at low temperatures. One might expect to find the few typical widespread
arctic-alpine annuals in those microenvironments with the longest periods of stable,
moderate temperatures. This might not be true of most of the Sierran alpine annuals,
which being desert-derived are mostly in the warmer and dryer sites such as sandy,
south-facing slopes.
   The small stature of arctic and alpine perennials is essentially genetic (Turesson, 193I ;
Clausen, Keck & Heisey, 1940, 1948; Mooney & Billings, 1961; Lona, 1963; Landolt,
I 967) but phenotypic plasticity often allows considerable morphological adjustment.
On the other hand, extreme dwarfism, without much phenotypic plasticity, is in-
herent in certain populations of widely distributed tundra species (Oxyria digyna,
Polygonium viviparum). These dwarf plants show up in cultivations in growth chambers
and in phenotypic form in herbarium collections. I n the Arctic, they are principally
from the ‘high Arctic’ in such places as Ellesmere Island and Thule, Greenland. We
have designated these as ‘arctic dwarfs’. Similar genetic ‘alpine dwarfs’ occur on
many high exposed mountains. For example, on Medicine Bow Peak, Wyoming, near
the summit at 3650 m. there is a population of dwarf Oxyria, and similar populations
of Trisetum spicatum (Clebsch, 1960) and Thlaspi alpestre (Rochow, 1967). Plants from
these populations are much smaller than plants of the same species 300-400 m. lower
on the mountain when grown side by side in controlled conditions. Since the site near
the summit was above the Pleistocene glaciers that buried the east face of the peak,
and is somewhat isolated at the top of a 300 m. cliff, there is the possibility that these
populations of dwarf plants represent descendants of populations that survived full
glacial conditions on the nunatak summit of the peak.
    The most severe alpine or arctic environments are of two general types: (I) late-
lying snowbank areas where the growing season is extremely short, and (2)windswept,
dry ridges. Both of these microenvironments become more severe toward the ‘high
Arctic’ and the ‘high alpine’. Plants are absent in the most severe situations in either
late-lying snowbanks or exposed ridges. However, at the limits of plant growth in
both cases, plants tend to be cryptogams-for the most part mosses in the snowbeds
 and crustose lichens and sometimes mosses on the windswept, rocky barrens of the
hilltops. In the latest-lying snowbeds of the Scandinavian mountains, for example,
 Gjaerevoll (1956) reports that vascular plants are almost lacking and the sites are
 dominated by various moss communities. Similar situations exist throughout the
 Arctic and particularly in alpine regions, although mosses are relatively unimportant
 in alpine snowbeds in the western American mountains where the transition from
 grasses and sedges to bare ground and rock under late-lying snow is a sharp one
 (Billings & Bliss, 1959).T h e ‘high Arctic’, north of 77O is so dry and cold (Savile, 1961,
 1964; Holmen, 1957) that snowbed vegetation is considerably restricted in size and
 also in flora. I n places, snowbed vegetation may even be absent in the sense that it
 occurs farther south in moister climates. Where snowdrifts do occur in the high
 Arctic, much of the water probably sublimes into the dry summer air; some of the
 meltwater may seep rapidly into the porous, rocky soil to emerge lower down slope
                       The ecology o arctic and alpine plants
                                    f                                                  497
as the main source of water for other communities in this dry, barren land (Savile,
1961, 1964).
   Windswept ridges provide the most severe environments in terms of temperature,
 drought stress, and wind abrasion. Beyond the limits of any vascular plants, crustose
 and sometimes fruticose lichens will be found. In the high Arctic, hilltops and ridges
 are virtually unvegetated except for such lichens and the grey moss Rhacomitrium
 lanuginosum, with an occasional plant of Saxifraga oppositifolia (Savile, 1961).   Polunin
 (1951) that such barrens are ‘desolate in the extreme’ and that ‘most areas have
 scarcely a plant to be seen’. Similar, but smaller, areas occur on high alpine ridges but
the terrain is usually so varied that the term ‘barrens’ is scarcely appropriate except
 perhaps on the high Tibetan Plateau and in some of the high mountain ranges of the
 Eurasian and American deserts (the White Mountains of California, for example).
   Comparable domination of moist areas by moss carpets and dry, rock areas by
crustose and fruticose lichens characterizes the maritime Antarctic (Longton, 1967).
Mosses become much less important in continental Antarctica and the only plants in
the most severe environments are crustose lichens (Greene et al. 1967).
   At the limits of vascular plants in snowbeds, there is apt to be a variety of life forms :
graminoid, small herb, cushion plants, prostrate shrubs such as Salix arctica and
S. herbacea. On windswept ridges, however, small-leaved dwarf rosette, moss-like,
or cushion plants with occasional grass tussocks represent vascular plants at their
limit of tolerance. In the Arctic, such plants are Saxifraga oppositifolia, Papaver
radicatum, Erysimum pallasii, Saxifraga tricuspidata, S . caespitosa, Cerastium arcticum,
Luzula confusa, Poa abbreviata, Hierochloe alpina, and others (Savile, 1961,1964;
Polunin, 1951 Holmen, 1957). Plants of these species not only can withstand the
winter cold and wind without a snow cover but also tolerate the drought of summer in
the most severe environment the Arctic has to offer.
   Similar rosette, moss-like, or cushion plants and dwarf graminoids characterize the
windswept rocks of high alpine locations. However, high on the rocky peaks or gipfels
there are also miniature ‘snowbeds’ in the crevices of the rocks; these shelter and
supply water to plants of a softer nature such as Ranunculus which depend upon melt-
water to a greater extent than do the cushion plants. In many cases, these softer plants
reach higher elevations than do the moss-like or cushion plants on the more exposed
sites. Such ‘snow-cranny’ plants extend well up into the zone of permanent snow
where their only plant associates are crustose lichens and a few moss tufts. Reisigl &
Pitschmann (1958)    have compiled data for the Austrian and Swiss Alps on the highest
elevation reached by alpine vascular plants. Cushion or moss-like plants such as
Minuartia sedioides (3825 m. on the Monte Rosa), Saxifraga moschata (4000m. on the
Finsteraarhorn), Draba $adzinensis (4150 m. on the Rimpfischgrat), and Androsace
alpina (4200m.on the Matterhorn) go well up onto the snowy peaks. Saxifraga
oppositifolia, the outlier of vascular plants in the windswept Arctic, goes only to
3540 m. on the Triftjoch, also windswept. But equally high are the relatively soft
‘snow-cranny’ plants Linaria alpina (4150 on the Rimpfischgrat), Poa alpina
(41 m. on the Rimpfischhorn) and Ranunculus glacialis, which reaches the highest
level of any vascular plant in the Alps at 4270 m. on the Finsteraarhorn. A parallel
498                    W. D. BILLINGS H. A. MOONEY
situation exists at the other end of the world in New Zealand, where Ranunculus
grahami exists in snow-crannies well above the permanent snow line as high as 2900 m.
on Malte Brun (Fisher, 1965); this is about the altitudinal equal of any vascular plant
in New Zealand. Fisher believes that this species with its soft foliage overwinters in
the dormant form (as do R. glacialis and Oxyria digyna in the Northern Hemisphere)
and that overwintering by evergreen rosettes could not easily endure the frequent
avalanches and compacted ice covers of the high-altitude winters.
    Webster (1961)lists the highest elevations known to be reached by vascular plants.
Since the lowest elevation on the list is 5500 m., all of the locations are, of course, in
the Himalayas, Karakorams, or Andes. The highest vascular plants listed are Lagotis
glauca, Potentilla saundersiana, Arenaria sp., and Pedicularis sp. photographed by
Oleg Polunin at 5945 m. on Mohala Bhanjyang, Nepal. Since Webster’s report,
Swan (1967) has found Stellaria decumbens as high as 6136m. on Makalu in the
Himalayas. This apparently was not flowering, and yet, not far below, at 6100 m.,
Parrya lanuginosa, Pegaeophyton scapgorum, and Gentiana urnula were in bloom as
early as 27 May. While the Stellaria and Parrya, at least, are small-leaved cushion or
moss-form plants, Lagotis glauca and perhaps some of the others have relatively large
leaves and are of the same life-form as Ranunculus glacialis. The Lagotis, Potentilla,
Arenaria and Pedicularis at 5945 m. on Mohala Bhanjyang were on a gravelly fell-field.
The plants on Makulu were confined to what Swan calls the ‘rock-base niche’, where
small snowdrifts are captured and provide meltwater as a result of thermal re-radiation
from the rocks. The ‘ rock-base niche’ is the ecological equivalent of our ‘snow-cranny ’.
This protected and water-supplying microenvironment allows vascular plants to reach
their altitudinal limits in the cold, windy, and dry environments of the high peaks
where most snow sublimes or blows away; thus, lack of liquid water is the principal
limiting factor.
    As in the polar lands, lichens and, to a lesser extent, mosses occur beyond the limits
 of vascular plants on most high mountains. This is true particularly of crustose
 lichens on the rocks in the mountains of Europe and North America, at least. Mattick
 (1950) states that lichens exist as high as 6200 m. in the Himalayas. According to
 Swan (1967)’ however, lichens are scarce or lacking at these high elevations where
 vascular plants occur above 6100 m. He ascribes this to the lack of surface water from
 snowmelt. Snow-blast could also be a factor, as Rudolph (1967) notes in Antarctica.
 At elevations in the Himalayas below 5750 m., subsurface liquid water from snow-
 melt allows the development of little ‘oases’ of vascular plants in the scree. Here, there
 are lichens tangled in with the small herbaceous plants. The relative aridity due to low
 temperatures and snow sublimation thus confining plants to portions of scree slopes
 with underground meltwater sources or to rock-base niches is strongly reminiscent
 of Savile’s (1961)observations in the high Arctic on Ellef Ringnes Island.
    Apparently, lichens can withstand environments with lower ‘growing ’ season
 temperatures than can vascular plants, if there is a supply of surface liquid water at
  some time during the year. If there is not such a surface supply, rooted vascular plants
  may actually do quite as well if not better than lichens. The former situation prevails
  in places such as the relatively moist west coast of the Antarctic Peninsula, where the
                      The ecology of arctic and alpine plants                        499
richness of the lichen vegetation is unequalled in most other parts of the world
(Rudolph, in Greene et aZ. 1967),but the vascular vegetation is depauperate and
consists of only two species. Of the cu. 350 species of lichens known from Antarctica,
ca. IOO have been found only on the west coast of the Antarctic Peninsula. The east
coast of the peninsula, being drier and colder, has only ca. 20 yoas many lichen species
as the whole peninsula. The lichen flora and vegetation of the rest of the continent is
relatively depauperate, a fact related to cold and aridity. The same decrease in luxuri-
ance in lichen vegetation may be seen in a transect from the relatively moist low Arctic
to the arid high Arctic. And, of course, we have the situation of 'frozen aridity'
presented by Swan for the high Himalayas.
   So, provided there is an occasional supply of water, the lichen life form seems more
efficient and better fitted to extremely cold summer situations than are most vascular
plants. Why is this? There appear to be several advantages to the lichen life-form in
a cold environment. First, the highest daytime temperatures are close to the rock- or
soil-air interface where the lichens exist. This can produce meltwater and allows
photosynthesis in the lichens to operate efficiently. The very low night temperatures
at the same interface do not damage lichens. Furthermore, Lange (1965) found    has
optimum photosynthetic temperatures for several species of alpine and antarctic
lichens to be around 5" C. All of these lichens had positive net photosynthesis in light
even at several degrees below 0" C. (as low as - 1 ' C. in Parmelia coreyi from Victoria
Land, Antarctica, and StereocauZon atpinurn from the Otztaler Alps). Presumably,
most of the water in the thallus was frozen at these temperatures (Scholander, Flagg,
Hock & Irving, 1953).Some CO, uptake occurred at temperatures as low as -24" C.
in lichens from a lowland dry site in Europe, so there appears to be little correlation
of site with the ability of a lichen to take up CO, in the light at low temperatures. The
lowland European lichens may well be more active in the winter than they are in the
dry summer, while the reverse is probably true of the polar and alpine lichens.
   Another possible advantage of lichens over vascular plants in a cold or dry environ-
ment is the ability to absorb water rather quickly into the photosynthetic system.
If temporary liquid water is available, it can move almost directly into the algal cells
which are at a favourable temperature for photosynthesis, thus circumventing the
need to enter a conducting system (with its resistances) by way of soil and roots. It
would not be surprising to find some high-altitude vascular plants with similar rapid
absorption by leaves. The lack of roots in lichens and mosses has another advantage:
Hedberg (1964)    reports both kinds of plants riding in loose balls or plates on the tops
of active solifluction areas on Mt Kenya.
   Lastly, lichens are long-lived perennials-up to I 300 years for Rhixocarpon geo-
graphicum in the Alps and 4500 years in West Greenland (Beschel, 1961). such,    As
there is no need to produce new biomass each year. Near the limits of plant growth
this is a real advantage, since favourable years for growth may be far fewer than un-
favourable ones. Lichen growth may be thought of as being 'opportunistic', thus
taking advantage of temporary environmental events to an extent not possible in most
vascular plants. The reverse is also true: lichens can remain dormant for long periods
of time at low temperatures, and when thawed resume normal physiological activity.
500                   W. D. BILLINGS H. A. MOONEY
Lange (1966) measured the photosynthesis rates of Cladonia alcicornis at various
temperatures then froze the thalli at -15' C. for periods of time to up I I O weeks;
photosynthesis rates were normal after thawing.

      (2) Physiological ecology of the life-cycle in arctic and alpine vascular plants
  T h e ability of a vascular plant to survive and reproduce in arctic or alpine environ-
ments is based on the physiology of the whole plant in its natural milieu. Bliss (1962b)
has provided an excellent summary of this adaptational ecology of arctic and alpine
plants. Also, Larsen (1964) has reviewed the roles of physiology and environment in
the distribution of arctic plants, while Tranquillini (1964) has done somewhat the
same thing for plants at high altitudes. Here, we shall summarize and evaluate the
current state of knowledge concerning the physiological ecology of both arctic and
alpine plants. As we have done with Oxyria digyna (Mooney & Billings, 1961), the
basis of the presentation will be the life-cycle, starting with the seed.

Seed dormancy and germination
    T h e weather during some growing seasons in the Arctic or in the high mountains
is so cold that flowering and fruiting are seriously hampered and little or no viable
seed is produced. However, in other seasons, some plants may have abundant seed
crops. Because of the shortness of the summer season, such seeds usually ripen not
long before the return of winter conditions. Indeed, they may not ripen at all during
the year of flowering but, as in Braya humilis in north-east Greenland, the developing
fruits may become dormant, go through the winter, and ripen the following summer
(Serensen, 1941).
    Because of the early onset of winter, one might anticipate finding almost universal
seed dormancy as a protective mechanism, but such is not the case. Given a suitable
environment in the laboratory, fresh seeds of most alpine and tundra plants germinate
rather easily (Soyrinki, 1938, 1939; S~rensen,1941; Bliss, 1958; Bonde, 1965a, b ;
Sayers & Ward, 1966). Seeds of some species, however, do show dormancymechanisms
of one kind or another (Pelton, 1956; Amen, 1965; Rochow, 1967). I n understanding
the extent and role of seed dormancy in alpine species, we are aided by Amen's recent
review of the subject (1966). Of sixty-two alpine species listed by Amen (based on the
individual works of Amen, Bliss, Bonde and Pelton), only about 40% showed dor-
mancy at time of harvest. Carex, Trifolium, and Salix are the principal genera having
 dormant seeds. Most dormancy is caused by seed-coat inhibition (Carex, Luxula,
 Thlaspi, Trifolium) and can be overcome by scarification. A few species, such as
Erythronium grandiforum, require a chilling period of ca. 4' C., while Cerastium
beeringianum and Saxifraga rhomboidea go through after-ripening phases at higher
temperatures. Amen's review indicates that seed dormancy is more common among
 dominant and abundant species; he suggests that such dormancy may contribute to
their success.
    There is some indication of ecotypic variation in dormancy or degrees of dormancy.
 I n Oxyria digyna, germination is better in light than in dark (Mooney & Billings, 1961);
 this indicates that some seeds in a population have a light requirement for germination
                      The ecology of arctic and alpine plants                       501
 while others do not. This light requirement trait in Oxyria is more frequent in
 American alpine populations (72-97 yo of viable seeds being light-requiring) than in
 arctic ones (20-64 % of viable seeds being light-requiring).
   Intrinsic seed dormancy, then, apparently is not common among arctic or alpine
 species. Even such dormancy as does exist is easily overcome by scarification, chilling,
light, or simply elapsed time. We can conclude that dormancy in nature is principally
 under the control of the tundra environment and its low winter temperatures. Seed
usually matures too late in the season to meet the right combination of temperature
 and moisture for germination.
   Optimum germination temperatures, as measured by speed and completeness of
germination, are surprisingly high for most tundra species (20-30" C.) ; such tempera-
tures are comparable to those at which seeds of most temperate and tropical species
germinate. In our work with Oxyria (1961) the optimum temperature was 20" C.
 There was still germination at 10' and 30°, but none at 3" C. Strangely, the southern-
most alpine populations showed better germination at 10"C. than did those from the
Arctic, while the arctic populations did better at 30" C. than did most of the alpine
collections. R. H. Wagner (personal comm.) tested some of our Oxyria seed collections
on a temperature gradient bar germinator and found curves similar to ours for alpine
populations from the Donjek Mountains, Yukon, and Niwot Ridge, Colorado. How-
ever, an arctic population from Pitmegea River, Alaska, showed IOO % germination
at 10-12' C. and no germination at 8-10" C., a sharp cut-off indeed.
   Amen (1966) states that there is no evidence to suggest that any alpine seeds can
germinate at temperatures below 10" He, of course, is referring to constant tempera-
tures; Wagner's data are also from constant temperatures. Sayers & Ward (1966) also
got no germination of any alpine seeds at a constant 5" C. This seems surprising since
the seeds of several cereal grasses and weeds are reported to germinate at temperatures
as low as 3-5" C. (Mayer & Poljakoff-Mayber, 1963). Also, Juhren, Hiesey & Went
(1953) found that seeds of several species of Poa, Agropyron, and Bromus from western
North America will germinate fairly well at a constant temperature of 3-5"C.
   The only evidence which we have found of germination in a tundra species at a
constant temperature of 5" C. is that reported by Holtom & Greene (1967) for
Deschampsia antarctica, one of the two antarctic angiosperm species. The other,
Colobanthus crassifolius, had no germination at 5" C. However, seeds of both species
germinated well if exposed to temperatures which fluctuated between 5' and 18" C.
This result confirms what apparently is a typical reaction to temperature in germina-
tion of tundra seeds. For example, we (1961) tested seeds of four populations of
Oxyria at temperatures which cycled once every 24 hr. between 13" and 2" C. (mean
7.5' C.), thus approximating summer temperatures on the Alaskan arctic coastal plain.
Germination (as high as 65 %) took place in all populations. Vickery (1967) got similar
results ;germination occurred in all eleven populations tested in two western American
species of Mimulus at temperatures alternating between 9" and 4" C.
   Field temperatures in arctic and alpine sites are not constant. The germination that
occurs there is thermally opportunistic, mostly at temperatures lower than the labora-
tory optimum, and over a much longer period of time. For example, S~lrensen      (1941)
                      W. D. BILLINGS H. A. MOONEY
found that germination in several species at Eskimonaes in north-east Greenland could
take place in early June even though the surface soil and seeds were frozen for many
hours each day. Apparently, arctic or alpine seeds germinate better at low alternating
temperatures than they do at low constant temperatures; but more research is needed.
   When does germination occur in the field? The best answers to this question may be
found in the observations of Soyrinki (1938,1939)and Ssrensen (1941).        With their
observations and the available laboratory temperature data as guides, germination
appears to take place a week or so after snowmelt in the early summer after soil surface
temperatures get up to IOO or 1' the daytime, and before the soil dries out. Most
germination in late summer is impeded by lack of soil moisture in the upland tundras
and most (but not all) of the species there produce non-dormant seed. Late soil
moisture is available in most wet meadows; many (but not all) of the principal species
in such places have seed dormancy mechanisms (Amen, 1966): Carex, Luxula,
Erythronium, Saxifraga rhomboidea, Deschampsia caespitosa, Polygonum bistortoides.
However, some plants in such wet places do produce seed that germinates the same
year. Bliss (1958) found that seeds of Salixplanifolia var. monica in the Rocky Moun-
tains ripen in July and germinate immediately in the wet habitat. On the other hand,
seeds of Salix brachycarpa do not ripen until August, when its site has become rather
dry; these seeds overwinter in viable form and germinate early the next summer.
   We can conclude that almost all seed germination in alpine locations, at least, takes
place in early summer after snowmelt during the year following seed production.
Most species lack a seed dormancy mechanism and the elapsed year (or more) between
 seed production and germination is environmentally imposed. There are a few species,
however, which do have an intrinsic seed dormancy mechanism of one kind or
another; the commonest type is caused by a hard and impermeable seed coat. Many
but not all of these species with seed dormancy are dominant and abundant in moist
 or wet alpine meadows.
    Since tundra environments may not be suitable for germination every year, we
would like to know how long such seeds may remain viable. Even under good germina-
tion conditions, there is often considerable intrapopulational variation in how long it
 takes a seed to germinate. In one sample of snow tussock (Chionochloa rigida) seed
from the New Zealand mountains, Mark (1965b)found (at 2 ' C. in the dark) that
 after 50 days only 31.57~ germinated, 40% had germinated at 109 days, and the
 last germination did not take place until 1450days, almost 4 years after the start, at
 which time 94% of the seed had germinated and the experiment was terminated.
 Such intermittent intrapopulational germination over a long period of time could have
 survival value for a tundra population near the limits of growth.
    If it took almost four years for some of Mark's snow tussock seed to germinate in
 the laboratory at 2' C. with moisture, these same seeds might have retained viability
 in the cold field environments for many years before the right combination of con-
 ditions allowed germination. In our laboratory at Duke University we have over
 1000collections of arctic and alpine seeds stored dry at -18" C. Most of these seeds
 retain their viability indefinitely in this frozen condition; our oldest collections are
 now 15 years old.
                       The ecology of arctic and alpine plants                        503
   A possible answer to how long arctic seeds may remain viable is provided by the
recent paper of Porsild, Harington & Mulligan (1967) the germination of seeds of
Lupinus arcticus recovered from permanently frozen burrows of the collared lemming
(estimated age 10,000   years B.P.) in unglaciated central Yukon. The seeds may have
survived at least 10,oooyears of freezing and 12years of dry, normal temperatures after
they were collected and before they were given to the National Museum of Canada.
Six of the seeds germinated within 48 hr. on wet filter paper (presumably at room
temperature) ;the seedlings were transferred to pots, and at least one has subsequently
flowered. This may not be too surprising when one recalls that Scholander et aE. (1953)
found a steep drop in respiratory rate in living arctic plants when frozen, so that the
respiratory Ql0 at -5' C. was between 20 and 50, where it levelled off with further
drop in temperature. As Scholander and his co-workers suggested at that time, if the
survival time of an isolated organism depends upon the rate at which its energy stores
are used up, an organism capable of surviving 'a modest 10days' at oo C. would last
for 1000years at -23" C.with a Qlo of 50, and for I,OOO,OOOyears at -42" C. These
are conservative estimates when one considers that the Lupinus arcticus seeds were
capable of surviving 12years of ordinary temperatures after recovery and before being
put in moist conditions suitable for germination. It would not be surprising if viable
lichens (see Lange, 1966) and seeds could be recovered from permafrost cores far
older than 10,ooo years.
   We can conclude that if seeds of certain arctic and alpine plants remain frozen in
soil or perhaps glacial ice, their period of viability may be very long. Certainly, such
viability could span time periods as long as full glacials and, in the right places, even
interglacials. Thus, delayed germination could act in lieu of refugia. It might allow
the re-establishment of a species in situ tens of thousands of years later after the retreat
of the ice and permafrost. This could help to explain some apparently rapid 're-
invasions ' and some puzzling species distributions in the Arctic.

Seedling establishment
   Even less is known of seedling establishment in tundras than is known about
germination requirements. Considerable information is provided by Soyrinki (I 938,
1939). Wager (1938)has also made observations on seedling establishment in the
   Since most seed germination in the tundra occurs in early summer, the seedling has
only a few weeks to develop a root system and to produce enough carbohydrates to
allow survival through the following winter. There are several reasons why years may
pass between episodes of successful seedling establishment of a species in a given
tundra location. First, the temperature must be warm enough for germination to take
place. Secondly, this must occur early enough in the summer to allow time for good
growth before the return of temperatures constantly below freezing, and in some
years these come unusually early. In regard to this point, Bliss (1958)notes that
seedling reproduction is more common in those species growing on deeply thawed
soils in the Arctic. Thirdly, the seedlings must not be exposed to drought in the latter
half of the summer before the root system has penetrated to a reliable water supply.
504                   W. D. BILLINGS H. A. MOONEY
The more severe the tundra (high Arctic, high alpine), the fewer are the years in
which all of these conditions are met and in which, therefore, there is a good chance
for seedling survival.
   Perhaps it should not be surprising that germination temperature requirements are
fairly high. If temperature thresholds were lower, germination could be triggered in
a season too cold for adequate photosynthetic and respiratory rates in the seedling-
or too late in a moist season in which the seeds were produced. Establishment would
thus be hampered by too little root and leaf growth to prevent death by drought,
frost-heaving, or inadequate carbohydrate storage.
   Seedling growth is very slow under natural conditions and at the end of the first
season the seedling is still very tiny in most species. Wager (1938), at Kangerdlugssuak
(68" 30' N.) in East Greenland, found that Oxyria seedlings at the end of their first
summer had two cotyledons and two leaves of 2.5 mm. diameter. Comparably small
seedlings were typical of other species. In Saxifraga oppositifolia, growth was so slow
that the true leaves did not develop until the second year.
   We have grown thousands of Oxyria seedlings during our research activities and
have watched seedling development in a qualitative way. At temperatures averaging
20" C. seeds of all populations (arctic and alpine) germinate readily and the cotyledons
appear above ground in 3-5 days. Shoot growth then becomes rather slow but root
growth goes on rather rapidly with the development of a tap root several centimetres
long within a week or two, By the end of a month, two or three leaves have appeared;
at this time, shoot growth is resumed at a fairly rapid pace. In nature, this resumption
of shoot growth does not occur in most Oxyria populations until the second year.
The first field growing season is devoted primarily to root-system establishment. This
is important to survival since the surface soil may become very dry in late summer in
many alpine locations and even in the Arctic. There is an interaction also between
 frost-lifting of seedlings with too-shallow root systems during the ensuing winter and
 subsequent greater susceptibility to drought-killing in the second summer (Mark,
    Such lifting by needle-ice action or frost-thrusting is a real danger. It has been
 shown by Hedberg (1964) for the high African mountain plants, where seedlings
 starting in open soil are rapidly eliminated by frost-lifting which occurs diurnally on
most nights of the year. In spite of root competition, seedlings of many afro-alpine
 species stand a better chance of survival in closed vegetation rather than in open soil.
 The result is a patchiness in vegetation in the upper parts of the alpine zones of
 Mt Kenya and Kilimanjaro.
   While there are advantages from the standpoint of light and moisture for a seedling
 starting in the open, there are also advantages in a needle-ice or solifluction area for
 seedlings to have the protection of well-anchored mature plants. This may be at least
 a partial explanation for the observations of Griggs (1956) in the fell-fields of the
 Rocky Mountains that seedlings of many alpine plants are more common and survive
 better in the cushions of Silene acaulis and similar plants than they do in the open.
 Such cushion plants with their deep tap-root systems are the pioneers in bare gravels
subject to frost action and other disturbances.
                      The ecology of arctic and alpine plants                            505
   Reproduction by seedlings appears to be most common and most important in the
drier alpine sites than in any other arctic or alpine environmental type. However,
seedling establishment every year is necessary only in the few alpine annuals such as
Koenigia. Even here, it is probable that the seeds may remain viable in the ground for
several years. With the abundant and slow-growing perennials, seedling establishment
need only occur now and then for the population to be maintained. In the Arctic and
in wet and moist meadows of alpine locations, vegetative reproduction by rhizomes or
runners is more common by far than reproduction by seeds.

Chlorophylls and other pigments
   Standing crops of chlorophyll in tundra vegetation have been measured by Bliss
(1966) for alpine communities on Mt Washington (1917 m.) in New Hampshire and
by Tieszen & Johnson (1967) for arctic communities on the northern coastal plain of
Alaska. The combined chlorophyll a + b contents for seven community-types on
Mt Washington ranged from 0.18 td 0.90 g./m.2; in four arctic types the range was
0.32-0.77g./m.~. On a land-area basis, chlorophyll content is roughly the same in
these two relatively moist tundra regions. Furthermore, they are equivalent to chloro-
phyll contents in temperate herbaceous communities of Minnesota as reported by
Bray (1962). Comparable land-surface chlorophyll data do not seem to be available for
the drier and higher mountains of western North America.
   On the basis of individual plants within single species, Henrici (1918) in Switzerland
found that on a fresh-weight basis lowland forms (250-450 m.) had over twice as much
chlorophyll as alpine forms (1700-2450 m.) of the same species. Her results are on a
relative basis only, with Urtica dioica being used as a standard. Seybold & Egle (1940),
however, found that, on a leaf-area basis, alpine plants in general had just as much
chlorophyll as lowland plants. They did confirm Henrici’s results that some alpine
plants have less chlorophyll than Urtica on a fresh-weight basis-but so did some
lowland plants.
   There are no data available for chlorophyll content in alpine plants of western
North America as compared to lowland plants of the same species in the same
latitudinal region. However, chlorophyll determinations on leaves of Oxyria dkyna
collected and preserved in the field showed lower values for plants at 3292 m. in
Wyoming than for plants at I 158 m. in Alaska on both fresh-weight (0.387 mg./gfw. u.
0.636) and leaf-area (0.674 mg./dm.2 v. 0.867) bases (Mooney & Billings, 1961). We
also determined chlorophyll contents on plants of eight populations of Oxyvia (four
for the arctic, four from alpine North America) grown from seed under controlled
conditions in two simulated environments : arctic and alpine. Under simulated arctic
conditions, plants from high-altitude alpine populations had less chlorophyll than did
those from the low-altitude arctic populations. The rate of decrease with source
altitude was about the same for both bases: fresh-weight or leaf-area. Under simulated
alpine conditions, there was slightly less chlorophyll in the high-altitude populations
than they had in the arctic chamber. However, the low-altitude arctic populations,
when grown under alpine conditions, had more than twice as much chlorophyll as they
showed under arctic conditions. This was true on both a fresh-weight basis and a leaf-
   33                                                                    Biol. Rev. 43
506                    W. D. BILLINGS H. A. MOONEY
area basis. These data indicate that within this single arctic-alpine species, there is less
chlorophyll in plants of alpine populations than in plants of arctic populations, and
that this difference has both genetic and environmental bases.
    The striking increase in chlorophyll content in the arctic populations under alpine
conditions can possibly be explained by the 8 hr. of darkness out of every 24 which
provided a respite from photo-oxidation. This respite had little efFect on the alpine
populations, which were from locations where such a respite is part of the natural
environment. Clebsch (1960) also found that chlorophyll content on a fresh-weight
basis increased dramatically when an arctic population of Trisetum spicatum was
grown under simulated alpine conditions. T h e genetic potential for higher chlorophyll
contents in arctic races was also noted by Mooney & Johnson (1965) using alpine and
arctic populations of Thalictrum alpinum in a growth chamber programmed for
simulated arctic conditions ; chlorophyll contents of the arctic populations were about
25 yo higher than that of an alpine population.
    Since solar radiation intensities are much higher in alpine environments than in
lowland or arctic situations, one would expect more photo-oxidation of chlorophyll
at high altitudes. I n addition to the genetically determined lower chlorophyll contents
of alpine Oxyrias as compared to arctic ones, the lower field values obtained in alpine
 Oxyrias may be due to daily photo-oxidation. Montfort (I950), for example, found
that plants of most (but not all) lowland species showed much chlorophyll destruction
in the bright light of high altitudes, and that even some alpine plants showed evidence
of chlorophyll damage.
    McWilliam & Naylor (1967) have shown that there is, indeed, an interaction
 between temperature, light and genetic system in chlorophyll content at low tempera-
tures and in bright light. Corn (Zea mays) develops almost no chlorophyll at 16"C.
 and 4500f.c., but at 28" C. chlorophyll accumulates normally. O n the other hand,
 wheat accumulates chlorophyll in a normal manner at both temperatures. They found
 that the lack of chlorophyll at 16"C. in corn was not due to a block in the pigment syn-
 thesizing mechanism but rather to the photodestruction of chlorophyll prior to its
 stabilization in the membrane structure of the chloroplast lamellae. This prolongation
 of the photosensitive stage in chlorophyll synthesis in corn at low temperatures does
 not occur in wheat, which is better adapted to low temperatures. T h e analogy to alpine
 plants o. lowland plants is obvious.
    Shakhov et ul. (1965) suggest that some of the chlorophyll damage at high altitudes
 may be due to ultra-violet radiation. They irradiated arctic and alpine plants in the
 field with additional U.V. and detected changes in chloroplast structure and pigment
 content. They also found that, in the light, there was some recovery from such U.V.
 damage by photoreactivation.
     I n this regard, the alpine U.V. climate and its effects on plants and their pigments
 has recently been investigated by CaldwelI (1968). Almost all U.V. damage to higher
  plants in nature is the result of u.v.-B (2800-3150 A). While direct beam solar u.v.-B
 increases with altitude, sky u.v.-B decreases; the result is only a modest rate of increase
  in global u.v.-B with elevation. On a cloudless summer day in Colorado, Caldwell
  found that integrated u.v.-B irradiance at 4350 m. was only cu. 26 yo greater than at
                       The ecology of arctic and alpine plants                            507
 147om. Leaves of alpine plants (including Oxyria digyna) in a simulated alpine
 environment in the laboratory were subjected to precisely defined u.v.-B irradiation
 (1-4 x 1015quanta cm.-2 sec.-l at 2967 , for 40 min.) without the presence of longer
wavelengths. Chlorophyll breakdown and tissue destruction occurred in some of these
 plants. Photoreactivation of such damage took place when leaves were exposed to
longer wavelengths (2.0 x 1o16quanta cm.-2 sec.-l at 4358 A for 4-5 hr.) immediately
 after doses of u.v.-B irradiation. In the field, natural filtration of u.v.-B by epidermal
pigments (mainly flavonoid compounds) in alpine plants combined with photo-
reactivation are the most likely adaptations to the u.v.-B climate.
    I n regard to chlorophyll content in tundra plants, we come to the following tentative
conclusions: (I) chlorophyll content per unit land area is probably not very different
between arctic and alpine communities occupying the same position on the soil moisture
gradient; (2) within a single species, there is less chlorophyll in alpine forms than in
arctic forms whether on a leaf-area or fresh-weight basis; this difference is both geneti-
cally and environmentally controlled; (3) the evidence is less clear for lowland TJ. alpine
forms in the same latitude but alpine forms apparently have less chlorophyll; (4)    u.v.-B
irradiation and perhaps other short wavelengths break down chloroplasts but longer
wavelengths in the alpine environment simultaneously allow photoreactivation of this
damage; ( 5 ) some species of plants apparently have too slow a rate of protochloro-
phyllide synthesis to keep ahead of photo-oxidation of chlorophyll in bright light at
low temperatures; (4) the lower chlorophyll content of alpine forms causes greater
reflectance and may be involved in the heat balance of the leaf.
   The bright light and low temperatures which result in low chlorophyll contents also
allow a build-up of anthocyanins in the leaf and young stem. This can be seen in the
high mountains, particularly early in the season just after snowmelt and again near the
end of the season just before the leaves die back. The same effect may be observed in
growth chambers by lowering the temperature while light intensity remains the same.
   In Oxyria, plants of some populations appear to produce much more anthocyanin
than do plants of other populations. In North American Oxyrias, plants of alpine
populations have more than those of arctic populations, but there seems to be no
consistent correlation with elevation of source. Some alpine populations are antho-
cyanin-producers, others produce very little. Among European Oxyrias, even some
arctic ones from Lappland have anthocyanin in fair quantity.
   Bjorkman & Holmgren (1958) found a good correlation of autumn anthocyanin
content with source altitude in Solidago virgaurea plants growing in a uniform garden
in Uppsala. Alpine and subalpine ecotypes from 870 to 1350 m. source elevations had
4-5 times as much anthocyanin in their leaves as did plants from two sea-level eco-
types. Apparently there is a strong hereditary tendency to produce anthocyanins in
alpine populations of both Oxyria and Solidago as compared to lowland populations.
   What are the functions of anthocyanins in the leaves of alpine plants? First, it is
possible that anthocyanins act as a ‘sink’ for excess sugars during cool, bright weather
or under conditions of poor nitrogen or phosphorus nutrition. Secondly, they may act
as absorbers of ultraviolet radiation in the epidermis or mesophyll (Cappelletti, 1929-
30). Also, Biebl (1942) found that destruction of epidermal cells in onion bulbs by
508                   W. D. BILLINGS H. A. MOONEY
u.v.-B was in inverse proportion to their anthocyanin content. Caldwell (1968) pro-
vides additional evidence with Sedum rosea, an alpine species, where green epidermis
transmitted almost 3 times as much U.V. as did epidermis with much anthocyanin.
He also found that red leaves of Oxyria digyna and Geum rossii from the Rocky
Mountains were consistently less susceptible to tissue destruction by irradiation of
2967 r f than were green leaves of the same species. Red leaves of both species had
reduced U.V. transmission. Colourless flavonoids in the epidermis also reduced U.V.
transmission in some species. From these results, it seems clear that an epidermis
rich in anthocyanin makes a good U.V. filter.
   A third possible function of anthocyanin in alpine leaves and young stems may be
the absorption of energy at other wavelengths with resultant increase in leaf tempera-
ture. We do not have any evidence for this but suggest it as a possibility.

Photosynthesis and respiration
   The key to successful adaptation to tundra or alpine environments is the develop-
ment and operation of a metabolic system which can capture, store, and utilize energy
at low temperatures and in a short period of time. The short, cold tundra ‘growing
season’ is at the opposite end of an environmental gradient from the moist, warm, year-
long environment of a tropical rainforest. In between lie the seasonal warm summer-
cold winter conditions of temperate forests, meadows, and most crop land. An under-
standing of the workings of the photosynthetic-respiratory systems of tundra plants
will be aided if we ask how such systems compare with their temperate and tropical
counterparts in regard to seasonality, temperature effects, light effects, and the avail-
ability of metabolic gases and water. Fortunately, we have the reviews of Pisek (1960),
Bliss (1962b),Tranquillini (1964)) and Larsen (1964)which can heIp us with the answer.
   Seasonality o carbon metabolism. As Gessner (1960) has pointed out, it is not the
high temperatures nor high light intensities which are the reasons for high productivity
in tropical rainforests but the uniformly good photosynthetic conditions throughout
the year which are responsible. Thus, in true tropical rainforest, photosynthesis has
little or no seasonality. Away from this ideal extreme, all other kinds of vegetation
show some degree of annual cycle in photosynthesis with extreme seasonality being
shown by plants of deserts and tundras.
   The photosynthetic seasons of arctic and alpine tundras are fully as short as those
of desert vegetation but somewhat more regular, The principal limiting factor is low
temperature, and of course, in all but the windy ridge habitat, the duration of snow
cover. The snow-free period decreases with elevation (see fig. 3, Pisek, 1960; also
Winkler & Moser, 1967) and towards the centres of snowbeds so that, at last, any
snow-free period is so short that there is no photosynthetic season. In alpine regions,
thus, the growing season of photosynthetic activity ranges from 3 or 4 months where
the snow melts early in June to none at all. For most alpine plants, the season ranges
from 6 to 10 weeks (Bliss, 1956; Billings & Bliss, 1959; Holway & Ward, 1963) with
low temperatures and snowmelt governing the start of the season; relative soil drought,
return of low temperatures, shortened photoperiod, and perhaps carbohydrate accumu-
lation govern the end of the period. In years of unusually great snow accumulation,
                      The ecology of arctic and alpine plants                      509
the plants may not be uncovered at all or for only a few days of activity before the new
snow and low temperatures of September return (Billings & Bliss, 1959; Moser, 1967).
   Photosynthetic activity in alpine plants occurs throughout most of the snow-free
period when temperatures during the daylight hours are near or above 0" C., but
photosynthesis is not uniform throughout the season. Net photosynthesis of shoots is
relatively low early in the season because of high respiratory rates associated with
rapid growth immediately after snowmelt (Hadley & Bliss, 1964; Bliss, 1966). How-
ever, some of the respiratory carbon dioxide may be recycled by photosynthesis inside
the hollow young stems of rapidly growing plants around snowbanks (Billings &
Godfrey, 1967), thus making up, in part, the photosynthetic deficit during early
growth. As the microenvironment warms up, photosynthesis increases as long as
water is available, According to Glagolev & Filipov (1965), maximum photosynthesis
is reached during the flowering stage.
   In the Arctic, perhaps because of the longer photoperiod, plants may start growing
somewhat earlier than in most alpine situations. Initiation of growth can be as early as
late April or early May in Saxifraga oppositifolia on dry sites in northern Norway and
as early as late May even at 73" 30' N. at Myggbukta in north-east Greenland
(Serrensen, 1941).At the latter location, flowering of Saxifraga occurs before the mean
daily temperature reaches 0" C. Zubkof (1935) also noted that Saxifraga oppositifolia
was the first plant to show signs of life in the spring on Novaya Zemlya at 76" 14' N.
Growth started before daily mean temperatures became positive, so that the metabolic
period of Saxifraga at this high Arctic location was 107 days while the duration of
positive mean daily temperatures was only 78 days. Other dry site plants, notably
many Cruciferae, often start growth early in the Arctic according to Serrensen. Soil
drought usually puts an end to the vegetative season of these plants before low tempera-
tures become effective.
   A common phenomenon both in arctic and alpine locations is the speeding up or
telescoping of phenological events in plants released from snow-cover relatively late
in the summer. Plants released from snow by early or mid-June will take much longer
to reach maturity and have a longer photosynthetic season than those of the same
species released in mid-July. This has been commented upon by Serrensen (1941)for
the Arctic and by Billings & Bliss (1959) and Rochow (1964) for alpine situations.
While these later plants grow faster in shorter time than plants released earlier, they
are smaller and produce less dry matter.
   Effect of temperature on photosynthesis and respiration. Temperatures of air and soil
decrease, in a general way, with increasing latitude and altitude so that arctic and
alpine plants not only must survive low temperatures but spend much of their growing
seasons in such temperatures. What are the relationships between temperature and
metabolic rates in such plants?
   In evaluating such temperature effects, one must attempt to separate those inherent
in the genotype and those brought about by environment, since both sources of varia-
tion are present in the phenotype. There appears to be some degree of genotypic con-
trol over rates of net photosynthesis. Mooney, Wright & Strain (1964) found that
plants of alpine and subalpine species in the White Mountains of California reached
510                     W. D. BILLINGS H. A. MOONEY
maximum photosynthesis at lower temperatures than did plants from lower elevations
when all were grown under uniform greenhouse conditions. T h e same kind of situation
occurs between arctic and alpine populations of both Oxyria digyna and Thalictrurn
alpinurn (Mooney & Billings, 1961;Mooney & Johnson, 1965) with arctic populations
having optima at lower temperatures than do alpine populations grown in the same

      r      +3

       '0     0

       2 -3



                       50        60        70        80        90       100       110
    Fig. 3. Effect of temperature on rates of net photosynthesis and dark respiration of plants
    from an arctic and an alpine population of Oxyriu digyna grown at low temperatures (4-10"C.)
    and at high temperatures (21-32' C.). The effect of environmental history is greater than the
    ecotypic effect of the two widely separated populations. Arctic population seed source:
    Breiddalur, Iceland ; alpine population : Niwot Ridge, Colorado (from Billings & Godfrey,

  Even though the genotypic effect is marked, the effect of environmental regime is
greater. Plants grown under low temperatures not only show optimum photosynthesis
shifted toward lower temperatures (Fig. 3) but the upper limit of positive net photo-
synthesis occurs at a decidedly lower temperature than that in plants acclimated to
higher temperatures (Mooney & West, 1964; Billings & Godfrey, 1968).
                      The ecology of arctic and alpine plants                       511
   The absolute temperature minimum for photosynthesis is also lower for alpine
plants than in plants of lowlands or warmer climates. Pisek, Larcher & Unterholzner
(1967) found that, in summer, Ranunculus glacialis and Oxyria digyna can carry on
photosynthesis at temperatures as low as -6" C. while Mediterranean broad-leaved
evergreens such as Citrus and Laurus had no net photosynthesis below - I " C. Some
lichens can photosynthesize at temperatures as low as - 4 C. but there appears to be
only a slight relationship of this ability with the coldness of their natural environments
(Lange, 1965).
   There is evidence that dark respiration rate at all temperatures is higher in alpine
and arctic plants than in lowland or middle-latitude plants. This effect is both geneti-
cally and environmentally controlled. We (Mooney & Billings, 1961) found that an
Oxyria population from the Arctic had higher respiration rates than did one from an
alpine location in Colorado when grown under the same controlled environment.
Clebsch (1960) found the same thing in arctic and alpine populations of Trisetum
spicatum. However, Billings & Godfrey ( I 968) found relatively small genetic differences
in respiration rates among seventeen latitudinal populations of Oxyria grown in
uniform environments. On the other hand, there were great increases in these rates
when the plants were grown at lower temperatures. Genetic differences in respiration
rates within this species, then, appear to be far less important than the effects of
environmental temperature (see Fig. 3). Furthermore, these environmental effects are
reversible within a matter of days or even hours in a new environment.
   Bjorkman & Holmgren (1961) found that leaves of an alpine ecotype of Solidago
virgaurea had higher rates of respiration at the same temperatures than did leaves of
a coastal ecotype when grown under the same conditions. Mooney (1963) found the
same effect in alpine and lowland races of Polygonum bistortoides. At least a partial
explanation of these genetic and environmental effects on respiration rate in alpine
plants is provided by Klikoff (1966), who found that the oxidative rates of mito-
chondria in four altitudinal populations of Sitanion hystrix (grown under uniform
conditions) showed a direct relationship at 20° C., with source elevation.
   Eflects o light on photosynthesis. Photosynthetic adjustment to light in tundra plants
 is also both genetically and environmentally controlled. Mooney & Billings (1961)
found that arctic Oxyrias reached photosynthetic light saturation at lower light in-
tensities than did alpine populations. Clebsch (1960) had similar results in Trisetum
spicatum. Bjorkman & Holmgren (1963), however, did not find this latitudinal effect
in Solidago virgaurea from 55" 55' N. and 69" 35' N. This may have been because the
latitudinal span was not as great as in the Oxyria and Trisetum populations and possibly
because both Solidago populations were from relatively low elevations in contrast to
the high source elevations of the alpine Oxyrias and Trisetums. They did find, how-
ever, a great effect of light intensity acclimation in these two populations. Plants grown
at low light intensities (3 x 104 erg. cm.3. sec.-l) reached saturation at much lower
intensities than those grown at high intensities (15 x with the
result that maximum net photosynthesis rates that could be attained were almost
twice as high at 30 x 104 erg. cm.-2. sec.3 in the plants grown in bright light.
    Temperature-light interactions. In nature, temperature and light interact with still
5 12                  W. D. BILLINGS H. A. MOONEY
other factors (and environmental history) to produce the daily photosynthetic curve.
Relatively few measurements of photosynthesis of arctic and alpine plants have been
made under field conditions. Cartellieri (1940) using a conductometric method, ob-
tained maximum net photosynthesis rates of 9-13'5 mg. CO, .dm-, (two surfaces). hr-l
with plants of several species in the Tirolean Alps. These compare quite well with our
figures of 4.15-13.4mg. CO,. dm-2 (two surfaces). hr-l (average = 9-32) using whole
plants of four alpine species in an infrared gas analyser system in the Rocky Mountains
(Billings, Clebsch & Mooney, 1966). Moser (1965) has measured photosynthesis in
Ranunculus glacialis, Geum reptans and Oxyria digyna with a gas analyser at 3190 m.
on the Hoher Nebelkogel in the Tirol but his results have not yet been published.
   I n the Arctic on Jan Mayen Island (71"N.) and Cornwallis Island (75O N.),
Warren Wilson (1960), using the detached leaf method, found that for several
species (including Oxyria digyna), the net assimilation rate was between 0-5 and
0.8 g .drn-,. week-l. This is about half the value obtained using the same method for
plants in temperate regions. Warren Wilson ascribes this depression in rate to low
temperatures, since mean July solar radiation at Resolute on Cornwallis Island is
5 10 langleys/day, almost twice the daily rate at London. Since the detached-leaf
method ignores respiratory losses in the rest of the plant, Warren Wilson (1960, 1966)
cautions against comparing such rates with gas analyser results and coming to an
erroneous conclusion that arctic plants in the field have high photosynthetic rates.
   Bjorkman & Holmgren (1963) found in Solidago ztirgaurea that the optimal tempera-
ture of photosynthesis increases with light intensity. Conversely, increased temperature
results in a higher light-saturation level. Scott & Billings (1964) also found in several
species of alpine plants from Wyoming that more light was needed for compensation
and saturation with increasing temperature. Warren Wilson (1966) suggests that the
lower light saturations of arctic plants at low temperatures may be due to accumulation
of assimilates.
   Carbon dioxide and oxygen ejfects. There is some evidence that alpine plants may be
able to carry on photosynthesis at lower carbon dioxide concentrations than lowland
or arctic forms. Using a closed infrared analyser system, Billings, Clebsch & Mooney
(1961) found that with Oxyria plants of equal age and from a uniform environment,
a subalpine population from Montana was more efficient at all carbon dioxide levels
than an arctic sea-level population. T h e subalpine population also had positive net
photosynthesis at carbon dioxide concentrations as low as 0.1mg./l., which approxi-
mates the carbon dioxide tension at an altitude of 12,200 m.; the arctic sea-level
plants could not do this. Mooney, Strain & West (1966) found no apparent photo-
synthetic acclimation to differences in carbon dioxide tensions in plants of four
species after being grown at two widely different elevations: 1250 m. and 3094 m.
   While almost no work has been done on the effects of oxygen concentrations on field
 metabolic rates of alpine plants, rapidly decreasing oxygen concentrations with in-
 creasing altitude would lead one to suspect a pronounced effect on dark respiration
rates, at least. On the other hand, normal oxygen concentrations (21 %) inhibit photo-
synthesis in vascular plants (Bjorkman, 1967); the photosynthetic rate increases at
light saturation as oxygen concentration is decreased to zero. There may be some
                      The ecology of arctic and a&im plants                           513
enhancement of photosynthesis at high altitudes due to low oxygen concentrations,
but controlled field data are lacking.
   Synthesis of information on metabolism. It is obvious that metabolism of arctic and
alpine plants is controlled both by genetic variation and by past and present environ-
ments of the phenotype. The actual diurnal and seasonal course of photosynthesis
and respiration is the resultant of complex interactions between these controls. With
the information available at present, the effect of immediate past and present environ-
ment on the metabolism of an individual arctic or alpine plant seems to outweigh
intraspecific genetic differences. These plants seem to have ample phenotypic plasticity
in metabolism, particularly in regard to temperature. A large part of their ability to
operate in cold environments is probably due to such an ability to acclimate their
metabolism to low air temperatures. Ability to absorb solar radiation and thus raise leaf
and stem temperatures may be helpful also in acclimation. More information is badly
needed on the metabolic plasticity of the phenotypes of different ecotypes of alpine and
arctic species. To do this, gas-exchange methods need to be standardized, environ-
mental preconditioning controlled, and genetic stocks recognized.

Effects of water availability and use
   High mountains are rather efficient producers of precipitation, and as a result, most
alpine areas outside the trade-wind tropics are moist and covered with snow through
most of the year. However, some mountainous areas, notably those in western North
America, are subjected to summer drought of varying degrees of intensity. Much of the
Arctic, because of low temperatures and cloud, is relatively moist, but the lowlands
of the high Arctic are dry, even desert-like, because of low precipitation. I n view of the
occurrence of drought in certain arctic and alpine locations, we may ask what effect
drought stress has on the plants.
   A distinction should be made between drought stress during the summer growing
season and drought stress produced by frozen soils and dry winds of winter. Tran-
quillini ( I 964) indicates that plants in European alpine communities seldom are
subjected to drought during the summer. However, during the winter, in exposed,
snowless sites at timberline and above, evergreen shrubs and small pioneer trees are
subjected to rigorous drought stress. On sunny days, such plants lose a great deal of
water by cuticular transpiration in spite of closed stomates. This water cannot be
replaced because of the frozen soil, and water stress becomes more severe as the winter
goes on. The leaves lose much of their water and the osmotic value rises to over 40
atmospheres (Tranquillini, 1963, 1964; Larcher, 1963). Shrubs covered with snow,
however, actually gain water by absorption so that the osmotic value drops during the
winter. But if such plants (for example, Rhododendron ferrugineum) are denuded of
snow in late winter, they lose water rapidly and may be killed since their drought
resistance is low. O n the other hand, certain dwarf or prostrate shrubs such as Loise-
leuria procumbens, characteristic of open, snowless, windswept places, have low trans-
piration rates all year and tolerate the winter drought very well. According to Larcher,
Loiseleuria can also take up surface meltwater late in the winter by shallow, adventitious
roots when most of the soil is frozen.
514                    W. D. BILLINGS H. A. MOONEY
   Winter drought stress also prevails in the mountains of western North America,
with the same general patterning into snow-protected and snow-free environments.
Lindsay (1967) measured leaf-water potential in Picea engelmannii at timberline
(3300 m.) in the Medicine Bow Mountains of Wyoming and also in the closed forest
250 m. lower in elevation. In the windswept, relatively snow-free timberline location,
leaf-water potential stayed near or below -30 bars from September to early June,
and only during midsummer did it rise to near -15 bars. With the drought of late
summer, leaf-water potential decreased to its low winter values. In the forest location,
leaf-water potential was higher throughout the year. Because of the deep snow cover
in the forest, soil moisture was relatively more available throughout the winter; here,
leaf-water potential did not go much below -20 bars, and this at the same time that
water potential at timberline was -34 bars. Abies Zusiocarpa showed similar results.
   The dry ridge-late snowbank environmental gradient is much sharper and extreme
in the mountains of western North America than it is in most of Europe. This is due,
partially, to the more regular incidence of late summer drought and the gradient is
particularly pronounced in the Sierra Nevada of California. Here, the sandy granitic
slopes and ridges are xeric environments in the long summer drought but the meadows
below snowbanks are usually green and moist throughout the summer (Klikoff, 1965a).
We (Mooney, Billings & Hillier, 1965) have found that in the Sierra Nevada specific
transpiration rates of alpine plants are closely related to the water supply of the habitat.
Plants from moist sites have higher transpiration rates and little control over transpira-
tion as compared with the more efficient water use of dry-site plants. Thus, water
availability in the dry Sierran summer has much to do with the sharp vegetational
patterning characteristic of the high Sierra.
   Water stress which develops in these plants along the dry meadow-wet meadow
gradient during the summer greatly affects photosynthesis Iates and thus productivity.
Klikoff (1965b) found that photosynthesis in Calamugrostis breweri (in the uniform
environment of a growth chamber) decreased to almost zero at water potentials below
 - 10 bars but photosynthesis of Carex exserta, characteristic of dry meadows, was
operating at almost maximum values at - 10bars and at 25 % of maximum at water
potentials as low as -20 bars.
   It is clear that in both alpine and high arctic areas, windswept ridges and windward
slopes are the principal stress environments, and that these are particularly severe in
winter. The cushion plant, prostrate evergreen shrub, and dry tussock life forms are
adapted as much to this winter drought stress as to that of summer. The summer
drought on such dry sites comes earlier than in the moist meadows both in alpine and
 arctic regions. S~rensen   (1941) has pointed out that summer drought may stop growth
in such plants in north-east Greenland while temperatures are still favourable for
 growth. Billings & Bliss (1959) noticed the same effect in the alpine vegetation of
Wyoming. Such cessation of growth due to late summer drought may be involved in
the hardening preceding onset of dormancy. For example, Holway & Ward (1965)
 have noted that, in Geum turbinatum, time of dormancy onset is closely related to soil
moisture with dormancy coming earlier in dry years.
                      The ecology qf arctic and alpine plants                       515
Annual cycle o growth
   Breaking of dormancy. Dormancy may be broken in arctic and alpine plants as early
as April or May, or as late as August or September. The keys to time of starting of
new growth are the melting of the snow-cover or, on snow-free areas, the increase of
soil and lower air temperatures to about oo C. As Ssrensen (1941) has pointed out,
dormancy may be broken and growth started in many arctic plants before the mean
temperature of the day arises to oo C. ; the plants make use of even the few hours of
positive temperatures around noon. Liquid water is another requirement for early
growth; some growth may take place even under the old snow providing water is
available on the soil surface or in the upper soil. A relatively long photoperiod appears
to be important in breaking the dormancy of perennating buds in some species
(Oxyriu digyna), but without positive temperatures there is no photoperiodic effect
(Billings, Godfrey & Hillier, 1965).
   Growth. Once dormancy is broken, growth of shoots is very fast. In spite of the low
temperatures, it is here that the perennial nature of the tundra herbaceous plant is
of decided advantage. The carbohydrates stored in roots and rhizomes are rapidly
translocated to the young shoots and leaves (Mooney & Billings, 1960). In the bright
light and low temperatures after snowmelt, these are not only incorporated into new
tissue but a considerable amount into anthocyanins, so that the young vigorous shoots
are often dark red. Such anthocyanins are effective in absorption of solar u.v., thus
preventing damage to tissues (Caldwell, 1968).
   Another common attribute of some fast-growing herbaceous stems of plants is the
lack of pith in many species. The resultant hollow stem is efficient in the saving of
materials, the rapidity of its growth, and the advantage of internal photosynthesis with
recycling of carbon (Billings & Godfrey, 1967). Such stems can grow several centi-
metres in height in a day and the high respiration rate of this early-season growth pro-
vides an additional supply of carbon dioxide which is utilized within the hollow stem,
where temperatures in sunlight may be as much as zoo C. above that of ambient air.
These early phases of shoot growth are thus somewhat independent of temperature in the
outside environment. After leaf expansion and development of a large supply of chloro-
phyll, however, further growth is rather closely tied to temperature until late in the
season, when drought, photoperiod, and carbohydrate accumulation become involved.
   Growth in the cold spells of arctic and alpine regions is fraught with several
problems in mineral nutrition. Nitrogen is often limiting in the cold soils of the Arctic
(Russell, 1 9 4 0 ~Dadykin, 1952) probably due to the low rates of bacterial activity
and also the relative lack of availability of certain other nutrients. Nitrogen build-up
in pioneer arctic and subarctic ecosystems is aided by bacterial nodules on legumes and
particularly on plants of Dryus (Lawrence, Schoenike, Quispel & Bond, 1967).
Nitrogen apparently is not so often a limiting factor in many alpine environments,
although Bliss (1966) found that dry-matter production on Mt Washington could be
increased significantly only by adding large amounts of nitrogen fertilizer. Cold
soils also appear to inhibit growth by suppressing the absorption of phosphorus
(Korovin, Sycheva & Bystrova, 1963).
                                   AND H. A. MOONEY
                      W. D. BILLINGS
   Carbohydrate and lipid storage and use. The general course of the carbohydrate cycle
has been described by Russell (194ob) for arctic plants and by Mooney & Billings (1960)
for alpine plants. Carbohydrates are stored in large amounts in roots and rhizomes
at the end of the growing season. During the winter, because of low temperatures and
dormancy, there is little use of these underground reserves. With the breaking of
dormancy in early summer, the stored carbohydrates are utilized in the rapid early
growth of the shoot and are depleted to a large degree. During this time, growth is SO
rapid that respiration far exceeds photosynthesis and food is used up faster than it
can be made (Hadley & Bliss, 1964). This situation is maintained until shoot growth
is 75-90y0 completed. Then, the respiration rate drops, shoot growth slows down,
and replenishment of carbohydrate reserves commences. By the onset of dormancy,
underground carbohydrate reserves are back to the highest level of the year.
   The carbohydrate cycle is modified and changed during periods or whole growing
seasons of cold, cloudy weather. Fonda & Bliss (1966) on Mt Washington found that
an early August cold period during 1963 reduced shoot carbohydrates in Carex
bigelowii from 22% to 13 yo and rhizome carbohydrates from 22% to 17’5%. How-
ever, the warmer, sunny days later in the month allowed replacement of the utilized
reserves. If there had been no more sunny weather, it is obvious that the plants would
have entered winter with less than optimum supplies. Local differences in microsite
also modify the carbohydrate cycle. Rochow (1964) found that root and shoot sugar
contents in Caltha leptosepala were highest in those plants released from snow-cover
late in June. Plants released in the middle of July had lower sugar levels and were
unable to increase these levels because of the short growing season remaining. There
appears, also, to be some tendency for carbohydrate content to decrease with increasing
elevation, as we (Mooney & Billings, 1965) found in roots of Calyptridium umbellatum
in California.
    The principal reserve foods in alpine and arctic plants are starches, sugars, and
lipids. In herbaceous plants these are mostly stored in roots, rhizomes, corms and
bulbs, since few leaves remain during the winter. However, in low evergreen shrubs
such as Diapensia lapponica and Ledum groenlandicum, Hadley & Bliss (1964) found
that old stems and old leaves provided the principal storage reservoir for lipids and
also for some carbohydrates. A similar situation occurs in the small evergreen alpine
shrubs of New Zealand, where Bliss & Mark (unpublished) found 15-20% lipid
 content in the leaves of Celmisia viscosa.
    There appears to be a close relationship between sugar content and cold resistance,
 although other factors such as environmental history are very important. Eagles (1967),
for instance, found that a Norwegian population of Dactylis glomerata had significantly
 higher concentrations of soluble carbohydrates (fructosans) at lower temperatures than
 did a population from Portugal at the same temperatures; fructosan content in sheaths
 was rapidly reduced by rising temperatures. Sugar content (as opposed to starch) is
often high at snowmelt and again just as dormancy sets in (Mooney & Billings, 1960).
 Parker (1963), in his recent review on cold resistance in woody plants, cites the almost
 universal correlation of increase in sugars with winter-hardening ; raffinose appears to
 be very important in this regard. Zavadskaya, Fel’dman & Kamentreva, (1964),
                      The ecology of arctic and alpine plants                     517
 working with Dactylis glomerata, Leucanthemum vulgare, Campanula persicifolia and
 Gagea lutea, say it may be concluded that in the species studied, high cold resistance
goes hand in hand with accumulation of raffinose'.
   Flowering. Since almost all alpine and arctic perennials have pre-formed flower
buds (S~rensen,    1941 and others), the flowering in a particular season depends to a
great extent on environmental conditions during the previous one or two summers.
The production of flowers from these buds depends upon the completion of the initial
stage of leaf and stem growth after snowmelt, upon temperature, and in some cases
(particularly in arctic species or ecotypes) upon the interaction between temperature
and photoperiod.
   In Oxyria digynu, we (1961) found that no populations flowered under a 12 hr.
photoperiod, populations from around 40' latitude flowered at 15 hr., but that arctic
populations would flower only in very long photoperiods or in continuous light. Alpine
populations flowered also under continuous light. Billings et al. (1965), working with
seventeen Oxyriu populations under four photoperiodic regimes in the laboratory,
found the same thing with more precision. The higher the latitude of origin, the
greater the number of hours of continuous light or the longer the photoperiod needed
for flowering. This is true not only for the elongation and flowering of the pre-formed
bud but for the production of the pre-formed flowering primordia themselves. Both
phenomena are tied to carbohydrate availability and neither flowering nor flower
primordia initiation take place until the carbohydrate level is sufficient for normal
survival. At short photoperiods, arctic populations of Oxyria seem to have difficulty
in meeting this carbohydrate requirement; this may be partially responsible for the
lack of initiation or aborting of inflorescences in such ecotypes when grown under
relatively short photoperiods (Billings et al. 1965).
   The photoperiodic effect in Oxyria is also tied to temperature; flowering occurs
more rapidly at higher temperatures than at low temperatures under the same photo-
period. Each latitudinal population of Oxyria has its own combination of temperature
and photoperiod ; alpine populations are better adapted to shorter photoperiods and
higher temperatures and arctic populations better fitted to long photoperiods at lower
temperatures. Mark (1965 b) found that flowering in snow tussock (Chionochloa rigida)
in New Zealand depends upon the right combinations of long photoperiod and mild
temperature minima. Since these relatively high minima do not occur every year,
flowering in this grass tends to be irregular.
   It has been suggested (S~rensen,   1941; Went, 1964) that some species of arctic or
alpine plants may have a chilling requirement. Clebsch (1960) found this requirement
in Trisetum spicatum, while Billings et al. (1965) found that although chilling was not
a requirement in Oxyria, it did greatly speed up flowering in that species. Chilling
requirements seem to be particularly important in bulbous plants such as Erythronium
grandiflorum (Caldwell, 1967).
   Pollination. Both arctic and alpine plants face pollination problems caused by low
temperatures which confine insect activities to a few weeks, and even then mainly to
the sunny daylight hours (Mani, 1962). The principal agents, of course, are insects,
birds, and wind. Schroeter (1926) cites some data of Raddes from the high mountain
5 18                  W. D. BILLINGS H. A. MOONEY
flora of the Caucasus which shows that 89% of the species are insect-pollinated and
I I % are wind-pollinated. Schroeter’s comparable figures from the Arctic (after
Aurivillius) show that 32-38 yo are wind-pollinated. A few species are pollinated by
humming-birds in some alpine regions (California: H. G. Baker, personal comm.)
but these are completely lacking from the Arctic.
   The principal kinds of pollinating insects in alpine locations are short-tongued
bees, bumblebees, flies, butterflies, and moths. Bumblebees are more common above
timberline in the Northern Hemisphere mountains than any other kinds of bees.
In many places, they are the only bees present; they fly only during the daylight hours
when air temperature is above about 10’C. (Mani, 1962; L. W. Macior, personal
comm.). Flies can and do work at lower temperatures and in dimmer light. As one
ascends above 4000 m. in the Himalaya, bee-pollinated flowers disappear completely
from the biota and one finds in their place a dominance of brightly coloured flowers
pollinated by Lepidoptera and Diptera (Mani, 1962). Muller (1881) found a similar
situation in the Alps.
    Long-tubed flowers in Northern Hemisphere alpine areas are usually bumblebee-
pollinated while small, flat flowers are mostly ‘fly flowers’ (L. W. Macior, personal
comm.). While some species of bumblebees (e.g. Bombusfrigidus) are common in parts
of the Arctic, Mosquin & Martin (1967) found that on Melville Island ( 7 5 O N.),
bumblebees visited only scented flowers of Astragalus alpinus, Petasites frigidus, and
perhaps Parrya arctica. They did not visit Pedicularis arctica, P. sudetica or Oxytropis
arctica, which were unscented, or any of thirty-six other species with scentless flowers.
Mosquin and Martin consider that Diptera are more important flower visitors in the
Arctic than bees.
    One of the striking differences between the alpine flowers of New Zealand and those
of the Northern Hemisphere high mountains is the absence of bright-coloured flowers
in alpine New Zealand and the prevalence of small, flat, white or yellow flowers.
Heine (1937) explains this, at least in part, by pointing out that there were no long-
tongued bees native to New Zealand, so that most insect pollination in the alpine zone
there is by small insects, flies, and short-tongued bees.
    Heterostyly and dioeciousness, while present in alpine locations of the middle-
latitudes, are less important in the Arctic, with heterostyly being rare or not known
there (Mosquin, 1966). In the most severe environments, self-pollination, apomixis,
or vivipary are often the rule and cross-pollination is relatively rare.
    Seedproduction. Seed production may or may not follow pollination depending upon
whether the weather remains favourable or not. At higher latitudes and altitudes,
years when any seed are produced become fewer, and even in relatively good years the
amount of viable seed may be low. Holway & Ward (1965), for example, noted that
several species of plants ( Trijolium parryi, Artemisia spithamea, Campanularotundifolia)
bloomed abundantly in the alpine regions of Colorado but failed to set seed. They
examined several hundred pods of the Trifolium and found only twenty-eight seeds.
Bliss (1956) noted that seed production decreases with an increase in severity of micro-
environments in the same general region so that late snowbankpopulations, for example,
of a given species may have poor seed-set while seed may be produced in abundance
                      The ecology of arctic and alpine plants                         519
by plants of the same species in more favourable sites nearby. There is a tendency
for seed weight to decrease with altitude in some species (Mark, 1965b).
   Many arctic and alpine plants, however, do produce seeds rather regularly and in
apparent abundance. Such seed, depending upon the species, ripens from mid- July
to September. It is usually soon dispersed, often by wind or by falling from the follicle
or capsule. In general, seed ripening is too late in most species for germination in that
growing season. I n fact, Porsild (1951)   points out that fruits of some arctic species do
not mature and dehisce until after snow arrives in the autumn and are dispersed with
the drifting snow of winter.
   Vegetative reproduction.As sexual seed production decreases and becomes unreliable
in the most severe environments, vegetative reproduction seems to increase. Apomixis
is particularly prevalent in tundra environments (Love, 1959). So, also, i the pro-
duction of bulbils, as in Polygonum viviparum and a number of grasses. Grass bulbils
frequently develop into young plants in the inflorescence ('vivipary ') ; these young
plants may even flower (Trisetum spicatum growing in controlled conditions, Clebsch,
1960). In nature, the little plants fall to the ground while vegetative, and may become
established. Such vivipary has often been deemed a taxonomic character (e.g. in Poa
alpigenu var. colpodea), but Clebsch found in several populations of normally non-
viviparous Trisetum spicatum that vivipary could be induced by manipulating the
experimental environment.
   The principal means of vegetative reproduction in both arctic and alpine plants is by
rhizomes. Layering is also important in some species, particularly in cushion plants
and prostrate shrubs. Intraspecific variation in rhizome production is sometimes
coupled to environmental conditions. For example, we (1961)           found that in North
America, arctic forms of Oxyria digyna have rhizomes and often reproduce by them
into clonal colonies. Western American Oxyrias south of about 60"latitude do not have
rhizomes and thus reproduce by seed. Rhizome production in Oxyria is geneticdly
controlled (Fig. 2).
   Onset of dormancy. Under natural conditions, tundra plants beyond the seedling
stage are seldom killed by the onset of winter weather. Some plants, such as Bruya
humilis in north-east Greenland, may be caught by such weather in flowering or
fruiting condition without any damage and resume growth and fruiting the following
year (Serrensen, 1941).  Normally, however, there is a period of shortening dayIength,
lowering temperatures, and often increasing drought which tend to 'harden' the plant
and induce a dormancy which prepares it for the oncoming winter. We (1961)           found
in all the American populations of Oxyria digyna with which we worked that a photo-
period of 12 hr. even at high temperatures induced the formation of a perennating bud
which carried the shoot primordia through experimentally maintained winter (freezing)
conditions. In further work, Billings et al. (1965)  found that there was some tendency
for arctic and subarctic Oxyrias to initiate perennating buds at photoperiods as long
as 15 hr. and to have well-developed buds by the time the photoperiod had been
dropped to 13 hr. Alpine populations from both America and Europe, however, did
not form perennating buds until the photoperiod was down to 12hr., at which day-
length the bud was formed rather quickly.
520                   W. D. BILLINGS H. A. MOONEY
   A somewhat similar photoperiodic effect on hardening itself was observed by Biebl
(1967) during some experiments in which he artificially shortened the summer photo-
period (early July) of some plants on Disko Island (West Greenland) to 8 hr. After
10 days of this treatment, plants of Salix glauca ssp. callicarpaea, Betula nana, Vacci-
nium uliginosum and Empetrum hemaphroditum had increased resistance to heat.
Additionally, the leaves of Betula nana showed an autumnal red colouring and were
significantly more resistant to a 24 hr. freezing treatment. A clue to such a photo-
periodic effect may be found in the work of Irving & Lanphear (1967). They showed
that cold hardiness could be induced in Acer negundo, Viburnumplicatum tomentosum
and Weigela jlorida by subjecting these woody plants of the temperate zone to short
days followed by low temperatures. T h e same hardening could be brought about in
long days by manually removing the leaves. These observations suggest the presence
of a hardiness inhibitor which originates in leaves exposed to long days.
   Shortened photoperiod and low temperatures are not the only factors involved in
the onset of dormancy in alpine plants, and perhaps in some arctic plants also. Both
Bliss (1956) and Billings & Bliss (1959) noted that cessation of growth and onset of
dormancy in August in the Rocky Mountain alpine zone was determined by factors
other than low temperature. I n the 1959 paper it was hypothesized that soil drought
limited growth and hastened dormancy. Holway &Ward (1965) found just this kind
of response in Geum turbinatum in noting that the time of dormancy up until mid-
September was controlled by the year-to-year variation in soil moisture. It seems
obvious that in a wet year some other factors must eventually control dormancy in this
species after mid-September ;most likely shortened photoperiod and low temperatures
provide the stimulus for such a late move into dormancy.

                            IV. PRIMARY PRODUCTIVITY
   Estimates of primary productivity and yield in alpine and arctic vegetation have
been assembled by Bliss (1962~).   Since then, Scott & Billings (1964) and Bliss (1966)
have published more detailed analyses of field standing crops and productivities in two
widely separated alpine regions. All values in this section are in grams dry weight
unless otherwise specified.
   One of the characteristic attributes of arctic and alpine tundras is the great pro-
portion of standing crop which lies underground. Aleksandrova (1958) found as much
as 810 g./m.2 of roots and rhizomes in an arctic tundra turf community. We found
(1960) 1060 g./m.2 beneath a community of Geum turbinatum in the alpine tundra of
Wyoming. I n a similar mesic site, Scott & Billings (1964) measured a maximum
underground standing crop of over 1100      g./m.2 while at the same time the above-
ground crop was about 250 g./m.2; over 80 % of the biomass was underground. Xeric
sites in the same vicinity had maximum underground biomass of less than 400 g./m.2
with above-ground crops of about 200 g./m.2. From these figures the relationship of
the moisture gradient to standing crop in a cold climate is obvious; the effect on the
underground crop seems to be greater than on the crop of shoots. T h e data presented
above are maximum values for their sites; average values are somewhat lower. Never-
                      The ecology of arctic and alpine plants                            521
theless, even larger underground standing crops have been measured in alpine areas;
 Bliss (1966)  found 3634 g.lm.2 in a sedge meadow on Mt Washington and Bliss &
Mark (1965)    measured living root biomass of 2120g./m.2under an alpine Celmisiu-Poa
herbfield in New Zealand. The smallest standing crops, as might be expected, are in the
most severe environments : colder, dryer, or those in the latest-melting snowbank
   Productivity may be based on an entire year, on the growing season, or on a single
 day. Since tundra is not productive at all for 70-907~ the year, annual figures are
low compared to most other terrestrial communities. Bliss (1962~) that annual
shoot productivities of tundra ecosystems are within the range 40-128g.. m.-2. year-l
but may be as low as 3 g.. m.-2. year-l in the high Arctic. Since these are for above-
ground material only, they should be multiplied by a factor of about 3 for total pro-
ductivity. Such figures would indicate that tundras on an annual basis are less
productive than any other terrestrial ecosystems except those of extreme deserts
(Rodin & Bazilovich, 1964).
   While annual productivities of tundras are quite low, daily productivities during
the short, cold growing seasons of 30-75 days are much more comparable to pro-
ductivities in other parts of the world. In fact, they usually exceed comparable rates
from deserts, grasslands, and on better sites even may be greater than those of culti-
vated crops (Scott & Billings, 1964).In terms of average shoot production alone during
the growing season, alpine communities range from about 0.5 to 5-0 g.. m.-2. day-l
(Bliss, 1966), if root productivity is included the rate may be as high as I I g., m.-2.
day-l on moist sites (Scott & Billings, 1964).Arctic productivity is generally somewhat
less than that in the better alpine sites although, except for those of Aleksandrova
(1958),  arctic data are rare. New Zealand alpine productivity is also lower than that of
Northern Hemisphere sites on a daily basis because of low daytime temperatures.
However, since the New Zealand alpine growing season is 2-3 times longer, the
annual productivity is about the same (L. C. Bliss, personal comm.).
   Up to the present time, almost all tundra productivity has been presented in terms
of dry weight. Better productivity estimates could be made by utilizing caloric values.
Bliss (1962~)   found that caloric values for alpine shrubs and herbs are significantly
higher than for many tropical and temperate plants. This could indicate that tundra
productivities are even higher relative to those in milder regions than the dry weight
figures imply. Highest caloric values are in the evergreen prostrate shrubs, reflecting
their higher lipid contents.
   A very practical question may be asked: 'How can productivity be increased in
tundra ecosystems?' The most obvious answer would be to increase the factors which
are limiting: temperature, soil nutrients, or water. We (Billings et ul. 1965)have some
controlled experimental evidence with Oxyriu digynu shoot yield to indicate that
productivity, in this species at least, is environmentally controlled in the field and
can be increased a great deal by increasing the temperature in which the plants are
growing. Under continuous light in a cold temperature regime (daily cycle 4 1 ' C.)
for 48 days, an alpine Oxyriu population from Niwot Ridge, Colorado, had an average
shoot productivity of 4.40 g.. m.-2. day-l, while one from Thule, Greenland, in the
   34                                                                    Biol. Rev. 43
                      W. D. BILLINGS H. A. MOONEY
same regime had 1-66g.. m.-2. day-l. These are quite comparable to field figures for
a similar growing season. Under a warm regime (16-27"C.), however, the rate for the
Niwot population rose to 21.20 g.. m.-2. day-l while that from Thule was 5-58 g.. m.-2.
day-l, both far above field figures for their respective regions. Even neglecting the
relatively ideal growing conditions, the genetic potential for higher productivity
appears to be present even in the dwarf Thule population from the high Arctic.
Because of root carbohydrate depletion, the daily rates were decreasing at the high
temperature at the end of the experiment. However, they were increasing at that time
under the low-temperature regime so that the Niwot population was producing shoot
dry weight at the rate of 6.28 g.. m.?. day-l. At an intermediate temperature regime
(7-18" C.), the rate for Niwot was holding steady at ca. 16 g. .m.-2. day-l, a very high
rate by field standards, and with carbohydrates apparently being maintained. T h e
high Arctic population, however, could hold its own only at the low temperatures.
   The addition of certain mineral nutrients will increase productivity in certain
locations. Bliss (1966) found under the cold field conditions of M t Washington that
addition of large amounts of nitrogen fertilizer would increase productivity. Similar
experiments in other alpine regions have often showed no increase in productivity,
however. Additions of phosphorus might also increase productivity in some tundra
   Since drought stress is a common phenomenon in some western American alpine
environments, the addition of irrigation water would probably increase productivity
i n such places. However, because of the postulated role of drought in the onset of
dormancy, the effect of late irrigation on survival should be studied before its value
can be estimated.
   T h e real problem in increasing tundra productivity is that most of the biomass
increase is underground and not available for harvest by domestic animals or by any
other means. Moreover, if one is to assume that most wild tundras are in a steady state
over a long period of years, any method which tends to deplete these underground
reserves is self-defeating since they are necessary to the survival of the producing
plants themselves.

                                    V. SUMMARY
   ' How are plants adapted to the low temperatures and other stresses of arctic and
alpine environments?' At present it is not possible to answer this question completely.
Much work remains to be done, particularly on low-temperature metabolism, frost
resistance, and the environmental cues and requirements for flowering, dormancy,
regrowth, and germination. However, in brief, we can say that plants are adapted to
these severe environments by employing combinations of the following general
characteristics :
   I . Life form: perennial herb, prostrate shrub, or lichen. Perennial herbs have
greatest part of biomass underground.
   2. Seed dormancy : generally controlled by environment ; seeds can remain dormant
for long periods of time at low temperatures since they require temperatures well
above freezing for germination.
                      The ecology of arctic and alpine plants                       523
   3. Seedling establishment: rare and very slow; it is often several years before a
seedling is safely established.
   4. Chlorophyll content: in both alpine and arctic ecosystems not greatly different
on a land-area basis from that in temperate herbaceous communities. Within a single
species there is more chlorophyll in leaves of arctic populations than in those of alpine
   5. Photosynthesis and respiration:
   (u) These are at high rates for only a few weeks when temperatures and light are
   ( 6 ) Optimum photosynthesis rates are at lower temperatures than for ordinary
plants; rates are both genetically and environmentally controlled with phenotypic
plasticity very marked.
   (c) Dark respiration is higher at all temperatures than for ordinary plants; rate is
both genetically and environmentally controlled, with phenotypic plasticity very
pronounced, i.e. low-temperature environment increases the rate at all temperatures.
   ( d ) Alpine plants have higher light-saturation values in photosynthesis than do
arctic or lowland plants ; light saturation closely tied to temperature.
   ( e ) There is some evidence that alpine plants can carry on photosynthesis at lower
carbon dioxide concentrations than can other plants.
   (f) Annual productivity is low, but daily productivity during growing season can be
as high as that of most temperate herbaceous vegetation. Productivity can be increased
by temperature, nutrients, or water.
   6 . Drought resistance: most drought stress in winter in exposed sites is due to
frozen soils and dry winds. It is met by decreased water potentials, higher concentra-
tions of soluble carbohydrates, and closed stomates. Little drought resistance in
snowbank plants. Alpine plants adapted to summer drought stress can carry on photo-
synthesis at low water potentials; alpine or arctic plants of moist sites cannot do this.
   7. Breaking of dormancy: controlled by mean temperatures near or above oo C., and
in some cases by photoperiod also.
   8. Growth: very rapid even at low positive temperatures. Respiration greatly
exceeds photosynthesis in early re-growth of perennials. Internal photosynthesis may
occur in hollow stems of larger plants during early growth. Nitrogen and phosphorus
often limiting in cold soil.
   9. Food storage: characteristic of all alpine and arctic plants except annuals.
Carbohydrates mostly stored underground in herbaceous perennials. Lipids in old
leaves and stems of prostrate evergreen shrubs. Depleted in early growth, and usually
restored after flowering.
   I 0. Winter survival : survival and frost resistance are excellent after hardening.
Cold resistance closely tied to content of soluble carbohydrates, particularly raffinose.
   I I. Flowering: flower buds are pre-formed the year before. Complete development
and anthesis dependent upon temperature of the flowering year and also, in some
cases, upon photoperiod.
   12. Pollination: mostly insect-pollinated in alpine regions and even in Arctic, but
to a lesser extent. Wind-pollination increasingly more important with increasing
524                                    AND H. A. MOONEY
                          W. D. BILLINGS
latitude. Diptera more important than bees in the Arctic and in the highest
   13. Seed production: opportunistic, and dependent upon temperature during
flowering period and latter half of growing season.
   14. Vegetative reproduction : by rhizomes, bulbils, or layering. More common and
important in Arctic than in alpine areas.
   15. Onset of dormancy: triggered by photoperiod, low temperatures, and drought.
Dormant plant extremely resistant to low temperatures.
  This review is an outgrowth of research supported by National Science Foundation Grants GB-1219
and G-5501, for which we are appreciative.

                                        VI. REFERENCES
ALEKSANDROVA, (1958). An attempt to measure the over- and underground mass of tundra plants.
             V. D.
 Bot. Zh. SSSR 43, 1748-61. (In Russian.)
AMEN, D. (1965). Seed dormancy in the alpine rush, Luzula spicata L.Ecology 46, 361-4.
AMEN, D. (1966). The extent and role of seed dormancy in alpine plants. Q. Rev. Biol. 41,271-81.
BESCHEL, E. (1961). Dating rock surfaces by lichen growth and its application to glaciology and
 physiography (lichenometry). In Geology of the Arctic, vol. 11, pp. 1044-1062. Ed. G. 0. Raasch.
BIEBL, (1942). Wirkung der UV-Strahlen auf die Plasmapermeabilitat. Protoplasma 37, 1-24.
BIEBL, ( I 967). Kurztag-Einflusse auf arktische Pflanzen wahrend der arktischen Langtage. Plantu 75,
BILLINGS, D.& BLISS, C. (1959). An alpine snowbank environment and its effect on vegetation,
          W.              L.
  plant development and productivity. Ecology 40, 388-97.
BILLINGS, D., CLEBSCH, E. C. & MOONEY, A. (1961). Effect'of low concentrations of carbon
          W.               E.                     H.
  dioxide on photosynthesis rates of two races of Oxyria. Science, N . Y . 133, 1834.
BILLINGS, D., CLEBSCH, E. C. & MOONEY, A. (1966). Photosynthesis and respiration rates of
                           E.                    H.
  Rocky Mountain alpine plants under field conditions. Am. Midl. Nut. 75,34-44.
          W.                    P.
BILLINGS, D. & GODFREY, J. (1967). Photosynthetic utilization of internal carbon dioxide by
  hollow-stemmed plants. Science, N . Y. 158, 121-3.
          W.                  P.
BILLINGS, D. & GODFREY, J. (1968). Acclimation effects on metabolic rates of arctic and alpine
  Oxyria populations subjected to temperature stress. Bull. ecol. SOC. Am. 49, 68-69 (abstr.).
BILLINGS, D., GODFREY, J. & HILLIER, D. (1965). Photoperiodic and temperature effects on
                             P.               R.
  growth, flowering, and dormancy of widely distributed populations of Oxyria. Bull. ecol. SOC. m . 46,
  189 (abstr.).
BILLINGS, D. & MARK, F. (1961). Interactions between alpine tundra vegetation and patterned
           W.             A.
  ground in the mountains of southern New Zealand. Ecology 42, 18-31.
           W.                H.
BILLINGS, D. & MOONEY, A. (1959). An apparent frost hummock-sorted polygon cycle in the
  alpine tundra of Wyoming. Ecology 40,16-20.
BJURKMAN,. 0 (1967). Photosynthetic inhibition by oxygen in higher plants. Yb. Carnegie Inst. Wash. 65,
BJBRKMAN,0. & HOLMGREN, (1958). Preliminary observations on anthocyanins and other flavonoid
  compounds and respiration rates in different ecotypes of Solidago virgaurea. Physiologia PI. 11, 154-7.
BJ~RKMAN, HOLMGREN, (1961). Studies of climatic ecotypes of higher plants. Leaf respiration
            0&.                P.
  in different populations of Solidago virgaurea. K . LantbrH6gsk. Annlr. 27,297-304.
BJORKMAN, & HOLMGREN, (1963). Adaptability of the photosynthetic apparatus to light intensity
            0 .                P.
  in ecotypes from exposed and shaded habitats. Physiologia PI. 16,889-914.
BLISS,L. C. (1956). A comparison of plant development in microenvironments of arctic and alpine
  tundras. Ecol. Monogr. 26,303-37.
BLISS, C. (1958). Seed germination in arctic and alpine species. Arctic IT, 180-8.
BLISS, C. ( 1 9 6 2 ~ )Net primary production of tundra ecosystems. In D e Stoflproduktion der Pfianzen-
      L.               .                                                i
  decke, pp. 3 5-46. Ed. H. Lieth. Stuttgart. 155 pp.
BLISS, C. (1962b). Adaptations of arctic and alpine plants to environmental conditions. Arctic 15,
  1 17-44.
BLISS, C. ( 1 9 6 2 ~ ) Caloric and lipid content in alpine tundra plants. Ecology 43,753-7.
                          The ecology of arctic and alpine plants                                  525
BLISS, C. (1963). Alpine plant communities of the Presidential Range, New Hampshire. Ecology 4,
BLISS,L. C. (1966). Plant productivity in alpine microenvironments on M t Washington, New Hamp-
  shire. Ecol. Monogr. 36, 125-55.
BLISS,L. C. & MARK, F. (1965). Alpine microenvironments and plant productivity on Rock and
  Pillar Range, New Zealand. Bull. ecol. SOC. m . 46, I I I (abstr.).
BOCHER, W. (1963). Phytogeography of Middle West Greenland. Meddr. Grmland 148(3), 1-289.
BONDE, K. ( 1 9 6 5 ~ ) Studies on the germination of seeds of Colorado alpine plants. Univ. Colo. Stud.
         E.                 .
  Ser. Biol. 14,16 pp.
BONDE, K. (1965b). Further studies on the germination of seeds of Colorado alpine plants. Univ. Colo.
  Stud. Ser. Biol. 18,3 0 pp.
BOYSEN-JENSEN,       P. (1932). Die Stoflproduktion der P’anzen. Jena. 108 pp.
BRAUN-BLANQUET,        J. (1948). L a Vegetation alpine des PyrenLes orientales. Barcelona. 306 pp.
BRAUN-BLANQUET, JENNY, (1926). Vegetationsentwicklung und Bodenbilding in der alpinen
                       J. &
  Stufe der Zentralalpen. Neue Denkschr. schweiz. naturf. Ges. 63,181-349.
BRAY, R. (1962). The primary productivity of vegetation in central Minnesota, U.S.A., and its
  relationship to chlorophyll content and albedo. In Die Stoffproduktion der P’anzendecke, pp. 102-16.
  Ed. H. Lieth. Stuttgart. 155 pp.
BRITTON, E. (1957). Vegetation of the arctic tundra. Oregon State Univ. 18th Biol. Colloquium
  ‘Arctic Biology’, pp. 26-61.
BROCKMANN-                 H.
                 JEROSCH, (1919). Baumgrenze und Klimacharakter. Schweiz. Beitr. geobot. Landesaufn.
  6, 1-255.
BROWN, & JOHNSON, L. (1965). Pedo-ecological investigations, Barrow, Alaska. Tech. Rep. US.
           J.                 P.
  Army Materiel Command, Cold Regions Res. and Engin. Lab., Hanover, New Hampshire, no. 159. 3 2 pp.
CALDWELL, L. (1967). Erythronium: Comparative phenology of alpine and deciduous forest species in
  relation to environment. M.A. thesis, Duke University. 72 pp.
CALDWELL, M. (1968). Solar ultraviolet radiation as an ecological factor for alpine plants. Ecol.
  Monogr. 38 (3), (in the Press).
CAPPELLETTI, (1929-30). La flora alpina ed i suoi adattamenti alle radiazioni ultra-violette del sole.
  Atti Zst. veneto Sci. 89 (2), 291-321.
CARTELLIERI, (1940). Uber Transpiration und Kohlensaure-assimilation an einem hochalpinen
  Standort. Sber. Akad. Wiss. Wien. Math.-Nut. Klasse 149,95-143.
CLARK, M. & MARR, W. (1966). Insolation and other environmental factors during 1965 in three
         J.                   J.
  ecological regions in the Front Range, Colorado. Final Rep. Contract DAqg-ogz-AR046. U.S. Dept.
  Def., Wash., D.C.
CLAUSEN, KECK,D. D. & HIESEY,W. M. (1940). Experimental studies on the nature of species.
  I. Effect of varied environments on western North American plants. Carnegie Inst. Wash. Pub. no. 520.
  452 PP.
CLAUSEN, KECK,D. D. & HIESEY, M. (1948). Experimental studies on the nature of species
             J.,                            W.
  111. Environmental responses of climatic races of Achillea. Carnegie Instn Wash. Pub. 581. 129 pp.
CLEBSCH, E. C. (1960). Comparative morphological and physiological variation in arctic and alpine
  populations of Trisetum spicatum. Ph.D. dissertation, Duke University 241 pp.
COSTIN, B. (1967). Alpine ecosystems of the Australasian Region. In Arctic and alpine environments,
  pp. 55-87. Ed. H. E. Wright, Jr. and W. H. Osburn. Bloomington, Indiana. 308 pp.
DADYKIN, P. (1952). Osobennosti povedeniia rastenii na Kholodnykh pochvakh. Akad. Nauk S S S R ,
  Znst. Merzlotovedeniia, Moscow. 279 pp.
DAHL,E. (1956). Rondane: Mountain vegetation in south Norway and its relation to the environment.
  Skr. norske Vidensk-Akad. (Mat. naturv. Kl.), 1956 (3). 374 pp.
DAHL, (1963). On the heat exchange of a wet vegetation surface and the ecology of Koenigia islandica.
  Oikos 14, 190-211.
DAUBENMIRE, F. (1941). Some ecological features of the subterranean organs of alpine plants.
  Ecology 22, 370-8.
DAUBENMIRE,      R. (1954) Alpine timberlines in the Americas and their interpretation. Butler Univ. bot.
  Stud. 11, 119-36.
EAGLES, F. (1967). Variation in the soluble carbohydrate content of climatic races of Dactylisglomemta
  (cocksfoot) at different temperatures. Ann. Bot. 31,645-5 I.
FISHER, J. F. (1965). The alpine Ranunculi of New Zealand. N . Z . Dept. Sci. Zndust. Res. Bull. 165,
  192 PP.
FONDA, W.& BLISS,L. C. (1966). Annual carbohydrate cycle of alpine plants on Mt Washington,
  New Hampshire. Bull. Torrey bot. Club 93, 268-77.
GATES, M. & JANKE, (1966). The energy environment of the alpine tundra. Oecologia PI. I, 39-62.
        D.                   R.
                         W. D. BILLINGS H. A. MOONEY
GESSNER, ( I 960). Die photosynthetischen Leistungen von Pflanzen besonderer Standorte. ( d ) Die
  Assimilationsbedingungen im tropischen Regenwald. Handb. PfEphysiol. 5 (z),492-505.
GJAEREVOLL, (1956). The plant communities of the Scandinavian alpine snow-beds. K . Norske
  Vidensk. Selsk. Sky. 1956 ( I ) , 1-405.
             T.                 L.
GLAGOLEV, A. & FILOPOV, A. (1965). Characteristics of photosynthesis of plants in conditions of
  high altitudes of the Pamir Range. Problems o Botany 7, 120-32. Akademia Nauk, U.S.S.R. (in
GOOD, (1964). The geography of the flowering plants, 3rd ed. New York. 518 pp.
GREENE, W. et al. (1967). Terrestrial life of Antarctica. Antarctic M a p Folio Ser. fo. 5. Am. Geogr.
  Soc., N . Y . 24 pp. I I pl.
GRIGGS, F. (1956). Competition and succession on a Rocky Mountain fellfield. Ecology 37, 8-20.
HADLEY, B. & BLISS, C. (1964). Energy relationships of alpine plants on Mt Washington, New
  Hampshire. Ecol. Monogr. 34,331-57.
HANSON, C. (1953). Vegetation types in northwestern Alaska and comparisons with communities in
  other arctic regions. Ecology 34, I I 1-40.
HEDBERG, (1964). Features of afroalpine plant ecology. Acta Phytogeogr. suec. 49, 1-144.
HEINE, M. (1937). Observations on the pollination of New Zealand flowering plants. Trans. R . Sac.
  N.Z. 67, 133-48.
HENRICI,   M. (1918). Chlorophyllgehalt und Kohlensaure-Assimilation bei Alpen- und Ebenen-
  Pflanzen. Verh. nuturf, Ges. Base1 30,43-136.
HODGSON, J. (1966). Floral initiation in Alaskan Gramineae. Bat. Gaz. 127, 64-70.
HOLM, (1922). Contributions to the morphology, synonymy, and geographical distribution of arctic
  plants. Rep. Can. Arctic Exped. 5, (B), 1-139.
HOLMEN, (1957). The vascular plants of Peary Land, North Greenland. Meddr Grenland 124 (9),
HOLTOM, & GREENE, W. (1967). The growth and reproduction of antarctic flowering plants. Phil.
           A.             S.
  Trans. R. Sac. B 252, 323-37.
HOLWAY, G. & WARD, T. (1965). Phenology of alpine plants in northern Colorado. Ecology 46,
           J.               R.
HOWELL, T. (1951). The arctic-alpine flora of three peaks in the Sierra Nevada. Leafl. west. Bot. 6,
HUSTICH, (1948). The Scotch pine in northernmost Finland and its dependence on the climate in the
  last decades. Acta bat. fenn. 42, 1-75.
                                F. 0.
IRVING, M. & LANPHEAR, (1967). The long day leaf as a source of cold hardiness inhibitors.
  PI. Physiol. 4,   1384-8.
JENNY-LIPS, (1948). Vegetation der schweizer Alpen. Zurich. 240 p.
JOHNSON, W., VIERECK, A., JOHNSON, E. & MELCHIOR, (1966). Vegetation and flora. In
            A.                 L.             R.                  H.
  Environment of the Cape Thompson Region, Alaska, ch. 14, pp. 277-354. Ed. N. J. Wilimovsky and
  J. N. Wolfe. U.S. Atomic Energy Comm., Wash., D.C. 1250 pp.
JOHNSON, L. & BILLINGS, D. (1962). The alpine vegetation of the Beartooth Plateau in relation to
           P.                  W.
  cryopedogenic processes and patterns. Ecol. Monogr. 32, 105-35.
JUHREN, HIESEY, M. &WENT,F. W. ( I 953). Germinations and early growth of grasses in controlled
         M.,          W.
  conditions. Ecology 34,288-300.
KLIKOFF, G. (1965a). Microenvironmental influence on vegetational pattern near timberline in the
  central Sierra Nevada. Ecol. Monogr. 35, 187-21 I.
KLIKOFF, G. (1965b). Photosynthetic response to temperature and moisture stress of three timber-
  line meadow species. Ecology 46,516-17.
KLIKOFF, G. (1966). Temperature dependence of the oxidative rates of mitochondria in Danthonia
  intermedia, Pentstemon davidsonii and Sitanion hystrix. Nature, Lond. 212 (5061),529-30.
KOROVIN, I., SYCHEVA, F. & BYSTROVA,A. (1963). The effect of soil temperature on the amount
            A.              Z.
  of various forms of phosphorus in plants. Soviet Pl. Physiol. 10,109-12.
LANDOLT, (1967). Gebirgs- und Tieflandsippen von Blutenpflanzen im Bereich der Schweizer Alpen.
  Bat. Jb. 86,463-80.
LANGE, L. (1965). Der C0,-Gaswechsel von Flechten bei tiefen Temperaturen. Planta 64,
         0 L.
LANGE, . (1966). C0,-Gaswechsel der Flechte Cladonia alcicornis nach langfristigem Aufenthalt bei
  tiefen Temperaturen. Flora, Jena 156,500-2.
LARCHER, (1963). Zur spatwinterlichen Erschwerung der Wasserbilanz von Holzpflanzen au der
  Waldgrenze. Ber. naturw.-med. Ver. Innsbruck 53, 125-37.
LARSEN, (1964). The role of physiology and environment in the distribution of arctic plants. Univ.
         J. A.
   Wis. Dept. Meteorol. Tech. Rep. 16,70pp.
                           The ecology of arctic and alpine plants                                    527
LARSEN, A. (1965). The vegetation of the Ennadai Lake area N.W.T. : Studies in subarctic and arctic
   bioclimatology. Ecol. Monogr. 35, 37-59.
LAWRENCE, B., SCHOENIKE, E., QuisPEL, A. & BOND, (1967). T h e role of Dryas drummondii in
                D.                  R.                          G.
   vegetation development following ice recession at Glacier Bay, Alaska, with special reference to its
  nitrogen fixation by root nodules. J . Ecol. 55, 793-813.
LINDSAY, H. (1967). The effect of environmental factors on the leaf water balance of conifers. M.Sci.
  thesis, Univ. Wyo., Laramie. 60 pp.
LONA, (1963). I1 nanismo di molte piante alpine e essenzialmente genetico. G. Bot. ital. 70, 560-564.
LONGTON, E. (1967). Vegetation in the maritime Antarctic. Phil Trans. R. SOC. B 252,
L ~ v E , (1959). Origin of the arctic flora. In Problems of the Pleistocene epoch and Arctic area, ed. G. R.
  Lowther. McGill Univ. Mus. Publs no. I .
MANI,M. S. (1962). Introduction to high altitude entomology. Insect life above the timber-line in the
  north-west Himalaya. London. 302 pp.
MARK, F. ( 1 9 6 5 ~ )Central Otago: vegetation and mountain climate. N.Z. Geogr. SOC.,
         A.             .                                                                       Misc. Ser. 5,
  69-9 I .
MARK, F. (1965b). Flowering, seeding, and seedling establishment of narrow-leaved snow tussock,
   Chionochloa rigida. N . Z . J . Bat. 3, 180-93.
MARR, W. (1961).Ecosystems of the east slope of the Front Range in Colorado. Univ. Colo. Stud., Ser.
  Biol. 8, 134 pp.
MARR, W. (1966).Atmospheric factors of Front Range Colorado, regional ecosystems. Bull. ecol. SOL.
  Am. 47, 198 (abstr.).
MATTICK, (1950). Die Flechten als ausserste Vorposten des Lebens im Gebirge. Mont. e Uomini 2,
          A.                                A. (1963). The germination of seeds. New York. 236 pp.
MCVEAN, N. & RATCLIFFE, A. (1962). Plant communities of the Scottish Highlands. Monogr.
              D.                     D.
  Nature Conserv. no. I. 445 pp.
MCWILLiAM, J. R. & NAYLOR, W. (1967). Temperature and plant adaptation. I. Interaction of
  temperature and light in the synthesis of chlorophyll in corn. PI. Physiol. 4,      1711-15.
MONTFORT, ( I 950). Photochemische Wirkungen des Hohenklimas auf die Chloroplasten photolabiler
  Pflanzen im Mittel-und Hochgebirge. Z. Naturf. 5 B, 221-6.
MOONEY, A. (1963). Physiological ecology of coastal, subalpine, and alpine populations of Polygonum
  bistortoides. Ecology 4,812-16.
            H.                   W.
MOONEY, A. & BILLINGS, D. (1960). The annual carbohydrate cycle of alpine plants as related to
  growth. Am. J . Bot. 47, 594-8.
MOONEY, A. & BILLINGS, D. (1961). Comparative physiological ecology of arctic and alpine
              H.                   W.
  populations of Oxyria dignya. Ecol. Monogr. 31,1-29.
MOONEY, A. & BILLINGS, D. (1965). Effects of altitude on carbohydrate content of mountain
             H.                    W.
  plants. Ecology 46, 750-1.
            H.                 W.                  R.
MOONEY, A., BILLINGS, D. & HILLIER, D. (1965). Transpiration rates of alpine plants in the
  Sierra Nevada of California. Am . Midl. Nut. 74, 374-86.
            H.                  A.
MOONEY, A. & JOHNSON, W. (1965). Comparative physiological ecology of an arctic and an alpine
  population of Thalictrum alpinum L. Ecology 46, 721-7.
MOONEY, A., STRAIN, R. & WEST,M. (1966). Photosynthetic efficiency at reduced carbon dioxide
            H.             B.
  tensions. Ecology 47, 490-1.
MOONEY, A. & WEST,M. (1964). Photosynthetic acclimation of plants of diverse origin. Am.J. Bat.
MOONEY, A., WRIGHT, D. & STRAIN, R. (1964). The gas exchange capacity of plants in relation
            H.               R.                 B.
  to vegetation zonation in the White Mountains of California. Am. Midl. Nut. 72, 281-97.
MOSER, (1965). Temperatur- und Lichtabhangigkeit der Photosynthese sowie Frost- und Hitzeresistenz
  der Blatter von drei Hochgebiagspflanzm (Ranunculus glacialis, Geum reptans, Oxyria digyna).
  Doctoral diss., Univ. Innsbruck.
MOSER, (1967). Einblicke in das Leben von Nivalpflanzen. Jb. Ver. Schutze Alpenpfl. u.-Tiere 32,
  1-1 I.
MOSQUIN, (1966). Reproductive specialization as a factor in the evolution of the Canadian flora.
  In The evolution of Canada’s flora, pp. 43-65. Eds. R. L. Taylor and T. A. Ludwig. Toronto. 137 pp.
MOSQuiN, T. & MARTIN, E. H. (1967). Observations on the pollination biology of plants on Melville
  Island, N.W.T., Canada. Can. Fld Nut. 81, 201-5.
MULLER, (1881). Die Alpenblumen, ihre Befruchtung durch Insekten und ihre Anpassung an dieselbe.
PARKER, (1963). Cold resistance in woody plants. Bot. Rev. 29, 123-201.
                                       AND H. A. MOONEY
                          W. D. BILLINGS
PELTON, (1956).A study of seed dormancy in eighteen species of high altitude Colorado plants. Butler
  Univ. bot. Stud. 13,74-84.
PISEK, (1960). Die photosynthetischen Leistungen von Pflanzen besonderer Standorte. ( a ) Pflanzen
  der Arktis und des Hochgebirges. Handb. PflPhysiol. 5 ( ) 375-414.
PISEK,A. & KNAPP,H. (1959).Zur Kenntnis der Respirationsintensitat von Blattern verschiedener
  Blutenpflanzen. Ber. dt. bat. Ges. 72, 287-94.
       A.,             W.                      R. (1967). Kardinale Temperaturbereiche der Photosynthese
  und Grenztemperaturen des Lebens der Blatter verschiedener Spermatophyten. I. Temperatur-
  minimum der Nettoassimilation, Gefrier-und Frostschadensbereiche der Blatter. Flora, Jena B 157,
POLUNIN, (1948).Botany of the Canadian Eastern Arctic. 111. Vegetation and ecology. Bull. natn.
  Mus. Can. 104,Biol. Ser. no. 32. 304 pp.
POLUNIN, (195) The real Arctic: suggestions for its delineation, subdivisions, and characterization.
            N.      I.
  J . Ecology 39, 308-15.
POLUNIN, (1959).Circumpolar arctic flora. Oxford. 514pp.
PORSILD, E. (1951).       Plant life in the Arctic. Can. Geogr. J. 42, 120-45.
           A.      HARINGTON, R. & MULLIGAN, A. (1967).
                                C.                   G.            Lupinus arcticus Wats. grown from seeds
  of Pleistocene age. Science, N . Y . 158,I 13-14.
RAUH, (1940).Die Wuchsformen der Polsterpflanzen. Bot. Arch. 40, 289-462.
RAUP,H. M. (1965).The structure and development of turf hummocks in the Mesters Vig District,
  Northeast Greenland. Meddr. Grrnnland 166 (3), 1-1 12.
          H.                    H.          Obere Grenzen von Flora und Vegetation in der Nivalstufe der
  zentralen dtztaler Alpen (Tirol). Vegetatio 8,93-129.
ROCHOW, F.(1964).Seasonal caloric and carbohydrate content of Caltha leptosepala, an alpine snowbank
  plant. M.A. thesis, Duke University, 77 pp.
ROCHOW, F. (1967).The ecology of Thlaspi alpestre in the Central Rocky Mountains along altitudinal
  gradients. Ph.D. diss., Duke University, 265 pp.
RODIN, E. & BAZILEVICH, N. I. (1964).         Biological productivity of the main types of vegetation in the
  Northern Hemisphere of the Old World. Dokl. Akad. Nauk S S S R 157 (I), 215-18. (In Russian.)
RUDOLPH, D. (1967). The distribution and environmental relationships of antarctic lichens. Bull. ecol.
  Sac. Am. 48, 70 (abstr.).
RUSSELL, S. (1940~).       Physiological and ecological studies on an arctic vegetation. 11. The develop-
  ment of vegetation in relation to nitrogen supply and soil micro-organisms on Jan Mayen Island.
  J . Ecology 28,269-88.
RUSSELL, S. (194ob).Physiological and ecological studies on an arctic vegetation. 111. Observations
   of carbon assimilation, carbohydrate storage and stomata1movement in relation to the growth of plants
  on Jan Mayen Island. J . Ecology 28, 289-309.
SAVILE, B. 0. (1961). The botany of the northwestern Queen Elizabeth Islands. Can. J. Bot. 39,
SAVILE, B. 0. (1964).General ecology and vascular plants of the Hazen Camp area. Arctic 17,
SAYERS, L. & WARD, T. (1966).Germination responses in alpine species. Bot. Gaz. 127,11-16.
          R.                R.
SCHOLANDER, FLAGG, HOCK, J. & IRVING, (1953).Studies on the physiology of frozen
                 P. F.,        W.,          R.               L.
  plants and animals in the Arctic. J . Cell. Comp. Physiol. 42,Suppl. I, 1-56.
SCHROETER, (1926).
                C.        Das Pflanzenleben der Alpen. Zurich. 1288pp.
SCOTT, &BILLINGS, D. (1964). Effects of environmental factors on standing crop and productivity
         D.               W.
  of an alpine tundra. Ecol. Monogr. 34,243-70.
             A.            K. (1940). Uber die Blettpigmente der Alpenpflanzen. Bot. Arch. 40,
                                     S. V.,          S.
             A.                                                         V.                     B.
   T h e problem of ultraviolet irradiation of plants in space phytophysiology. [Ul’trafioletovoeobluchenie
   rastenii kak problema kosmicheskoi fitofiziologii.] Prob. Kosmi. Biol. 4,474-86. (In Russian.)
SQRENSEN, (1941).        Temperature relations and phenology of the northeast Greenland floweringplants.
  Meddr. Grsnland 125 (9), 1-305.
 SOYRINKI, (1938,       1939).Studien uber die generative und vegetative Vermehrung der Samen-pflanzen
   in der alpinen Vegetation Petsamo-Lappland. I, 11. Suomal. elain-ja kasvit. Seur. uan elain. Julk. 11,
   1-323; 14, 1-405.
STOCKER, (I 963). Das dreidimensionale Schema der Vegetationsverteilung auf der Erde. Bey. dt.
   bot. Ges. 76, 168-78.
 STRAIN, R. (1966). The effect of a late spring frost on the radial growth of variant quaking aspen
   biotypes. Forest Sci. 12,334-7.
                          The ecology of arctic and alpine plants                                   529
SUKACHEV, (ed.) (1965). Studies on the flora and vegetation of high-mountain areas. Transl. from
          V. N.
 Russian of Problemy Bot. 5. Israel Prog. Sci. Transl. Jerusalem. 293 pp.
SWAN, W. (1967). Alpine and aeolian regions of the world. In Arctic and alpine environments,pp. 29-
 54. Ed. €1. E. Wright Jr and W. H. Osborn. Bloomington, Ind. 308 pp.
TAYLOR, W. (1955). The flora, vegetation and soils of Macquarie Island. A.N.A.R.E. Rep. B, 2,
TIESZEN, & JOHNSON, L. (1967). The pigment structure of some arctic tundra communities.
           L. L.
  Bull. ecol. SOC. Am. 48,5 5 (abstr.).
TIKHOMIROV,A. (1963). Contributions to the biology of arctic plants [in Russian]. Dokl. Akud. Nauk
  U .S .S .R. 154 pp.
TRANQUILLINI,    W. (1963). Climate and water relations of plants in the sub-alpine region. I n The Water
  Relutions of Plants, pp. 153-67. Ed. A. J. Rutter. Dorking, England.
TRANQUILLINI, The physiology of plants at high altitudes. Ann. Rew. PI. Physiol. 15,345-62.
                 W. (1964).
TROLL, (1957). Tropical mountain vegetation. Proc. 9th Pac. Sci. Congr. 20, 37-46.
TROLL, (1958). Zur Physiognomik der Tropengewachse. Jber. Ges. Freund. F6rder. rhein. Friedrich-
  Wilhelms Univ. 75 pp.
TURESSON, (1931). The geographical distribution of the alpine ecotype of some Eurasiatic plants.
  Heriditas 15,329-46.
TURNER, (1958). u b e r das Licht- und Strahlungsklima einiger Hanglage der &taler Alpen bei
  Obergurgl und seine Auswirkung auf das Mikrolima und auf die Vegetation. Arch. Met. Geophys.
  Bioklim. 8 B , 273-325.
VICKERY, K., JR. (1967). Ranges of temperature tolerance for germination of Mimulus seeds from
  diverse populations. Ecology 48,647-51.
WAGER, G. (1938). Growth and survival of plants in the Arctic. J. Ecology 26, 390-410.
WARD, T. & DIMITRI, J. (1966). Alpine tundra on Mt Catedral in the southern Andes. N.Z.
        R.                 M.
 J . Bat. 4, 42-56.
WARDLE, (1964). Facets of the distribution of forest vegetation in New Zealand. N . Z . J. Bot. 2,
  35 2-66.
WARREN             J.
         WILSON, (1960). Observations on net assimilation rates in arctic environments. Ann. Bot. 24,
WARREN   WILSON, (1966). An analysis of plant growth and its control in arctic environments. Ann. Bot.
  30, 383-402.
WATSON, E. (1959). Climate of Alaska. U S . Wea. Bur., Climatography of U.S. no. 60-49. 24 pp.
WEBERBAUER,    A. (1911). Die PfEanzenwelt der peruanischen Anden. Vol. 12. Die Vegetation der Erde, ed.
  A. Engler and 0. Drude. Leipzig. 352 pp.
WEBSTER, . L. (1961). The altitudinal limits of vascular plants. Ecology 42, 587-90.
WENT,F. W. (1953). Annual plants at high altitudes in the Sierra Nevada, California. Mudrono 12,
WENT,F. W. (1964). Growing conditions of alpine plants. IsraelJ. Bot. 13,82-92.
WILriwovsKY, N.J. & WOLFE, N. (eds.) (1966). Environment of the Cape Thompson region, Alaska.
  U.S. Atomic Energy Comm., Wash., D.C. 1250 pp.
WINKLER, & MOSER, (1967). Die Vegetationzeit in zentralalpinen Lagen Tirols in Abhangigkeit
            E.            W.
  von den Temperatur-und Niedersdlagsverhaltnissen. Ver@. Mus. Ferdinandeum Innsbr. 47,I 21-47.
                                                       I. E.
ZAVADSKAYA, FEL’DMAN, L. & KAMENTSEVA, (1964). Carbohydrate content and cold resist-
               I. G.,           N.
  ance of higher plant cells. (Transl. from Russian.) Dokl. Akad. Nauk SSSR 157 ( ) 995-7.
ZUBKOF, I. (1935). Duration of the vegetation period on the northern island of Novaya Zemlya.
  Ayctica 3. (Arctic Inst. U.S.S.R.) (In Russian with English summary.) Cited (p. 267) in Serensen