THE VEGETATION OF A DESERT MOUNTAIN
RANGE AS CONDITIONED BY
• CLIMATIC FACTORS

BY

FORREST SHREVE

WASHINGTON, D. C.

PUBLISHED BY THE CARNEGIE INSTITUTION OF WASHINGTON

1915

^

CARNEGIE INSTITUTION OF WASHINGTON
PUBLICATION No. 217

o

Copies of this Book
were first issued

OCT281915

PRESS OF J. B. LIPPINCOTT COMPANY
PHILADELPHIA

CONTENTS.

PAGE.

Introduction 5

Geography and Topography of the Santa Catalina Mountains 6

Vegetation of the Santa Catalina Mountains 11

The Desert Region 15

The Upper Bajadas 15

The Desert Arroyos and Canons 19

The Lower Desert Slopes 22

The Upper Desert Slopes 23

The Encinal Region 24

The Lower Encinal 25

The Upper Encinal 27

The Forest Region 29

The Pine Forest 31

The Fir Forest 33

Flora of the Santa Catalina Mountains 36

Phytogeographic Relationships of the Flora 36

The Desert Flora 36

The Encinal Flora 38

The Forest Flora 39

List of Characteristic Species 41

Climate of the Santa Catalina Mountains 46

Rainfall 48

Seasonal Distribution of Rainfall 48

Altitudinal Increase of Rainfall 51

Soil Moisture 59

Evaporation 63

Humidity 67

Temperature 69

Length of Frostless Season 70

Normal Altitudinal Temperature Gradient 75

The Absolute Minimum of Winter 79

Departures from the Normal Temperature Gradient due to Cold-air Drainage 82

Soil Temperature 86

Correlation of Vegetation and Climate in the Santa Catalina Mountains 88

The Normal Altitudinal Gradient of Vegetation 88

The Vertical Distribution of Individual Species 89

Physical Factors Involved in the Determination of the Normal Altitudinal Gradient

of Vegetation 91

Moisture Factors 92

Temperature Factors 94

The Role of Topographic Features in Determining Departures from the Normal

Altitudinal Gradient of Vegetation 97

The Role of Slope Exposure 97

The Role of Streams and Flood-plains 104

The Role of Topographic Relief 107

General Conclusions 109

3

THE VEGETATION OF A DESERT MOUNTAIN RANGE AS
CONDITIONED BY CLIMATIC FACTORS.

INTRODUCTION.

The southern half of the state of Arizona may be briefly character-
ized as a relatively level plain studded with numerous hills and moun-
tains. The plain rises from elevations of a few hundred feet along the
Colorado River to as much as 4,500 and 5,000 feet near the New Mexi-
can boundary. The lower elevations follow the Gila, Salt, San Pedro,
and other rivers, while the higher plains surround the loftier mountains
of the southeastern portion of the State. Between the Colorado River
and Tucson there are no mountains of commanding elevation, and the
area occupied by the scattered volcanic peaks and ranges is not more
than one-tenth of the total area of the region. To the eastward of
Tucson, however, a much greater percentage of the total area is occu-
pied by mountain ranges, a score of which reach elevations of over 8,000
feet. The general topographic configuration of the region has remained
unchanged throughout a long period of geological time, and the moun-
tains and hills have been subjected to prolonged erosion, the products
of which have served to build up the shelving plains which form the
intervening valleys.

Those portions of southern Arizona which lie below 4,000 feet are
covered with a low, open, desert vegetation, while the plains and valleys
of higher eleyation support a loose carpet of perennial grasses and
ephemeral herbs, together with a sparse representation of succulent and
semi-succulent types of plants. The higher mountain ranges exhibit
a graduated sequence of vegetation from that of the desert valleys,
through a scrub of evergreen oaks to forests of pine, spruce, and fir.
The bodies of mesophilous vegetation which occupy these isolated
mountain summits, and the stages which connect them with the vege-
tation of the desert, present innumerable phenomena of the greatest
interest to both physiological and floristic plant geography, and form
a most fruitful field of investigation.

The Santa Catalina Mountains are one of the most westerly of the
high ranges of southeastern Arizona, and rise from an approximate
basal elevation of 3,000 feet to a height of 9,150 feet. With respect to
their vegetation these mountains are typical of a large number, not
only in Arizona but in southern New Mexico and northern Mexico as
well. Their location within 20 miles of Tucson and their ready acces-
sibility from the Desert Laboratory have given opportunity for a study
of the distribution of their vegetation and for a measurement of some
of the physical factors upon which the existence and activities of the
vegetation depend. It is the purpose of the present paper to give a
brief description of the vegetistic features of the various altitudes and

5

6 VEGETATION OF A DESERT MOUNTAIN RANGE.

topographic situations in the Santa Catalina Mountains, to give the
results of the climatological instrumentation which has been carried
on, and to indicate in so far as possible the manner and degree in which
the successive altitudinal stages of vegetation are dependent upon the
gradients of climatic change by which they are accompanied.

Some of the instrumental records date from the summer of 1907,
the first year in which members of the Desert Laboratory became
interested in the mountains, but the principal part of the data to be
presented were secured in 1911 and subsequent years. The operation
of instruments and the study of vegetation have been made on visits
of 5 to 10 days, at intervals between April and October. From three
to nine such visits have been made to the mountains in each of the
summers since 1910.

The practical exigencies of the work have limited the character of
the instrumentation which could be carried out, but have not impaired
the accuracy of the data which it was possible to secure. There is no
respect in which the results herewith presented may be considered as
more than a mere outline of a large and widely ramifying botanical
problem. The central interest of the writer has been to determine
which of the major environmental factors are responsible for the chief
distributional features of the vegetation, and to ascertain something
regarding the intensities of the factors responsible for the distributional
limits of individual species, and thereby for the limitation of the types
of vegetation themselves. Such an inquiry into the correlations exist-
ing between physical conditions and the occurrence and activity of
plants may do much to explain general vegetistic phenomena, but it
does far more to open up the innumerable physiological problems which
must be well known at the outset to underlie these correlations.

GEOGRAPHY AND TOPOGRAPHY OF THE SANTA
CATALINA MOUNTAINS.

The Santa Catalina Mountains occupy the drainage divide between
the San Pedro River, a tributary of the Gila, and the Santa Cruz, a
river which seldom has sufficient flow to reach an outlet in the Gila.
The position of the mountains is between 110° 30' and 111° east longi-
tude and 32° 15' and 32° 35' north latitude. The general outline of the
range is roughly triangular (see plate 40), its southern base being at
about 3,000 feet (915 m.) elevation, its northeastern base (parallel to
the San Pedro River) lying at approximately 3,500 feet (1,065 m.).
To the northwest a broad grassy plain, 3,500 to 3,800 feet in elevation,
connects the Santa Catalinas with the lower Tortilla Mountains. To
the southeast a narrow pass, 4,300 feet (1,310 m.) in elevation, connects
with the closely adjacent El Rincon Mountains, which reach an eleva-
tion of 8,465 feet (2,580 m.). Southward from El Rincon range a pass
of 4,000 feet (1,220 m.) elevation leads to the Santa Rita range (9,432

GEOGRAPHY AND TOPOGRAPHY OF SANTA CATALINA MOUNTAINS. 7

feet, 2,875 m.). To the northeast of the San Pedro River rises the
Galiuro range of mountains, the main ridge of which is approximately
35 miles (57 km.) distant from the Santa Catalinas. The next moun-
tains encountered in passing northeastward are the Pinaleno or Graham
range, about 60 miles distant from the Santa Catalinas, and exceeding
them in altitude by about 1,400 feet (427m.). Beyond the upper
course of the Gila River lie the Gila Mountains, and still further to the
northeast the White Mountains, which reach an elevation of 11,280
feet (3,440 m.) in Escudilla Peak. The White Mountains present one
of the largest elevated land masses of the State, connecting to the
northeast, through the Mogollon Mesa, with the elevated region which
surrounds the San Francisco Peaks and supporting a heavy body of
forest which extends from the New Mexican boundary nearly to the
Grand Canon. East of the White Mountains the elevated country ex-
tends for about 75 miles (121 km.) into New Mexico, breaking up into
several diverging ranges which form a part of the Continental Divide,
draining westward into the Gila and eastward into the Rio Grande.

The Santa Catalina Mountains are thus seen to stand at the south-
western terminus of a series of isolated elevations which stretch away
from the southern edge of the Colorado Plateau. The valley of the
Rio Grande and its tributaries, several undrained basins, and the
valley of the Little Colorado combine to separate the entire chain of
elevations — San Francisco Mountains, Mogollon Mesa, White Moun-
tains, and the mountains of western New Mexico — from the Sangre
de Cristo, San Juan, and Jemez mountains of northern New Mexico,
which are virtually a part of the Rocky Mountain system. A consider-
able degree of isolation from the north is thus given to the entire series
of mountains in southeastern Arizona.

To the south and southeast of the Santa Catalinas an irregular but
close-set series of mountains gives them a connection with the Mexican
Cordillera which is much closer than their connection with the Rocky
Mountains. To the west the nearest forest-clad elevations are the San
Jacinto and San Bernardino Mountains of southern California, which
are about 300 miles (480 km.) distant.

The relative isolation of the Santa Catalina Mountains and the
directions in which they possess easy stages of connection with other
elevated regions are of first importance in relation to the genesis and
history of their flora, a subject which will be only briefly touched upon
in this paper (see p. 36).

The southern face of the Santa Catalinas, to which the present in-
vestigation has been confined, is built solely of gneiss of varying degrees
of hardness. The main ridge and the northern and eastern lateral
ridges are worn into a relatively rounded topography, while the south-
western corner of the range possesses rock of greater durability and is
correspondingly more rugged in topography.

8 VEGETATION OF A DESERT MOUNTAIN RANGE.

The highest elevations lie between Mount Lemmon (9,150 feet,
2,790 m.) and Green Mountain (7,900 feet, 2,410 m.), which are only
7 miles (11 km.) apart. Samaniego Ridge and Oracle Ridge extend
northward from the vicinity of Mount Lemmon, falling rapidly in
elevation and terminating in the high plain which lies in the direction
of the Tortilla Mountains. A very rugged ridge extends southwest-
ward from Mount Lemmon and terminates in Pusch Ridge. To the
south of the main ridge an extensive elongated drainage basin has been
developed which lies parallel to the south face of the range and finds
its outlet through Sabino Canon. Several important streams drain the
south slopes of the main ridge and are tributary to Sabino Canon. To
the east of Sabino two important drainages — Bear Canon and Soldier
Canon — drain the eastern end of the main ridge in the vicinity of Green
Mountain, and west of Sabino are Pedregosa, Ventana, Pima, and other
canons which rise in the rugged southwestern portion of the mountain.
All of these streams flow into the Rillito, a tributary of the Santa Cruz
which also drains a portion of El Rincon range. The north face of the
main ridge between Oracle and Samaniego ridges is drained by the
Canada del Oro, which flows at first north, then west, and finally south-
west, emptying into the Santa Cruz. On the northeast slopes of the
range the topography is relatively simple, the high elevations falling
away rapidly in the direction of the San Pedro River. A large number
of minor streams drain this region and give to the San Pedro perhaps
less than one-fourth of the total run-off of the mountains.

The main drainageway of Sabino Canon is the only one in the Santa
Catalinas which possesses a constant flow of water, which is due both
to the great extent of its cachment basin and to the fact that it has
its source on the north slopes of Mount Lemmon in the heaviest body
of timber on the mountain. In the Canada del Oro, in Bear and Soldier
Canons, as well as in a few of the larger canons of the north slopes,
water may be found at all times of the year in certain localities where
the local configuration of the valley or the occurrence of resistant dikes
of rock forces the underflow to the surface. During the rainy seasons
water may, of course, be found in any of the large drainageways. The
heavy local showers of summer often convert even the smallest stream-
ways into rushing torrents for a few hours.

The small size of the Santa Catalinas together with their elevation
gives a steep gradient to all of the major streams. The main stream
of Sabino Canon falls from 7,700 feet at Webber's Cabin to 3,700 feet
at the west end of Sabino Basin, a distance of 6 miles, or a gradient
of fall of 667 feet per mile. From the west end of the Basin to its'
emergence onto the desert this stream falls only 1,000 feet in a distance
of 5 miles. The Canada del Oro falls at a rate of 494 feet to the mile
from its source, just west of Mount Lemmon, to the confluence of its
main tributary from the west slopes of the Oracle Ridge, and at the

GEOGRAPHY AND TOPOGRAPHY OF SANTA CATALINA MOUNTAINS 9

rate of 200 feet to the mile from there to its emergence from the moun-
tain at 3,400 feet. The result of the passage of intermittent and tor-
rential streams through such steep drainageways has been the wearing
down of the stream beds to solid rock throughout almost the entire
drainage system of the mountain. There are no parks nor mountain
meadows, such as are present in some of the largest southwestern
mountains. The flood-plains and alluvial bottoms are all small and
scattered. The spots in which meandering streams may be found are
very few indeed. In Bear Canon a flood-plain nearly half a mile in
length has been formed as a result of a large body of highly resistant
rock, which has narrowed the canon and prevented the outwash of
erosion material. Similarly, in Soldier Canon there is a small flood-
plain below which the stream falls 300 feet in a very short distance
through a narrow gorge. Although Sabino Basin is a locality in which
several converging streams undergo a sudden reduction in their gradi-
ent of fall, there has not been any considerable deposition at that place.
On the contrary the region is one in which the streamways are bordered
and bedded by large boulders in a matrix of coarse sand and are sub-
jected to active scouring by the torrential floods of summer.

Whatever may have been the original form of the Santa Catalinas
they have been so far worked upon by erosion and weathering that
they now possess almost no relatively level areas or regions of inde-
terminate drainage. All of the higher portions of the main ridge and
of the lateral ridges as well are extremely narrow. The only localities
in which the topography broadens and is relatively level are at points
where several drainages have their origin, or places just above precipi-
tous cliffs. On the summit of Mount Lemmon there is a nearly level
area of at least 100 acres (see plate 36 A and B), from which a flat-topped
ridge extends eastward for half a mile, terminating in an abrupt drop,
in the course of which two narrow ridges have their origin. This re-
stricted area of nearly level land is a last relic of a portion of the original
structural form of the mountain, and it will not be many centuries
until it is reduced to the narrow form of the lower ridges.

In the eastern and central portion of the Santa Catalinas the gneiss
weathers readily and gives rise to a loam soil. The precipitate topog-
raphy gives little opportunity for the accumulation of this soil, and
it is thin in almost all localities. Throughout the lower portions of the
range, below the pine forest, the soil has the appearance of being ex-
tremely coarse by reason of the surface coating of angular fragments
from 1 to 5 mm. in diameter. Surface drainage is able to move this
material but slowly by reason of its size and angularity. Just beneath
it may be found a fine soil, still mingled with coarse particles but held
in place by the mulch of stones, which is analogous to " desert pave-
ment." The outcropping rock and larger boulders serve to retard
erosion and to preserve a soil sufficiently deep for shrubs and trees to

10 VEGETATION OF A DESERT MOUNTAIN RANGE.

find root. There are many deep soil-filled crevices through which the
roots of trees are able to penetrate to bodies of deep-seated soil of
favorable moisture content.

The restricted areas of alluvial soil in the desert and lower mountain
regions are of a fine sand or sandy loam and possess considerable humus,
in contrast to the soils of the slopes. At the forested elevations the
soil is similar to that of the evergreen oak region. The soil of the lower
pine belt is scarcely superior in depth or humus content to that of the
upper oak region. Above 7,500 feet, however, the amount of humus,
as well as the amount of surface litter, increases with the increasing
density of the stand of pines. On the north-facing slopes which are
clothed with fir forest the soil is not much if at all deeper than in the
heavy stands of pine, but is notably richer in organic matter.

The alluvial slopes which immediately surround the mountain are
so closely related to it in all of their physical and biological features
that it will be necessary in the following pages to give some considera-
tion to their vegetation. Throughout the arid southwest the long
straight profiles presented by the outwash slopes of the hills and moun-
tains form one of the characteristic features of the landscape. The
distinct character of these slopes is to be attributed to the manner in
which they have been laid down under conditions of torrential rainfall
and of violent and intermittent stream flow, and their distinctness from
the parabolic alluvial slopes of the humid regions has caused Tolman *
to designate them technically by their popular Spanish name "bajada."]'

The bajadas constitute almost the total area of all the intermontane
valleys of southern Arizona. The only portions of the valleys topo-
graphically separable from them are the stream beds, the flood-plains,
and the "play as" or undrained areas into which one or more streams
flow and deposit their load. To the student of vegetation there are
marked differences between the upper and lower portions of all bajadas.
The differences in the physical features presented by upper and lower
bajadas of the same elevation have been only superficially investigated ;
the differences in their vegetation are very obvious, as will be described.
The principal environmental features which appear to differentiate the
high and low bajadas are the coarser character of the soil in the high
bajadas, the possibility of higher soil moisture in them, at least at a
depth of several feet, and the greater development of calcareous incrus-
tations, or " caliche," in the soil of the low bajadas. The layers of
caliche lie near the surface in some places, while in others the upper-
most ones have been covered by deposition; they extend downwards
for a few feet in some cases, or more frequently recur to a depth of
100 feet or more.

The bajada of the southern face of the Santa Catalinas has been

* Tolman, C. F. Erosion and Deposition in the Southern Arizona Bolson Region. Jour. Geol.,
vol. 17, pp. 136-163, 1909.
t Pronounced bahada.

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 11

truncated at its lower edge by the Rillito, which flows at right angles
to the slope of the bajada, and on the western side of the range the
Canada del Oro has worn off the lower edges of the detrital slopes in
similar matter. The well-developed bajadas which lie between Pima
and Ventana Canons fall at the grade of 150 to 175 feet per mile.
Between Ventana and Bear Canons the uppermost portion of the
bajada has been worn away, so that at present a shallow valley lies
between the base of the mountain and the lower portion of the old
bajada, now cut into isolated and rounded hills. On the northeast
side of the Santa Catalinas the bajada which extends down to the San
Pedro River exhibits approximately the same grade as the bajada at
Pima Canon. Its surface is crossed, however, by so many drainages
from the steep northeast face of the mountain that the bajada region
consists of a series of rounded ridges extending out from the base of
the mountain, very unlike the relatively flat bajadas of the Santa Rita
and El Rincon ranges.

VEGETATION OF THE SANTA CATALINA MOUNTAINS.

The journey from the base to the summit of the Santa Catalina
Mountains brings to the eyes of the observer a constantly changing
panorama of vegetation. New types of plants are constantly being
encountered with increase of altitude, while types already familiar are
being left behind. There is no portion of the mountain, at least below
7,500 feet, in which a climb of 500 feet does not materially alter the
physiognomy of the surrounding vegetation. The course of the vege-
tational panorama is not merely a gradual transition from the open
desert of succulents and microphylls to the heavy fir forest which
occupies the summit of Mount Lemmon (plate 1). There are inter-
posed between these vegetations two distinct belts of plant life through
which this tremendous transition takes place.

The arborescent cacti and the trees and shrubs of the desert give
way gradually to evergreen oaks, leaf-succulents, sclerophyllous shrubs,
and perennial grasses. This open but arborescent vegetation reaches
a full development and then gives way to pine forest, with a distinctive
accompanying carpet of herbaceous perennials. The pine forest is
then, in turn, invaded by spruce and fir and the heavy stands of these
trees are accompanied by still another assemblage of shrubs and her-
baceous plants. The striking character of these gradations of vegeta-
tion is not due solely to the contrast between the varied vegetation
of the open desert and the monotony of the closed coniferous forests,
but is quite as largely due to the striking types of plants which are
to be found both in the desert and in the region of evergreen oaks.

A first and most general observation of these vegetational stages will
discover the distinctive regions of desert, of park-liks semi-desert and
of forest. The first is like the desert of the extensive bajada slopes
which surround the entire mountain; the second is similar to plant

12

VEGETATION OF A DESERT MOUNTAIN RANGE.

communities which may be seen in southern Texas and in California, as
well as in similar situations in Arizona and New Mexico; the last is
essentially like the great body of yellow pine forest which stretches
from southern Jalisco to British Columbia, or like the fir and Douglas
spruce forests of the Rocky Mountains. These three major regions

FTET
9,000

8,000
7,000
6,000
5,000
4,000
3,000

9,000
8,000
7,000
6,000
5,000
4,000
3,000

9,000
8,000
7,000

6,000 -

3,000 -

SOUTH

NORTH

Fir forest

Pine forest

Fouquieria splendens

Quercus emoryi

Dasyiirion wheeieri

Quercus hypoieuca

FIG. 1. — Diagram showing vertical distribution of Desert, Encinal, Pine
Forest, and Fir Forest in relation to slope exposure, together with dia-
grams showing effect of slope exposure on vertical distribution of Fou-
quieria splendens, Dasyiirion wheeieri, Quercus emoryi, and Quercus
hypoieuca.

constitute the most natural and easily distinguished subdivisions of
the vegetation, and depend for their distinctness on the radical dis-
similarity of the dominant types of plants in each. They may best be
designated by the simple terms Desert, Encinal,* and Forest. The

* The Spanish word "encinal" signifies a grove or forest of evergreen oaks, being derived from
encina, evergreen oak. The suitability of the word was suggested by Prof. J. W. Harshberger.

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 13

Encinal belt is essentially a region dominated by sclerophyllous trees
and shrubs and by semi-succulent perennials, with an open stand of
perennial grasses. It is what is designated by the Forest Service as
the " woodland type" of forest. The pine and fir forests are very
dissimilar in their floristic composition, but they are much more closely
alike vegetistically than are any two of the three major divisions which
have been made. A further and more careful examination of the stages
which connect the Desert with the Forest will discover not only the
inevitable gradations between the three major regions, but also
several minor features which cause constantly recurring departures
from the typical or ideal vertical distribution of the vegetation. The
influence of slope exposure on the vertical ranges of both the individual
species and the vegetation itself is a feature which these mountains
share with almost all extra-tropical mountains; the distinctive vegeta-
tion of flood-plains and streamways is also as clearly noticeable here
as in all arid and semi-arid regions; the occurrence of the lowland
species at higher altitudes on ridges than in the valleys is also a strong
differentiating feature.

In describing the salient physiognomic and floristic features of the
vegetation, and its distributional behavior, it is expedient to recognize
primarily the three major divisions of Desert, Encinal, and Forest,
and then to take into account secondarily the degree to which the
components of these regions intermingle and the extent to which the
topographic irregularities of the mountain cause an alternation and
interdigitation of the three regions.

The basal slopes of the mountain between 3,000 and 4,000 feet (915
and 1,220 m.) present few vegetational distinctions from the upper
bajadas, and almost no distinctions of flora. Between 4,000 and 5,000
feet (1,220 and 1,525 m.) there is a rapid elimination of all but a very
few of the characteristic desert species, and on north slopes at the
latter elevation nearly all of the dominant Encinal forms have made
their entry. The upper limit of the Desert may be placed at 4,000 feet
for north slopes and 4,500 feet (1,472 m.) for south slopes. The upper
edge of the Desert exhibits an attenuated occurrence of all of the larger
desert plants and the presence of many perennial grasses and semi-
woody plants which occur both in the Encinal Region and on the bajadas
of equal or slightly greater elevation in the neighboring portions of
Arizona. The extreme upper limit of desert forms is 7,000 feet (2,133
m.), an elevation which is reached by a single succulent species. Follow-
ing the dissimilarity of the lower and upper portions of the Desert Region
they have been described separately.

The Encinal Region extends from the occurrence of the first ex-
tremely open groves of evergreen oaks on north slopes at 4,000 feet up
to the first elevation at which the larger pines begin to dominate the
physiognomy of the vegetation, at about 6,300 feet (1,920 m.) on south

14 VEGETATION OF A DESERT MOUNTAIN RANGE.

slopes. A few of the dominant species of the Encinal are found on
the higher bajadas, above 3,500 feet (1,067 m.) elevation, but in the
mountains none of its species is to be found so low as this except on
north slopes or near arroyos of large drainage area. At 4,000 feet, on
north slopes, several of the larger Encinal plants are encountered, and
at 4,300 feet (1,310 m.), on north slopes, several additional dominant
species are found. Within the Encinal Region it is possible to recognize
a lower and an upper portion, distinguished chiefly by the openness of
the former and the closed character of the latter. The closed Upper
Encinal merges gradually into the Forest, losing some of its character-
istic species at 6,000 to 6,300 feet (1,830 to 1,920 m.), while others range
to 7,800 feet (2,380 m.) and a few to 8,300 feet (2,530 m.) on south
slopes.

The lowest occurrence of Forest is on north slopes at about 5,800
feet (1,768 m.) and on south slopes at about 6,300 feet (1,920 m.).
The Forest is at first rather open and is superposed, as it were, upon
the closed Encinal, but it becomes heavier and the Encinal elements
within it become more sparse at elevations of from 6,300 to 6,800 feet
(1,920 to 2,073 m.), according to the slope exposure. The upper limit
of Forest is not reached in the Santa Catalina Mountains at their
highest elevation of 9,150 feet (2,790 m.), nor in the adjacent Pinaleno
Mountains (Mount Graham) at 10,516 feet (3,205 m.). The forest of
yellow pine occupies all south slopes up to the summit of Mount
Lemmon. A forest dominated by fir, spruce, and Mexican white pine
occupies the north slopes above 7,500 feet (2,287 m.), the earliest
occurrence of these species being about 1,000 feet (305 m.) lower.

The description of vegetation which is given in the following pages
applies only to the south face of the Santa Catalinas. The north face
presents more abrupt slopes than the south, with most of its ridges
running north from the main ridge. This circumstance obscures the
influence of slope exposure, since it presents opposed slopes, facing
east and west, which are identical in their vegetation.* Furthermore
the north face of the range is mineralogically diversified, presenting
exposures of shale, sandstone, limestone, diorite, and gneiss, whereas
the south face presents an exposure of gneiss only, with a resultant
mineral identity of soils from base to summit. It has thus been possible
to carry out a study of climatic influences over a vertical gradient of
6,000 feet (1,830 m.) with uniform soil, and the east and west ridges
of the south face have furnished opposed north and south slopes at all
elevations.

* Differences between the vegetation of east and west slopes have been pointed out by Blumer
for El Rincon Mountain, but the differences noted were of another character from those commonly
existent between north and south slopes. See: Blumer, J. C. A Comparison between two Moun-
tain Sides. The Plant World, 13: 134-140. 1910.

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 15

THE DESERT REGION.*

Under the designation of "Desert" are comprised all those portions
of the Santa Catalina Mountains in which the vegetation is open, low,
and diversified in the assemblage of growth-forms, with a predominance
of microphyllous trees and shrubs and an abundance of cacti. Such
a vegetation is to be found covering the upper bajadas and extending
up the slopes of the mountain to elevations of 4,000 to 4,500 feet
(1,220 to 1,372 m.), according to slope exposure. The vegetation of the
Upper Bajadas will be described for the sake of the contrast which it
affords with the vegetation of the upper portions of the mountain, as
well as to give a picture of the plant life by which the mountain is
surrounded and from which it has derived many of its characteristic
species. The desert slopes of the mountain itself exhibit at first a close
resemblance to the bajada, and then lose most of the larger bajada
plants before the entry of the dominant plants of the Encinal region.
This circumstance admits of a subdivision of the Desert region of the
mountain into Lower Desert Slopes and Upper Desert Slopes. The
latter region is much poorer than the former in cacti and much richer
in grasses, both from the standpoints of the number of species and the
number of individuals. The Upper Desert is similar in vegetation to
many of the Upper Bajadas, such as those to the northwest of the Santa
Catalinas and to the east and west of the Santa Rita Mountains, and
might well be designated as " semi-desert " or " desert-grassland transi-
tion." It is, however, essentially similar to the desert plains in its
vegetational make-up, and in no part of Arizona does it serve as a
transition to true grassland. The largest canons of the Santa Catalinas
possess some plant communities that are radically unlike the vegeta-
tion of the desert itself, but not unlike the communities which surround
the springs and wells of the desert plains. These are the groups of
aquatic and palustrine plants which accompany the streams and pools
of the canons. The smaller canons and arroyos t present distinctive
features of vegetation, departing more and more from the large canons
and approaching more nearly the character of the desert areas away
from water. All of these areas have been treated as a part of the Desert
Region.

THE UPPER BAJADAS.

The Lower Bajadas of the Tucson region are covered by a vegeta-
tion in which Covillea tridentata (jediondia, creosote bush) is always the
predominant plant and is often almost the sole plant of more than
2 feet in height over areas many square miles in extent. The plants
which most commonly enter this community are Prosopis velutina
(mesquite) , Opuntia fulgida, Opuntia spinosior (both arborescent cylin-

* The word "region" is not here used in any of the technical senses in which it has been em-
ployed in phytogeography.

t The Spanish word arroyo is in common use in the southwestern United States to designate
streamways which are usually without water.

16 VEGETATION OF A DESERT MOUNTAIN RANGE.

dropuntias), Opuntia toumeyi, Opuntia blakeana (procumbent plato-
puntias), and Acacia paucispina.

The shorter and steeper Upper Bajadas which fringe the southern
and southwestern edge of the Santa Catalinas are clothed with a much
more diversified vegetation, in all respects similar to that of other
Upper Bajadas which lie below 3,500 feet (1,067 m.) in other localities
in southwestern Arizona. The freedom of the soil from caliche is here,
as elsewhere, responsible for the existence of a diversified vegetation
rather than a pure stand of Covillea.

The Upper Bajadas, as exemplified along the south face of the Santa
Catalinas at about 3,000 feet elevation (915 m.), bear what may be
regarded in many respects as the most highly developed type of desert
vegetation to be found in southern Arizona or northern Sonora. In the
Upper Bajadas may be found a greater number of species of perennial
plants than in any other distinctly desert situations. In them also
the number of individual perennial plants per unit area is greater than
in any areas outside the flood plains of such rivers as the Santa Cruz
and Gila. The only areas that compare with the High Bajadas in
these respects are the volcanic hills in which basaltic rock has weathered
to a fine clay which is very retentive of soil moisture, as is well exempli-
fied in Tumamoc Hill, the site of the Desert Laboratory. The andesitic
and rhyolitic hills in the vicinity of Tumamoc are much poorer than
it is in the number of individual plants per unit area, although perhaps
nearly as rich in their flora.

On the Upper Bajadas there often occur, in almost equal admixture,
from 15 to 25 perennial species of plants of such size as to dominate the
physiognomy of the vegetation. These same species may be found on
the more nearly level Lower Bajadas, but any one of them may often
be absent for many miles, may be sporadically represented by a few
individuals, or may occur in dense but local colonies (particularly in
the case of the cacti). Occasionally as many as 5 to 10 of the species
may be within sight at the same time.

The flora which characterizes the Upper Bajadas of the Santa
Catalinas ranges without substantial loss down to sea-level on the gulf
of California,* and the vegetation formed by their commingling may
be found as a belt covering the high bajadas which encircle all of the
mountain ranges and clothing all of the low basaltic hills. A climb of
2 hours from the base of the Santa Catalinas will discover greater
changes of vegetation and flora than can be encountered in the 150
miles (242 km.) between Tucson and Adair Bay.

The Upper Bajadas present the desert characteristics of openness
of stand, lowness of stature, and commingling of diverse vegetation
types. The first of these features is common to the vegetation of all

*SeeHornaday,W.T. Camp Fires on Desert and Lava. New York, Scribner, 1909. MaoDougal,
D. T. Across Papagueria. The Plant World, 11: 93-99, 123-131, 1908.

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 17

deserts, the last is at least characteristic of the less pronounced deserts
of the southwestern United States and of Mexico. The openness of the
stand is such that it is possible in all places to ride a horse through the
vegetation and to take whatever course the rider may wish, with only
occasional digressions of a few yards from the general direction of
travel. The stature of the vegetation is such that it would be possible
for the rider to keep almost constantly in view another mounted man
half a mile distant. The columnar giant cacti reach a maximum height
of 40 feet (12 m.) and the trees a height of 20 to 25 feet (6 to 8 m.).
The great bulk of the shrubs and succulents, however, are not more
than 6 feet (2 m.) in height, and many of them are less than 4 feet
(1.2 m.). Among the commonest vegetation types are stem-succulents,
microphyllous and sclerophyllous trees and shrubs, macrophyllous decid-
uous shrubs, perennial grasses, and root-perennial and ephemeral her-
baceous plants.

Largest and most conspicuous of the succulents is Carnegiea gigantea
(saughuaro, giant cactus), which is here in its optimum habitat and
very abundant (plate 3, B) . Among the microphyllous trees the most
abundant are Prosopis velutina, Acacia greggii, Acacia paucispina, and
the green-barked Parkinsonia microphylla (palo verde). The much-
branched arborescent types of cacti are represented by Opuntia versi-
color, which attains a maximum height of 12 feet (4 m.), and by Opuntia
fulgida and Opuntia mamillata (cholla), remarkable for the brilliance
of their glistening straw-colored spines. Opuntia blakeana, Opuntia
engelmanni, Opuntia toumeyi, and Opuntia discata are abundant repre-
sentatives of the platopuntia group. The evergreen Covillea is greatly
outnumbered by Fouquieria splendens (ocotillo) . The globular Ecliino-
cactus wislizeni (bisnaga) attains a height of 4 feet (1.3 m.) with an
even greater girth. Similar in form but never exceeding a foot in
height are Echinocereus fendleri and Mamillaria grahami. The sclero-
phyllous Simmondsia calif ornica (jojobe) and the relatively large-leaved
deciduous Jatropha cardiophylla are frequent, while a large number of
less striking shrubs are common, including Franseria deltoidea, Isocoma
hartwegi, Encelia farinosa, Zizyphus lycioides var. canescens, Lycium
torreyi, Momisia pallida, Krameria glandulosa, Trixis angustifolia var.
latiuscula, Crassina pumila, and Psilostrophe cooperi.

The seasonal rains of winter and those of summer cause activity of
foliation and growth on the part of all of the smaller shrubs. The
winter rains cause foliation in Parkinsonia and Fouquieria, but not in
Prosopis and the species of Acacia. Neither do they initiate growth
in Parkinsonia, Fouquieria, nor any of the cacti. The two widely
separated seasons of rain bring forth two wholly distinct sets of herba-
ceous ephemeral plants, at the same time that each season causes activ-
ity upon the part of some of the root-perennials. The ephemeral plants
may form a dense carpet over both the Upper Bajadas and the Lower

18

VEGETATION OF A DESERT MOUNTAIN RANGE.

Bajadas in seasons of well-distributed and copious rainfall. With less
favorable conditions they may form a very light cover or may be almosc
absent. The total flora of root-perennials and ephemerals is large, and
the relative abundance of the various species fluctuates tremendously
from spot to spot, and in the same spot it is by no means the same from
year to year. This flora is nearly identical with that of the basaltic hills
in the vicinity of Tucson, and has been fully listed by Thornber,* with
a subdivision of vegetation types.

In the following list have been brought together the names of the
most characteristic plants of the Upper Bajadas, grouped vegetistically
and briefly described. Asterisks indicate the relative abundance of
the species — three indicating that a plant is extremely common, two
that it is very common, and one that it is fairly common. Figures
follow the descriptions, indicating the average height of each species.
A comparison of all the maximum heights given will convey an impres-
sion of the low stature of the commonest components of the vegetation.

Vegetislic Grouping of the Characteristic Species of the Upper Bajadas.

***

Perennial Non-succulent Trees and Shrubs:
*** Acacia greggii, microphyllous, winter-
deciduous. 2 to 3 m.
** Covillea tridentata, microphyllous,
evergreen. 1 to 2.5 m.

* Crossosoma bigelovii, sclerophyllous,

evergreen. 1 to 1.5 m.

* Ephedra trifurca, aphyllous, green-

stemmed. 0.5 to 1 m.
Fouquieria splendens, macrophyllous,

drought-deciduous. 2 to 4 m.
"* Jatropha cardiophylla, macrophyllous,
winter-deciduous. 0.5 to 1 m.

* Kceberlinia spinosa, aphyllous, green-

stemmed. 0.5 to 1 m.
Krameria glandulosa, microphyllous,

evergreen. 1 to 2 m.
Lycium berlandieri, microphyllous,

evergreen. 1 to 2 m.

* Lycium fremontii, microphyllous,

evergreen. 1 to 2 m.
** Momisia pallida, sclerophyllous, ever-
green. 1.5 to 2.5 m.

* Olneya tesota, microphyllous, ever-

green (foliage occasionally winter-
killed). 3 to 6m.

*** Parkinsonia microphylla, microphyl-
lous, drought-deciduous, green-
stemmed. 2 to 5 m.

*** Prosopis velutina, microphyllous,

whiter-deciduous. 3 to 6 m.
** Zizyphus lycioides var. canescens,
microphyllous, evergreen 1 to 2 m.

**

**

Perennial Succulent Plants:

*** Carnegiea gigantea, columnar,

branched. 5 to 14 m.
** Echinocactus wislizeni, cylindrical.

0.5 to 1.5 m.
** Echinocereus fendleri, cylindrical, cae-

spitose. 0.1 to 0.4 m.
** Mamillaria grahami, cylindrical, soli-
tary or csespitose. 0.1 to 0.2 m.
** Opuntia blakeana, flat-jointed, pro-
cumbent.

*** Opuntia discata, flat-jointed, pro-
cumbent or semi-erect.
** Opuntia fulgida, cylindrical-jointed,

arborescent, 1 to 2 m.
*** Opuntia mamillata, cylindrical-
jointed, arborescent. 1 to 2 m.
** Opuntia toumeyi, flat-jointed, pro-
cumbent.
*** Opuntia versicolor, cylindrical-jointed,

arborescent. 1 to 4 m.
Perennial Shrublets (all less than 0.7 m. high):

* Coldenia canescens, sclerophyllous.
** Crassina pumila, dissected leaves.

*** Encelia farinosa, macrophyllous,

slightly drought-deciduous.
*** Franseria deltoidea, sclerophyllous.
*** Isocoma hartwegi, dissected leaves.
*** Kalliandra eriophylla, dissected leaves.
** Psilostrophe cooperi, sclerophyllous.

* Trixis angustifolia var. latiuscula,

sclerophyllous, slightly drought-
deciduous.

* Thornber, J. J. Vegetation Groups of the Desert Laboratory Domain. Carnegie Inst.
Wash. Pub. 113, Chapter IV, 1909.

VEGETATION OF THE SANTA CATALINA MOUNTAINS.

19

Ephemeral Summer-active Herbaceous Plants — Continued.

**

***

Root Perennials (all facultative evergreens):
*** Abutilon incanum, sclerophyllous.
Brodicea capitata var. pauciflora,

bulbous, linear leaves.
Cassia coz;esi'i,sclerophyllous,branched

leaves.

Dalea parryi, microphyllous.
Muhlenbergia porteri, semi-scandent.
** Pentstemon wrightii, macrophyllous.
** Perezia wrightii, macrophyllous.
** Verbena ciliata, macrophyllous, hairy.
Ephemeral Summer-active Herbaceous Plants:
** Bahia absinthifolia.
*** Bailey a multiradiata.
** Boerhaavia pterocarpa.

* Boerhaavia watsoni.

* Bouteloua aristidoides.
Cladothrix lanuginosa.
Euphorbia Jlorida.
Euphorbia melanadenia.

***
**

***

**
***

Ephemeral Summer-active Herbaceous Plants:
Continued.
Pectis papposa.
Wedelia iiwarnata.
Ephemeral Winter-active Herbaceous Plants.

* Actinolepis lanosa.
Anisolotus trispermus.
Baeria chrysostoma.

** Chorizanthe brevicornu.
** Cryptanthe intermedia.

* Eremocarya micrantha.
"** Gilia floccosa.

** Lepidium lasiocarpum.
** Lesquerella gordoni.

* Mentzelia albicaulis.

* Orthocarpus purpurascens.
Pectocarya linearis.
Plantago fastigiata.

*** Plantago ignota.

**
***

THE DESERT ARROYOS AND CANONS.

In crossing the Upper Bajadas it is often possible to detect, by means
of the vegetation, the approach to a very shallow drainageway through
which water runs for only a few hours after the severest summer rains.
The larger arroyos are still more conspicuous by reason of the still
heavier stand of vegetation along their margins, and in the largest
canons is found the culmination of the influence of surface streams
and underflows for the support of vegetation. The effect of the most
transitory of the small streams is merely the raising of the moisture of
adjacent soil to such a point that it will present favorable conditions
for plant activity for a longer time after the close of the rainy periods
than will the soil of the bajada in general. There is only a negligible
and short-lived underflow in these smallest arroyos, and their only
differences from the bajada are that in the rainy seasons they present
slightly more favorable conditions with respect to soil moisture and
that the effect of the rainy season is slightly prolonged in them, while
the periods of drought are correspondingly shortened. In the larger
arroyos there may not be a constant underflow, but there is at least a
relatively high percentage of soil moisture for periods of sufficient length
to greatly reduce the influence of the arid periods upon their plants.
In the largest arroyos and in the mountain canons themselves there is
either a constant underflow, maintaining high moistures in the soil of
the banks and bed of the arroyo, or else there is constant water, either
running or standing in pools.

The smallest arroyos, which are very frequent on the Upper Bajada
in the close proximity of the mountain, present no peculiar species, but
merely a closer stand of the same plants that are to be observed
throughout the bajada, notably Prosopis, Acacia greggii, and Momisia
pallida. Along somewhat larger arroyos are to be found still heavier
stands of the above species, together with Parkinsonia torreyana, Celtis

20 VEGETATION OF A DESERT MOUNTAIN RANGE.

reticulata, Baccharis sarothroides (batamote), Franseria ambrosioides,
Lycium fremontii, Verbesina encelioides, and Bebbia juncea.

In the canons and arroyos large enough to have a heavy flow of
storm water but not large enough to have even pools of water which
are constant throughout the year, there may be found several additional
species of plants which also occur on the sandy flood-plains of the
largest canons. Prominent among these are Chilopsis linearis, Hymeno-
clea monogyra, and Baccharis glutinosa, all of which are large shrubs
or in the case of Chilopsis may attain the size of small trees. Also
characteristic of these sands are Franseria ambrosioides, Rumex hymeno-
sepalus, Euphorbia pediculifera, Clematis ligusticifolia, and Calyptridium
monandrum.

In the largest canons of the south side of the Santa Catalinas it is
possible to witness the occurrence of communities of mesophilous,
palustrine, and aquatic plants which are limited in area but are made
up of species which stand strongly in contrast with the predominant
forms of the bajadas. The existence of streams and pools adjacent to
rocky slopes makes it possible in several places for Callitriche and
Isnardia to grow within 20 feet of Carnegiea and Fouquieria.

At the mouth of Soldier Canon the rocky slopes of the streamway
are clothed with typical bajada plants together with a few forms which
are particularly abundant on cliffs and in rocky situations, both in the
larger mountains of the region and in the volcanic hills. Among the
latter are Opuntia bigelovii, Hyptis emoryi, Lippia wrightii, Anisacan-
thus thurberi, Encelia farinosa, Eriogonum wrightii, Chrysoma laricifolia,
and Crossosoma bigelovii. Among the boulders bordering the stream-
way are Janusia gracilis, Plumbago scandens, Maurandia antirrhini-
flora, Mimitanthe pilosa, and Stemodia plumieri, as well as occasional
individuals of several species which are common away from streams
at elevations of 4,000 to 5,000 feet, as, for example, Dasylirion wheeleri,
Nolina microcarpa, Erythrina flabelliformis, Ingenhousia triloba, and
Mimosa biuncifera. When the sands of the arroyo have not been
recently scoured by floods they support scattered individuals of Ama-
ranthus palmeri (celite), Cassia leptocarpa, Nicotiana trigonophylla,
Bebbia juncea, Hymenoclea monogyra, Franseria xanthocarpa, Asclepias
linifolia, Baccharis sarothroides, Mentzelia gracilenta, and Carduus sp.

In Ventana, Bear, and Sabino Canons it is possible at all times of
the year to find small colonies of palustrine and aquatic plants, and
the vicinity of such localities is the optimum habitat for Prosopis and
Populus. An underflow passes out at the mouth of Sabino Canon which
is heavier and more constant than that of any other canon in the range;
this gives Sabino Canon its abundant mesophilous vegetation and also
causes the arroyo through which its flood waters reach the Rillito to
be occupied by a much richer stand of vegetation than is to be found
along any of the arroyos of the adjacent region. The sandy and

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 21

boulder-strewn bed is from 100 to 300 feet in width and the portions
covered by the storm floods vary from season to season. Trees not
only occupy the banks of the arroyo but occur scattered throughout
its bed, where they may persist for many years, only to be eventually
uprooted by some exceptionally severe freshet. Among these trees
are Populus sp., Fraxinus tourney i, Juglans major, Platanus wrightii,
Sapindus drummondii, Prosopis velutina, and Sambucus mexicana, some
of which reach a height of 50 to 60 feet (15 to 18 m.). Such shrubs as
Chilopsis linearis, Baccharis sarothroides, Baccharis glutinosa, and
Hymenoclea monogyra are also common in and along this arroyo.

The floor of Sabino Canon from its mouth to the Sabino Basin is
occupied by an irregular and broken forest of Prosopis, Populus, Frax-
inus, Platanus, and Salix (Salix wrightii and Salix sp.). Among these
common trees are scattered a few individuals of three of the oaks
characteristic of the Encinal region: Quercus oblongifolia, Quercus
arizonica, and Quercus emoryi. These oaks occur near the mouth of
the canon at an elevation of 2,700 feet (823 m.), although their lowest
occurrence on slopes, away from the proximity of an underflow, is at
4,200 to 4,500 feet (1,280 to 1,372 m.). Cupressus arizonica occurs in
the upper half of the canon and in Sabino Basin at elevations above
3,200 feet (975 m.). It is confined to the proximity of streams up to
an elevation of 6,000 feet, above which it is occasionally found on
slopes.

The shrubby vegetation of the floor of Sabino Canon includes all
of the species which have been mentioned as occurring in the smaller
canons and arroyos, together with a number of shrubs and perennials
which are more common along the arroyos and streams of the Encinal
region. Among the latter are those species mentioned as occurring
in Soldier Canon and also Vitis arizonica, Bouvardia triphylla, Amorpha
californica, and Brickellia calif ornica.

Other characteristic plants are the shrubs Dodoncea viscosa var.
angustifolia, Eysenhardtia orthocarpa, Indigofera sphcerocarpa, and the
woody climber Nissolia schottii.

Among the herbaceous plants common in and along the pools and
water-courses of Sabino Canon may be mentioned :

Callitriche sp.
Car ex sp.

Cerastium texanum.
Cyperus inflexus.
Helenium thurberi.
Hydrocotyle ranunculoides.
Isnardia palustris.
Juncus arizonicus.
Juncus bufonius.
Juncus interior.
Juncus sphcerocarpus.
Linaria canadensis.
Mecardonia peduncularis.

Mimulus langsdorfii.
Mimitanthe pilosa.
Montia perfoliata.
Myosurus cupulatus.
Phalaris intermedia.
Platystemon calif ornicus.
Polygonum sp.
Senecio lemmoni.
Specularia biflora.
Stachys coccinea.
Stemodia durantifolia.
Tagetes lemmoni.
Tillcea erecta.

22 VEGETATION OF A DESERT MOUNTAIN RANGE.

THE LOWER DESERT SLOPES.

On leaving the uppermost edge of the bajada and commencing the
ascent of the mountain over the rather abrupt slopes which lie between
the larger canons, a region is entered upon in which the physical con-
ditions differ from those of the bajada chiefly in the pronounced slope
exposure to the south, southwest, or southeast, and in the occurrence
of large masses of rock in situ, with the coarse soil limited to small
benches, pockets, and fissures. The vegetation of these lowest slopes
is very similar to that of the Upper Bajadas, and is composed of a
nearly identical flora. Prosopis, Parkinsonia, and Acacia are repre-
sented by smaller and less frequent individuals, and both the cylindro-
puntias and platopuntias occur somewhat less frequently. Carnegiea
gigantea is even more abundant on the slopes than on the bajadas,
being represented by smaller individuals, among which relatively few
have reached the size at which branching begins. For Carnegiea and
the above-mentioned trees the relatively rapid erosion of the soft gneiss
and the shifting of the shallow soil are apparently too great to permit
the attainment of great size or age. Fouquieria, Encelia, and Chrysoma
laricifolia are even more abundant on the slopes than on the Upper
Bajada, and Opuntia bigelovii, the most densely spiny of all the cylin-
dropuntias, is found exclusively on southerly slopes and cliffs, in very
rocky substratum, at elevations below 3,500 feet (1,067 m.). Olneya
tesota and Covillea have not been detected on the mountain slopes (see
plates 4 and 5).

The summer and winter ephemerals of the bajada are nearly all
to be found on the Desert Slopes of the mountain, but rarely in such
abundance as they attain on level ground. Among the most common
of the ephemerals and root-perennials to be observed in the summer
are Cladothrox lanuginosa, Pectis papposa, Euphorbia florida, Bcer-
haavia pterocarpa, Bouteloua aristidoides, Andropogon saccharoides,
Wedelia incarnata, Machceranthera tanacetifolia, Triodia mutica, Evol-
vulus arizonicus, Allionia gracillima, and Cassia covesii. The bases of
boulders and partially shaded ledges of rock are the habitats of Selagi-
nella rupincola, Cheilanthes lindheimeri, and Notholcena hookeri. The
ferns are not common and are conspicuous only during rainy periods,
but the Selaginella is abundant here and becomes even more so at
slightly higher elevations, where it frequently clothes the rocky walls
of steep canons to such an extent that their usual grayness is converted
to a vivid green a few hours after a heavy rain (see plate 7) .

The ascent from 3,500 to 4,000 feet (1,067 to 1,220 m.) witnesses
the first essential changes in the vegetation. At the latter elevation
nearly all of the typical desert forms may be found, but Opuntia has
become infrequent and Carnegiea gigantea, Echinocactus wislizeni, and
Fouquieria splendens are conspicuously confined to southerly slopes
(see plates 6 and 8) . Parkinsonia torreyana, which is confined to arro-

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 23

yos on the bajada, is found here growing with Parkinsonia microphylla,
which it eventually exceeds in vertical distribution by nearly 500 feet
(153 m.). Prosopis is even more abundant at 4,000 feet (1,220 m.)
than it is on the lowest slopes, and attains a trunk circumference of
6 feet (2 m.) at 4,200 feet elevation (1,280 m.), within 600 vertical
feet (183 m.) of its upper limit. Such common shrubs of the bajada
as Lycium, Zizyphus, Krameria, Jatropha, and Momisia are now very
sporadic in their occurrence, and the compact, hemispherical and
vividly green Chrysoma laricifolia has become very frequent and con-
spicuous, together with the white-tomentose Artemisia ludoviciana and
the less conspicuous Eriogonum wrightii.

On northerly slopes, just below 4,000 feet (1,220 m.), are encountered
the first individuals of the rosaceous tree Vauquelinia californica and
of Agave palmeri and Dasylirion wheeleri (sotol). Along the arroyos
the most conspicuous forms are Erythrinaflabelliformis, the large leaves
and brilliant scarlet flowers of which recall its tropical congeners,
Manihot carthaginensis, with leaves of striking form, and Ingenhousia
triloba (wild cotton), with tripartite leaves and large white flowers
which strikingly resemble those of the cotton plant.

THE UPPER DESERT SLOPES.

The slopes lying between 4,000 and 4,500 feet (1,220 and 1,372 m.)
constitute the upper edge of the desert. On these slopes all the char-
acteristic species of the bajada are confined to southerly slopes, and
all but half a dozen of them find their uppermost limits. On the Upper
Desert slopes Vauquelinia becomes common, although confined to
ledges of rock, and Juniperus pachyphlcea, Quercus oblongifolia, and
Quercus arizonica occur for the first tune away from canons. On the
northerly slopes, where these trees form the lowest attenuated edge of
the Encinal region, Dasylirion occurs in abundance together with the
lowest individuals of Nolina microcarpa (bear grass), Arctostaphylos
pungens (manzanita), Agave schottii, and Yucca macrocarpa.

The physiognomy of the Upper Desert slopes is made distinctive
from that of the Lower Desert slopes not only by the entrance of these
plants of striking form, and the exit of the desert species, but also by
the abundance of perennial grasses, root-perennials, and small shrubs,
which combine with the ephemeral plants, or their dead remains, to give
a much more complete ground cover than is to be found in any part of
the bajadas. The compact and extended patches of Agave schottii are an
important element in this low cover, and so are the scattered plants of
Boutelouarothrockiiand the bunches of Boutelouacurtipendula, Bouteloua
oligostachya, Muhlenbergia dumosa, Andropogon scoparium, Eragrostis
lugens, and Heteropogon contortus (see plates 8 and 9).

Commonest among the low shrubs and other perennials of the Upper
Desert are: Chrysoma laricifolia, Acacia suffrutescens, Eriogonum

24 VEGETATION OF A DESERT MOUNTAIN RANGE.

wrightii, Dalea wislizeni, Calliandra eriophylla, Hymenopappus mexi-
canus, Artemisia ludoviciana, Dalea albiflora, Asclepias linifolia, Fran-
seria tenuifolia, Baccharis thesioides, Ayenia microphylla, and Anisolotus
argensis. The commonest summer ephemerals are Eriogonum abertia-
num and Eriocarpum gracile.

The flood-plains and the banks and beds of the arroyos in the Upper
Desert are, in general, more like the arroyos of the desert in their
vegetation than like those of the Encinal region. The largest tribu-
taries of Sabino Canon are somewhat less rich in aquatic and palustrine
plants than the lower portion of the canon, and merely because of
their steeper gradient and less regular flow. The forest which occupies
the flood-plains of Sabino Basin is chiefly made up of Quercus emoryi,
Quercus arizonica, Platanus wrightii, and Cupressus arizonica. The
smaller flood-plains and arroyos of the Upper Desert have few of these
trees but occasional individuals of Populus and open thickets of Bac-
charis emoryi and Baccharis sarothroides, together with Franseria
ambrosioides, Ingenhousia triloba, Erythrina flabelliformis, Croton texen-
sis, Calliandra eriophylla, Brickellia californica, Gymnosperma corym-
bosa, Amorpha californica, Bouvardia triphylla, and Stachys coccinea
(see plate IOA).

THE ENCINAL REGION.

Some of the distinctive species of the Lower Encinal are found at 4,000
feet and other forms, characteristic of the Upper Encinal, extend upward
into the Forest Region as far as 8,000 to 8,600 feet. The Lower Encinal
may be said to have its commencement, however, in the open orchard-
like stands of Quercus oblongifolia and Quercus arizonica, which occupy
northerly slopes at about 4,300 feet. At an approximate elevation of
5,000 feet the open Encinal may be found on all slopes except the steepest
southerly ones, while on steep northern slopes it already forms nearly
closed stands. The dense stands of the Upper Encinal region begin to
appear on southerly slopes at about 5,800 feet, and persist to the eleva-
tion of 6,200 to 6,400 feet, where large pines begin to dominate the
physiognomy and the true forest may be said to begin.

The activities of growth and flowering which are so conspicuous on
the Desert in the season of winter rains are very much reduced in the
Lower Encinal and are practically absent in the Upper Encinal.
Leaves are retained by the evergreen oaks and the sclerophyllous shrubs
throughout the winter and are shed in April or May, simultaneously
with the first growth of shoots and the renewal of foliage. Extremely
few of the ephemerals which often carpet the Desert in January are
to be found in the Encinal region. There is some activity on the part
of root perennials in the Lower Encinal during the months of March
and April, and flowers may be found on species of Sphoeralcea, Calo-
chortus, Verbena, Pentstemon, Eriogonum, and Lesquerella. Such activ-
ity is commonly stopped by the advent of the arid fore-summer, and

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 25

relatively little activity is observable in May and June, at least away
from arroyos and springs. In the Upper Encinal the early months of
spring partake of the rest which is then predominant in the Forest
region. The deciduous trees begin foliation in late April or early May
and some of the root perennials are not far behind them in their earliest
activity. The duration of the arid fore-summer being slightly less in
the Upper Encinal than in the Lower Encinal, and its intensity being
also less, there is not so decided a break, among the herbaceous peren-
nials between the first activity of spring and that of the humid mid-
summer, as there is in the Lower Encinal and the Desert.

THE LOWER ENCINAL.

The species which chiefly characterize the Lower Encinal at its
desert edge have already been mentioned : Quercus oblongifolia, Quer-
cus arizonica, Juniperus, Vauquelinia, Dasylirion, Nolina, Yucca macro-
carpa, Arctostaphylos pungens, Agave palmeri, and Agave schottii. All
of these are much more abundant at 5,000 feet than at 4,500 except
Quercus oblongifolia, which is a tree of very narrow vertical range, rarely
occurring above 5,200 feet and reaching its limit at 5,600 feet on steep
south slopes. At 5,000 feet the Encinal has been augmented by the
appearance of the common trees Quercus emoryi and Pinus cembroides
(pinon) and by the shrubs Garrya wrightii, Mimosa biuncifera, Rhus
trilobata, and Rhamnus crocea var. pilosa (see plates HA, 15, and 16).

The only characteristic Desert species which persist throughout the
Lower Encinal are: Carnegiea gigantea, a single young individual of
which has been seen at 5,100 feet; Opuntia versicolor, which reaches
5,500 feet; Fouquieria and Echinocactus wislizeni, which reach 5,600
to 5,800 feet; and Mamillaria grahami, which ascends to 7,000 feet.
So far as known, no other plants occurring on the Bajadas or in any
of the other non-palustrine desert habitats range to elevations above
6,000 feet.* There are at least a few species found in canons and near
constant water which range from the elevation of the Desert to more
than 6,000 feet. Several of the typical desert genera are represented
at higher elevations by species which seldom range as low as the Upper
Desert region. Two species of Opuntia (platopuntias) are found
throughout the Encinal, growing in thin soil or on rocks, and reaching
their highest occurrence solely on ridges or upper slopes. One of these
species has been found on a sharp rocky ridge at 7,200 feet, which is
the highest known occurrence of a platopuntia in the Santa Catalinas.
Mamillaria arizonica ranges from the Upper Desert to nearly 7,000
feet; Echinocereus polyacanthos ranges from about 5,000 feet to 7,800
feet, which is the highest elevation at which any cactus has been found
in these mountains. Agave palmeri and Yucca schottii are also fre-

* This statement is made only with respect to the Santa Catalinas. The influence of the
character of the underlying rock and of the elevation of the surrounding desert each serves to
determine indirectly the vertical limits of desert species.

26 VEGETATION OF A DESERT MOUNTAIN RANGE.

quently found up to elevations of 7,000 to 7,200 feet, and the latter
reaches its uppermost limit at 7,800 feet (see p. 30).

The ground cover of low perennial plants, grasses, succulents, and
herbaceous species which has been mentioned as characterizing the Upper
Desert is likewise to be found throughout the Lower Encinal, but does
not form as close a carpet in the latter region as it does in the former (see
plates 12, 15, and 16). Throughout the year this irregular carpet does
much to lend character to the landscape, varying but little in its density
with the alternating seasons of vegetative activity and of drought rest.
The scattered polsters of Chrysoma laricifolia are green at all seasons, and
there is no change in the gray-green foliage of Eriogonum wrightii nor
in the white tomentose leaves of Artemisia ludoviciana. The perennial
grasses, many of the other perennial herbaceous plants, and all of the
ephemerals are either in a resting state or dead throughout the arid
fore-summer and the arid after-summer, but the only change which
their rest or death registers in the landscape is a change of its color tone
from a greenish gray to an almost uniform gray and grayish brown.

All of the low shrubs and root perennials which were mentioned as
characteristic of the Upper Desert are to be found occasionally or
commonly in the Lower Encinal, excepting Franseria tenuifolia and
Ayenia microphylla. The winter and spring ephemerals are extremely
few at 4,500 to 5,000 feet, but there is much activity of growth and
much blooming among the root perennials and low shrubs during the
months of February and March, and sometimes during the early part
of April. The humid mid-summer is a season of even greater activity
on the part of the smaller elements of the vegetation. Relatively few of
the conspicuous herbaceous plants which are active at 5,000 feet in the
mid-summer have extended upward from thebajada, and the number of
summer ephemeral species is very small as compared with the Desert.

Among the small shrubs, root perennials and other herbaceous plants
which are common during the humid mid-summer at 4,500 to 5,500
feet, in the Lower Encinal, may be mentioned :

Bacchans pteronoidcs.
Baccharis thesioides.
Bouteloua hirsuta.
Bouteloua rolhrockii.
Castilleja integra.
Cordylanthus ivrightii.
Crotalaria lupulina.
Dalea albiflora.
Dalea wislizeni.
Eriocarpum gracile.

Eriogonum pharnaceoides.
Euphorbia heterophylla.
Gilia multiflora.
Gnaphalium wrightii.
Hymenothrix wrightii.
Linutn neomexicanum.
Muhlenbergia gracillima.
Pappophorum wrightii.
Pentslemon palmeri.
Phascolus wrightii.

On the flood-plains and along the streamways of the Lower Encinal
may be found a greater number of individuals of the evergreen oaks
than on the surrounding slopes (see plate 10s), and also Juglans major,
Platanus wrightii, and Populus sp., not to mention the restricted occur-
rence of Cupressus arizonica. Shrubs occasionally found along the

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 27

arroyos are Rhus trilobata, Baccharis emoryi, Erythrina flabelliformis,
Bouvardia triphylla, Amorpha californica, Fendlera rupicola, Morus
celtidi folia, and the climber Vitis arizonica.

On the flood-plains of the larger canons in the Lower Encinal may
be found the lowest examples of several species which become common
in the forested region of the mountain, and these are indeed the lowest
members of the forest flora, if aquatics are excepted. At 4,900 feet
Ceanothus fendleri and Prunus virens are both to be found, growing
not only on a flood-plain but in the shade of evergreen oaks. Berberis
wilcoxii is found at 5,200 feet growing in shade near a constant spring,
and Rhamnus ursina is infrequent at 5,000 feet near streamways.

During the mid-summer there is an abundant stand of herbaceous
perennials and ephemerals on the flood-plains of the Lower Encinal,
giving them a much closer carpet of vegetation than is to be found on
the adjacent slopes. Abundant and characteristic among them are:

Artemisia sp.
Asclepias luberosa.
Brickellia californica.
Castilleja inlegra.
ChamcEcrista leptadenia.
Comandra pallida.
Cordylanthus wrightii.
Crotalaria hipulina.
Diodia teres.
Eriocarpum gracile.

Euphorbia crenulata.

Gymnolomia multiflora.

Hymenothrix wrightii.

Malvastrum sp.

Monarda pectinata.

Solanum douglasii.

Solidago sparsiflora var. subcinerea.

Sporobolus confusus.

Stachys coccinea.

Stenophyllus capillaris.

THE UPPER ENCINAL.

During the ascent from 5,000 to 6,000 feet the most notable change
in the vegetation is the gradual increase in the density of the stand of
evergreen trees and shrubs (see plates 18, 19, and 20), a change which
forms the chief distinction of the Upper Encinal from the Lower
Encinal. Quercus emoryi and Quercus arizonica are still the dominant
trees, while Pinus cembroides and Juniperus pachyphloea are somewhat
less common. Arctostaphylos pungens and Garrya wrightii are the most
common of the larger shrubs and Mimosa biuncifera of the smaller ones.
Dasylirion wheeleri, Nolina microcarpa, and Agave palmer i remain
abundant, at least on southern slopes, up to 6,000 feet and Agave
schottii remains common up to its upper limit at that elevation. With
the increasing abundance of the oaks, however, these semi-desert
species as well as the cacti become infrequent and are confined to the
summits of ridges and the crevices of rocks.

On steep north slopes, between 5,300 and 6,000 feet, many almost
pure stands of Pinus cembroides are to be found and also the lowest
individuals of Quercus reticulata, here a low-branched tree of 20 feet
in height. Pinus chihuahuana first appears at about 5,900 feet on
south slopes, being the only one of the trees which is not found at
much lower elevations on north slopes than on south ones — indeed it
is not common on north slopes at any elevation.

28 VEGETATION OF A DESERT MOUNTAIN RANGE.

The heaviest stands of the Upper Encinal constitute a relatively
dense thicket in which the trees are from 18 to 30 feet in height and
so closely placed that it is very difficult for a mounted man to make
his way among them. This is partly due to the fact that the oaks, the
juniper, and the pinion all branch freely from a point near the ground,
and partly to the size and hemispherical habit of Arctostaphylos, in
which many of the stiff branches are placed in a nearly horizontal
position near the ground. These dense stands of the Upper Encinal,
between 5,600 and 6,200 feet, are made up of the same species that
form the very open Lower Encinal in so far as concerns the trees, shrubs
and larger perennials. There are, however, many root-perennial her-
baceous plants in the Upper Encinal which are not to be found below
5,500 feet, nearly all of which extend upward into the lower portions
of the Forest region.

Quercus emoryi is still a common tree at 5,600 feet, Quercus arizonica
is replaced by the closely similar Quercus reticulata, Quercus hypoleuca
makes its first appearance, and Juniperus pachyphlcea and Pinus cem-
broides reach their maximum abundance between 5,500 and 6,500 feet.
Dasylirion, Nolina, and Yucca are still conspicuous elements of the
vegetation even in the most dense stands of oaks, but Agave schottii
is no longer found and Agave palmeri, like the cacti, is found only on
ridges and rocks. A common tree of the lower forest region, Arbutus
arizonica, is first found in the Upper Encinal, where its isolated indi-
viduals are conspicuously different from the oaks. The only trees of the
mountain, excepting the desert species, the ranges of which lie wholly
below the Upper Encinal, are Vauquelinia calif ornica and Quercus oblong-
ifolia, while Quercus arizonica reaches its upper limit in this region.

Phoradendron calif ornicum, the mistletoe, which so commonly infests
Prosopis and the other trees of the desert, is found throughout the
Desert region of the mountains, while in the Encinal Phoradendron
flavescens var. villosum is found on several hosts and Phoradendron
juniperinum is extremely common on Juniperus, but does not extend
with it to its highest occurrences.

The vegetation of the Upper Encinal is extremely poor in shrubs
of the type so common in the Upper Desert and still frequent in
the Lower Encinal. In the open spots there may be found a few
individuals of Artemisia ludoviciana, Parosela wislizeni, Anisolotus
argensis, and other dwarf shrubs of the Lower Encinal, while in the
shade of the heaviest stands of oaks are to be seen Pteris aquilina var.
pubescens, Muhlenbergia affinis, Polygala alba, Comandra pallida,
Hymenopappus mexicanus, Cordylanthus wrightii, Chenopodium fre-
montii, and other species of root-perennials. The vegetation of rocks
and exposed ridges is still suggestive of the desert, both in its physiog-
nomy and in its phyletic relationships. In the crevices of rocks, where
the amount of soil is extremely scant and the supply of moisture must

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 29

be very uncertain, Laphamia lemmoni is found, a small composite
known only from the Encinal region and only from this habitat. In
crevices more favorably situated with respect to moisture may be found
Heuchera sanguinea, and on north slopes at 6,000 feet, in moist crevices,
may be found the lowest colonies of Saxifraga eriophora, a plant which
occurs infrequently up to the summit of the mountain.

Beds of Selaginella rupincola are still to be found at 6,000 feet and
the several species of drought-resistant ferns, which are confined to the
shade of rocks at lower elevations, are common on the floor of the heavy
stands of Pinus cembroides, or grow among the boulders in more open
situations. Among these the most common are : Cheilanthes fendleri,
Notholcena sinuata, Notholcena ferruginea, and Gymnopteris hispida. None
of these species extend upward into the Forest region (see plate 14) .

The drier flood-plains and arroyos of the Upper Encinal are charac-
terized by the same oaks and evergreen conifers that occur on the
adjacent slopes, while the moister streamways bear a number of decidu-
ous trees and shrubs, notably Juglans rupestris and Platanus wrightii,
extending upward from streamways at lower elevations, and Prunus
virens, Rhamnus ursina, Rhus trilobata, Robinia neomexicana, and Rhus
elegantula. Less frequent are Ceanothus fendleri, Berberis wilcoxii, and
Bouvardia triphylla, and Vitis arizonica is still common. Pinus chihua-
huana is not infrequent along the drier arroyos at the lower edge of its
range, and Cupressus arizonica is found along the streams and on the
lower slopes of Sabino and Bear Canons and some of their tributaries.

The commonest herbaceous perennials of the flood-plains of the
Upper Encinal are:

Apocynum sp.
Artemisia dracunculoides.
Asclepias tuberosa.
Carduus rothrockii.
Euphorbia crenulata.
Geranium caespitosum.
Gomphocarpus hypoleucus.
Gymnolornia multiflora.
Monarda pectinata nutt.
Muhlenbergia sp.

Oenothera sp.

Pentstemon torreyi.

Picradenia biennis.

Pteris aquilina var. pubescens.

Rubus oligospermus.

Scnecio neomexicanus.

Solidago sparsiflora var. subcinerea.

Sporobolus confusus.

Thalidrum fendleri var. wrightii.

Zauschneria californica.

THE FOREST REGION.

One of the most striking changes encountered in the vegetational
gradient of the Santa Catalinas is that from the closed and relatively
low Encinal to the open forest of Pinus arizonica, with trees 50 to 60
feet in height. This pine, the Arizona yellow pine, is closely related
to Pinus ponderosa, the western yellow pine, and is the common tree
of the forested altitudes of the mountain, extending upward on south-
erly slopes to the summit of Mount Lemmon. The lowest stands of
pine which possess sufficient density to be regarded as forest occur on
northerly slopes at 5,800 to 6,000 feet, or on southerly slopes at 6,000
to 6,400 feet, the limits depending in each particular locality upon the

30 VEGETATION OF A DESERT MOUNTAIN RANGE.

steepness of the slope and its soil characteristics, particularly with
respect to the soil moisture supply (see plate 21).

Much more gradual and inconspicuous is the transition from the
Pine Forest to that in which Abies concolor (white fir) is the dominant
tree. This type of Forest occupies the northern slopes of the highest
summits and ridges of the range from 7,500 feet upward, but there are
no elevations in the Santa Catalinas sufficiently great to bring the Fir
Forest onto the south slopes.

Throughout the Pine Forest there are trees, shrubs, and herbaceous
plants which may be found in the Encinal, at least in its upper portion,
but only in the lowest edge of the Pine Forest may plants be found
which suggest the genera or vegetation types characteristic of the desert.
A single cactus (Echinocereus polyacanthos), a Yucca, and an Agave are
the sole representatives of the succulent and semi-succulent forms of
the lower elevations, and they are rare above 7,000 feet and absent above
7,800 feet.

The Pine Forest is not, however, without vegetational features which
suggest the effects of a climate not far removed in character from that
of the desert. The openness of the lowest stands of Pinus arizonica,
the high mortality among the seedlings of the pine, the character of
the foliage of the shrubs and herbaceous perennials, and the deep-seated
root systems of the latter plants, all point to the existence of a pre-
carious soil-moisture supply and to atmospheric conditions conducive
to active transpiration. In the Fir Forest none of these features is
observable, and the vegetation as a whole presents a much more
mesophilous aspect.

In the Forest region the winter is a season of almost absolute rest,
save for the photosynthetic activity which is doubtless carried on by
the conifers, and possibly by the evergreen oaks and shrubs. The
deciduous trees and shrubs are leafless from early or mid October until
April or May, and only a few herbaceous perennials are active during
this period, such as the evergreen species of Pyrola and the early vernal
plants, such as Frasera. The amount of activity on the part of the
perennial herbaceous plants during the arid fore-summer is largely
dependent on the amount of winter precipitation and the date of its
termination. In the lower portion of the Pine Forest it often happens
that almost all activity is in abeyance until the first rains of the humid
mid-summer, while in the upper Pine Forest and in the Fir Forest it
is always possible to find a majority of the common herbaceous plants
in activity in May and June. There is a notable scarcity of annual
plants above 6,000 feet, and the only ones that have been detected in
the Forest region are:

Androsace arizonica.
Bidens sp.
Cerastium sericeum.
Dalea polygonoides.

Drymaria sperguloides.
Drymeria tenella.
Muhlenbergia sp.

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 31

THE PINE FOREST.

In the lowest stands of Pine Forest many of the dominant Encinal
forms are still to be found, but in no case do the evergreen oaks fail
to become more and more scattered in occurrence as the forest of pines
becomes more dense. Quercus emoryi and Pinus cembroides are scarcely
concerned in the overlapping of the Chaparral and Forest, as the former
reaches its upper limit at 6,300 feet, while the latter becomes confined
to the rocky non-forested or lightly forested ridges at about the same
elevation, although it persists as a rare shrub to an elevation of 7,800
feet. Arctostaphylos and Garry a are likewise of infrequent occurrence
in stands of forest. The oaks which are characteristic of the closed
forest are Quercus reticulata and Quercus hypoleuca. The former is com-
monly a low-branching shrub which often forms thickets on the steep
slopes of the highest peaks, where it extends upward to about 8,600 feet.
The latter oak is a shrub near its lower and upper limits at 6,000 and
8,500 feet respectively, but attains a height of 40 feet and a girth of
3 to 4 feet between 6,500 and 7,500 feet. Juniperus pachyphlcea is of
occasional occurrence in the Forest up to 7,900 feet, and Arbutus
arizonica (Arizona madrona), at first infrequent, becomes common at
7,000 to 7,500 feet and reaches its upper limit at 7,800 to 8,000 feet.

The composition of the Forest itself is extremely simple from its
lower limit around 6,000 feet to 7,500 feet, and above that elevation
is equally simple on southerly slopes up to the summit of Mount
Lemmon. Pinus chihuahuana reaches its limit at about 6,700 feet and
forms a very inconsiderable portion of the forest throughout the upper-
most 500 feet of its vertical range. Pseudotsuga mucronata begins to
occur on steep northerly slopes at 6,100 feet and Pinus strobiformis
(Mexican white pine) at 6,800 to 7,000 feet, but neither begins to
affect the composition of the Forest in general until higher elevations
are reached. At 6,000 feet the streamways and flood-plains are char-
acterized by several deciduous trees in addition to the pines themselves.
Platanus wrightii is near its upper limit at this elevation, Juglans
rupestris, Prunus virens, and Acer interior are of frequent occurrence,
while at 6,500 to 6,800 feet are found the lowest individuals of Quercus
submollis and Alnus acuminata.

Throughout the Pine Forest are to be found a large number of her-
baceous perennials, a few of which occur in the Upper Encinal, the great
majority of which, however, accompany the closed stands of pine, with
additions and eliminations with increasing altitude. In addition to these
plants is another large group which is confined in occurrence to the near
proximity of streams and streamways; some of the members of the group
being thus restricted in occurrence at lower altitudes, while they are of
more general occurrence on heavily wooded slopes at higher elevations.

In the clear park-like stretches of Pine Forest where no evergreen
oaks happen to occur, the most conspicuous plants on the forest floor

32

VEGETATION OF A DESERT MOUNTAIN RANGE.

are the low thorny shrub Ceanothus fendleri, or varieties of it, and the
bunch-grass Muhlenbergia virescens. The commonest of the herbaceous
perennials are low, small plants such as Hedeoma hyssopifolia, Hous-
tonia wrightii, POOL fendleriana, Calliandra reticulata, and Calliandra
humilis, or else they are somewhat taller but relatively inconspicuous,
as Pseudocymopterus montanus var. tenuifolius, Erigeron neomexicanus,
Litkospermum muliiflorum, Lotus puberulus, and others. In the dense
shade of the Upper Encinal Pteris aquilina var. pubescens is common,
and it again becomes common in the pines above 7,500 feet, but is infre-
quent in the lower portion of the forest region.

The Pine Forest gives the impression of possessing a much richer
flora of herbaceous plants than is found in any other habitat of the
mountain. This impression is due to the fact that a large number of
species enter into the vegetation as very common components of it.
As there are almost no rare or infrequent species to be found in the
Pine Forest away from streams and springs, the total flora involved
is not so great as might be supposed on first examination. Following
is a list of the characteristic species found between 7,000 and 8,000
feet, the relative abundance of which is indicated by asterisks:

Characteristic Herbaceous Plants of the Pine Forest.

***

Achillea lanulosa.

Agastache pallidiflora.

Anisolotus puberulus.

Antennaria marginata.

Anthericum torreyi.

Apocynum scopulorum.

Bidens sp.

Brickellia grandiflora.

Calliandra reticulata.

Carpochoete bigelovii.

Castilleja gloriosa.

Cologania longifolia.

Commelina dianthifolia.

Desmodium arizonicum.

Desmodium grahami.

Dugaldia hoopesii.

Erigeron macranthus.

Erigeron neomexicanus.

Eupatorium arizonicum.

Eupatorium pauperculum.

Eupatorium rothrockii.

Geranium coespitosum.

Gilia thurberi.
** Gnaphalium decurrens.
** Gnaphalium wrightii.

Gomphocarpus hypoleucus.

Hedeoma hyssopifolia.

Hieracium discolor.

Houstonia wrightii.

Hymenopappus mexicanus.

Ipomoea muricata.

Koeleria cristata.

**

***

**
***

**
***

***

**

***

***

***

**
**

***

***

***

***

* Lathyrus graminifolius.
*** Lithospermum multiflorum.
*** Lupinus sp.

Microstylis montana.
Monarda pectinala.
** Monarda scabra.
*** Muhlenbergia virescens.

* Muhlenbergia sp.
Oenothera hookeri.
Onosmodium thurberi.
Panicum bulbosum.

** Pentstemon torreyi.
** Perityle coronopifolia.
** Phaseolus retusus.

* Pinaropappus foliosus.
Poa fendleriana.
Potentilla subviscosa.
Pseudocymopterus montanus

tenuifolius.
Pseudocymopterus montanus

purpureus.

Pteris aquilina var. pubescens.
Salvia arizonica.
Senecio neomexicanus.
Solidago bigelovii.
Solidago marshallii.
Stevia sp.

Tradescantia pinetorum.
Trifolium pinetorum.
Vicia americana.
Woodsia mexicana.

***

**
**

**

***
***

***

***

***
***
***

4- T ••

**

var.

var.

VEGETATION OF THE SANTA CATALINA MOUNTAINS. 33

The pure or nearly pure stands of Pinus arizonica which occur be-
tween 8,000 and 9,000 feet are increasingly poor in the evergreen oak
shrubs, which have disappeared at the latter altitude. The clumps of
young Quercus submollis give the forest its only deciduous element at
this altitude, and the low patches of Ceanothus, so common at 8,000
feet, give way at 9,000 feet to Symphoricarpos oreophilus and to the
much less frequent Holodiscus dumosus. Very many of the commonest
herbaceous perennials of the Pine Forests which lie between 7,000 and
8,000 feet do not reach 9,000 feet, or are replaced in the physiognomy
of the forest by closely related species. On the summit and southern
slopes of Mount Lemmon the commonest herbaceous plants are:
Kceleria cristata, Dugaldia hoopesii, Erigeron neomexicanum, Pteris
aquilina var. pubescens, Gnaphalium decurrens, Hieracium lemmoni,
Senecio sp., Antennaria marginata, Silene greggii, and Helianthella
arizonica. Throughout the higher Pine Forest Arceuihobium divari-
catum and Arceuihobium robustum are common on the trunks and limbs
of Pinus arizonica.

At about 6,800 feet Alnus acuminata, Acer interior, and Quercus
submollis become frequent along streams (see plates 30 and 31). The
first two are confined to this habitat throughout their vertical range,
while the oak, which is the only deciduous member of the genus in
these mountains, is found even in some of the driest situations above
7,600 feet. Quercus submollis occurs characteristically either as single
trees of considerable size, up to 40 feet in height and 4 feet in girth,
or else as crowded circumscribed groups of young trees, which doubt-
less owe their juxtaposition to the accidents of seed dispersal. Salix
taxifolia is also a common streamside shrub above 6,800 feet, and
in certain portions of the mountain Rosa fendleri is abundant in the
proximity of streams.

Herbaceous plants are to be found in increasing numbers at or near
the banks of streams between 6,000 and 7,400 feet. Prominent among
them are: Juncus arizonicus, Aquilegia chrysantha, Thalictrum fendleri
var. wrightii, Scrophularia sp., Trifolium pinetorum, Fragaria ovalis,
Potentilla thurberi, Hypericum formosum, Lobelia gruina, Agrimonia
brittoniana var. occidentalis, Gaura suffulta, and Tagetes lemmoni.

THE FIR FOREST.

Between 7,000 and 7,400 feet is a rapid change in the character of
the forest stands on northerly slopes, due to the increasing occurrence
of Pseudotsuga mucronata and Pinus strobiformis, the lower limits of
which have already been mentioned, and to the appearance of Abies
concolor. These three species occur in mixed stands together with
Pinus arizonica on northerly slopes up to about 7,500 feet, above which
elevation the latter becomes a very infrequent tree on slopes facing
directly north, although it still occurs in admixture with Pseudotsuga

34 VEGETATION OF A DESERT MOUNTAIN RANGE.

and Abies at 9,000 feet on eastern and western exposures. Above 7,500
feet Pinus strobiformis ceases to be confined to the proximity of streams,
and occurs in admixture with Pseudotsuga and Abies, but is not so
common as they in the heaviest stands of this type of forest. On the
north slopes of Mount Lemmon is a small colony of Abies arizonica.
which is not known from any other locality on the mountain.

Slopes of due south or southwestern exposure are held by Pinus
arizonica up to the summit of Mount Lemmon at 9,150 feet, with a
slight occurrence of Pseudotsuga and Pinus strobiformis above 8,000
feet. The Pseudotsuga and Abies forest is found in fine development
at 7,500 feet on steep north exposures, and reaches its maximum devel-
opment in stature and size of the trees on the north slopes of Mount
Lemmon at 8,500 to 9,100 feet (see plates 1 and 35). The altitude of
the Santa Catalina Mountains is nowhere sufficient to admit of the
occurrence of extended bodies of such forest, nor of their existence on
southerly slopes.

In the Fir Forest the last relicts of the Encinal have disappeared :
Quercus hypoleuca, Quercus reticulata, and Juniperus pachyphlcea are
nowhere to be found in association with Pseudotsuga and Abies,
although they may grow very near them on opposed slopes. Arbutus
arizonica, which is more common in the Pine Forest than in the Encinal,
is likewise absent from the Fir Forest. The deciduous Quercus submollis
and the widely distributed Populus tremuloides are the commonest
of the subordinate trees, the latter often becoming dominant over
areas of an acre or more in extent, where it ultimately gives way to
conifers.

The floor of the Fir Forest is much more heavily and continuously
shaded than that of the densest stands of pine, a circumstance which
is of great importance in determining the nature of the forest reproduc-
tion and also in conditioning the character of the shrubby and herba-
ceous vegetation. The dense shade, the heavy litter, and the high humus
content of the soil tend to preserve its moisture throughout the arid
fore-summer (see p. 61), so that the seedling trees and other plants of
these situations are very far removed from the desiccating influences
which are operative in the open Pine Forest. The Fir Forest near the
summits of ridges is somewhat more open than that which is found on
middle and lower slopes, and this difference is accompanied by a dis-
similarity in the herbaceous flora of upper and lower slopes. On the
latter may frequently be found communities of plants which differ little
in their specific make-up from the communities which occupy flood-
plains, although they are much less dense.

The heaviest stands of Abies and Pseudotsuga, like most heavy conif-
erous forests, are relatively poor in both shrubs and herbaceous plants.
A few of the shrubs common to the water-courses are to be found also
in the Fir Forest, such as Jamesia americana, Symphoricarpos oreo-

VEGETATION OF THE SANTA CATALINA MOUNTAINS.

35

philus, Ribes pinetorum, and Rubus neomexicanus. The trifoliate
maple, Acer glabrum, also occurs locally on the north slopes of Mount
Lemmon. The poverty in the stand of herbaceous species on the floor
of the Fir Forest is contrasted with the large number of species to be
found, which is probably not so great, however, as the number charac-
teristic of the open Pine Forest. Most common are: Bromus richard-
sonii, Cystopterisfragilis, Geranium ccespitosum, Frasera speciosa, Thalic-
trum fendleri var. wrightii, Galium asperrimum, Smilacina sessilifolia,
Osmorhiza nuda, Disporum trachycarpum, Viola canadensis var. rydbergii,
Oxalis metcalfii, Fragaria ovalis, Trifolium rusbyi, and Draba helleriana.
On Abies the parasitic Phoradendron bolleanum is not infrequent.

The banks of constant and intermittent streams and the narrow
flood-plains of the Fir Forest region form a series of habitats with
closely similar physical conditions and with nearly identical vegetation.
In them are to be found a greater abundance and variety of trees and
shrubs than occur in topographically analogous habitats at lower ele-
vations. Abies, Pscudotsuga, Pinus strobiformis, and even Pinus
arizonica occur in this habitat. Its commonest woody plants, however,
are those which do not occur in other situations, as Alnus acuminata,
Acer interior, Acer brachypterum, Salix scouleriana, Salix exigua, Salix
taxifolia, Sorbus dumosa, Cornus stolonifera var. riparia, Jamesia ameri-
cana, Sambucus vestita, Symphoricarpos oreophilus, Rubus arizonicus,
Ribes pinetorum, and Salix sp.

In this same series of habitats, which are the most elevated of the
moist habitats of the mountain, is the most dense stand of herbaceous
vegetation that occurs on the Santa Catalinas. This vegetation is rich
in species and varies in its make-up from place to place according to
the amount of soil moisture present and according to the openness or
shade. In the following list are given the characteristic plants of these
situations. The two species of Mimulus are the only plants invariably
confined to the immediate proximity of water. Such plants as Dugaldia
a,udAgrimonia, on the other hand, are found only in the unshaded flood-
plains. A comparison of this list with that just given for the floor of the
Fir Forest will show that the latter habitat has few distinctive species.

Characteristic Herbaceous Plants of Flood-Plains, Stream Banks, and Lower Slopes in the

Fir Forest.

**

* Aconitum columbianum.
Aclcea viridiflora.

*** Agrimonia brittoniana var. occiden-

talis.
** Agrostis scabra var. subrepens.

* Aralia humilis.

** Aspidium filix-mas.
*** Bromus richardsonii.
*** Car ex sp.
Car ex sp.

** Cerastium sericeum.

* Delphinium scopulorum.

**

** Disporum trachycarpum.

** Draba helleriana.
*** Dugaldia hoopesii.

** Epilobium novomexicanum.

** Equisetum robustum.
*** Frasera speciosa.
*** Galium asperrimum.
* Gentiana microcalyx.
*** Geranium ccespitosum.

** Glyceria nervata.

** Gyrostachys sp.

** Heracleum lanatum.

36

VEGETATION OF A DESERT MOUNTAIN RANGE.

** Humulus lupulus var. neomexicanus.
*** Hypericurn formosum.
*** Juncus brunnescens.
*** Juncus interior.

** Limnorchis sparsiflora.
* Lister a sp.

** Microstylis porphyrea.
*** Mimulus cardinalis.
*** Mimulus guttalus.
*** Osmorhiza nuda.
*** Oxalis metcalfii.

** Oxalis wrightii.

* Polygonum douglasii.
** Pyrola chlorantha.
** Pyrola secunda.
** Rubus arizonicus.
** Rudbeckia laciniata.
"** Scrophularia sp.
"** Smilacina amplexicaulis.
** Smilacina sessilifolia.
Solanum fendleri.
Thalictrum fendleri var. wrightii.
Viola canadensis var. rydbergii.
Viola nephrophylla.

**

***

FLORA OF THE SANTA CATALINA MOUNTAINS.

The wide range of physical conditions embraced within the area of
the Santa Catalina Mountains gives them a relatively large flora, which
has been estimated by Professor J. J. Thornber to be about 1,500
species. Although the exploitation of this flora is not completed it is
nevertheless sufficiently well advanced to show that elements are pres-
ent which are common to each of many diverse regions lying north,
south, east, and west.

The desert at the foot of the mountains stands in unbroken connec-
tion with the deserts of Sonora and Sinaloa. The Encinal and Forest
regions, on the other hand, are isolated from other areas possessing the
same physical conditions. Areas of Encinal are numerous and near,
both on the low desert mountains and on the elevated plains of southern
Arizona; while bodies of forest are to be found only at greater distances
and more remotely separated from each other. The floristic history
of the Encinal and Forest regions of the Santa Catalinas is quite as
intimately bound up with the controlling influences of climatic con-
ditions as is the present limitation of the vegetation. In fact the floras
of the two isolated regions are a resultant between the physical con-
ditions which they have presented in the remote and recent past and
the operation of natural agencies of dispersal.

PHYTOGEOGRAPHIC RELATIONSHIPS OF THE FLORA.

It would not be within the scope of this paper to enter upon a detailed
discussion of the floristic relationships of the isolated mountain areas
of Encinal and Forest in southern Arizona, even if all the evidence
bearing on such a discussion were now in hand. It will be instructive,
however, to point out very briefly some of the principal floristic rela-
tionships of the Santa Catalinas in order to demonstrate the extensive
and diversified area over which members of its flora may be found.

THE DESERT FLORA.

The flora which occupies the bajadas of the Santa Cruz valley and
the lower slopes of the Santa Catalina Mountains derives many species
from each of two Mexican desert regions, the one lying at low elevations

FLORA OF THE SANTA CATALINA MOUNTAINS. 37

between the Sierra Madre Occidental and the Gulf of California, in the
States of Sonora and Sinaloa, the other lying at higher elevations in
the States of Chihuahua and Zacatecas. There are strong diversities
of flora between these two Mexican deserts, although they do not fail
to have many species in common. The Sierra Madre forms an effective
barrier between them in Mexico, but north of the International Bound-
ary the continental divide is formed by scattered mountain ranges and
broad valleys rather than by a continuous elevated range, and these
valleys, lying between 4,000 and 5,000 feet, have permitted the inter-
mingling of species from the two desert floras, at the same time that
they have constituted a barrier to many species presumably unable to
withstand the winter temperature conditions of the elevated valleys.
The deserts which border the lower course of the Colorado River in
Arizona and California, the Mojave Desert, and other desert regions
in southern California and Nevada lying below 4,000 feet, possess a
very small number of distinctive species as contrasted with the two
Mexican desert regions, and have contributed almost no species to the
flora of the Santa Cruz valley, although many species of wide Mexican
occurrence are represented in both localities. The deserts of the Great
Basin have likewise contributed no distinctive elements to the flora
of the Santa Cruz Valley and the Desert region of the Santa Catalinas.

Among the many species characteristic of the Arizona-Sonora Desert
which do not cross the continental divide are: Carnegiea gigantea,
Parkinsonia microphylla, Encelia farinosa, Olneya tesota, Hyptis emoryi,
Franseria deltoidea, Simmondsia californica, Jatropha cardiophylla, and
Crossosoma bigelovii. Among the desert species which are common to
the Arizona-Sonora region and to the Texas-Chihuahua desert are:
Fouquieria splendens, Kceberlinia spinosa, Chilopsis saligna, Momisia
pallida, Coldenia canescens, Opuntia leptocaulis, Ephedra trifurca,
Hilaria mutica, and Bailey a multiradiata.

It would be possible to place perhaps 90 per cent of the desert flora
of southern Arizona in one or the other of the categories just mentioned.
There are a few local and endemic species, but very few species exhibit
ranges extending chiefly to the west, north, or east. Among the two
Mexican elements many species range far south of Mexico, as witness
the following, which are found in the deserts of Chile: Calandrinia
menziesii, Bowlesia lobata, Daucus pusillus, Parietaria debilis, and
Hydrocotyle ranunculoides. A large number of the genera found in the
Desert flora also possess representatives in the deserts of Argentine
and Chile, as: Covillea, Franseria, Encelia, Actinella, Kramer ia, Gutier-
rezia, Viguiera, Chorizanthe, Coldenia, Perezia, Menodora, Nama,
Amsinckia and many others. Other genera found in the Santa Cruz
Valley have many representatives in tropical South America or in the
West Indies, as Hyptis, Dodoncea, Erythrina, and Gymnolomia, or have
a world-wide representation, as Tragia and Stemodia.

38 VEGETATION OF A DESERT MOUNTAIN RANGE.

Enough has been said to show that both the specific and generic
relationships of the Desert flora are with the desert regions of Mexico,
the deserts of Argentine and Chile, and even with the moist tropical
regions of South America. The plants which dominate the Desert
landscape in southern Arizona are members of genera, or even of
species, which characterize a much greater area to the south than to
the north, and they are in the main members of genera which reach
their maximum development in number of species and in abundance
of individuals in similar desert regions. The plants of the Desert which
are of tropical relationship are usually the sole and northernmost repre-
sentatives of families or genera which are much more richly represented,
both in types and in individuals, in the tropical zone. These plants
are often so infrequent and inconspicuous as scarcely to interest the
student of vegetation, except for the fact that their seasonal behavior
and habitat relations are such as to give them the most moist conditions
which the Desert affords. Among them may be mentioned : Passiflora,
Stemodia, Maurandia, and Rivina.

The few members of genera of northern dominance, such as Populus
and Salix, or Anemone and Delphinium, are either to be sought in the
vicinity of streams and ponds, as is the case with the former two, or
are to be found in activity only in the late winter and early spring, as
is true of the latter two. The still fewer species of transcontinental
range are almost solely palustrine plants, as Cephalanthus occidentalis,
Scirpus americanus, Cyperus diandrus, and others, and are to be found
only in palustrine situations in Arizona.

THE ENCINAL FLORA.

The type of vegetation which is designated as Encinal in this paper
is found throughout southern Arizona and New Mexico at elevations
of 5,000 to 7,000 feet. It is pre-eminently a community of evergreen
oaks and nut pines, with many sclerophyllous shrubs. With many
floristic modifications this type of Encinal extends into western Texas,
Colorado, and inner California, usually as a belt connecting the treeless
plains or desert with the forested mountain tops. Encinal similar to
that of southern Arizona is found throughout the mountainous portions
of Sonora, Chihuahua, Sinaloa, and Zacatecas, and with many modi-
fications it extends still further south.

The dominant species of the Encinal of the Santa Catalinas range
far to the south along both sides of the Sierra Madre, whereas but few
of them range further north than the southern edge of the Mogollon
Plateau in central Arizona, and some of them not even so far as that.
The 14 commonest woody or semi-succulent perennials in the Encinal
of the Santa Catalinas are all plants of extended Sonoran and Chi-
huahuan distribution; all of them occur in southern New Mexico and
eight of them in western Texas. Only one of the plants reaches Cali-

FLORA OF THE SANTA CATALINA MOUNTAINS, 39

fornia and only one of them has been reported from Colorado. These
plants are:

Quercus oblongifolia.
Quercus arizonica.
Quercus emoryi, Tex.
Vauquelinia calif arnica.
Juniperus pachyphlcea, Tex.

Arctostaphylos pungens, Cal.
Garrya wrightii, Tex.
Dasylirion wheeleri, Tex.
Agave palmeri.
Nolina microcarpa.

Pinus cembroides, Tex.
Mimosa biuncifera, Tex.
Chrysoma laricifolia, Tex.
Eriogonum wrightii, Col., Tex.

The Encinal likewise comprises a number of plants which reach their
maximum occurrence on the Great Plains or else possess areas of dis-
tribution which are chiefly to the northeast of Arizona. Among these
are Bouteloua obligostachya, Bouteloua hirsuta, Bouteloua curtipendula,
Polygala alba, Artemisia ludoviciana, Artemisia dracunculoides, and
Stephanomeria runcinata.

The elements which are common to the flora of California are few,
as is true of the Desert, and are almost solely comprised in the follow-
ing: Zauschneria calif ornica, Amorpha calif ornica, Bouvardia triphylla,
and Brickellia californica, not to add Arctostaphylos pungens, which has
its maximum extension southward into Mexico. The Encinal contains
a number of forms which have been but recently segregated from well-
known species, among them Rhus racemulosa, Rhamnus ursina, and
Prunus virens. So little is known of the ranges of these species that
it is impossible to state in how far they may represent contributions
from distant floras or to what extent they represent forms that have
been differentiated in the Arizona-Sonora region.

The only northern element in the Encinal flora seems to be that
which has been mentioned as occurring also in the Great Plains, while
the mountainous regions of Colorado and Utah have contributed even
fewer species than has the Calif ornian region.

THE FOREST FLORA.

The Forest region of the Santa Catalinas possesses strong floristic
affinities both with the Mexican cordillera and with the Rocky Moun-
tains of Colorado and their southern extension in New Mexico. The
majority of the plants which take a conspicuous place in the vegetation
of the Forest are members of northern genera. Many of these members
are identical with Rocky Mountain species, while many others have
their chief range in the mountains of northern Mexico. There are also
representatives of a few genera which are distinctively Mexican, a few
species of northwestern relationship, and a few apparently of restricted
range in the desert mountains of Arizona and New Mexico.

As examples of the large Rocky Mountain contingent in the Forest
flora may be mentioned:

Abies concolor.
Pseudotsuga mucronata.
Disporum trachycarpum.
Salix scouleriana.
Populus tremuloides.

Acer glabrum.
Jamesia americana.
Symphoricarpos oreophilus.
Eraser a speciosa.
Dugaldia hoopesii.

Erigeron macranthus.
Heuchera rubescens.
Brickellia grandiflora.
Gilia thurberi.
Achillea lanulosa.

40 VEGETATION OF A DESERT MOUNTAIN RANGE.

Some of the members of this group are of wide distribution in the
north, as Populus tremuloides, Achillea lanulosa, and Disporum trachy-
carpum. In the northern mountain contingent are also a few species
which range eastward to the Atlantic coast, a few which are found at
least as far south as Maryland (Heracleum lanatum, Rudbeckia lacini-
tata, Apocynum androscemifolium, Vicia americana, and Asplenium tri-
chomanes), not to mention Achillea lanulosa, which scarcely deserves
separation from the cosmopolitan Achillea millefolium.

The relationship with northern California and the northwestern
states is weakly expressed in the occurrence of Salix lasiolepis and
Prunus emarginata. Genera characteristic of the sub-arctic regions are
sparingly represented at higher elevations by species of Primula, Saxi-
fraga, and Androsace.

Some of the most conspicuous components of the vegetation belong
to northern genera, but to species which are characteristic of the Mexi-
can cordillera, as Pinus arizonica, Pinus strobiformis, Alnus acuminata,
Salix bonplandiana, Quercus hypoleuca, and Quercus reticulata. Such
genera of herbaceous plants as Solidago, Eupatorium, Erigeron, Penlste-
mon, Mimulus, Potentilla, Gilia, and Gentiana — all of which are richly
developed in the Rocky Mountains — are chiefly represented in the
Santa Catalinas by species not found in Colorado nor Wyoming. The
extent to which these species are characteristic of the Arizona-New
Mexico region or are components of the flora of the higher Mexican
mountains is only partially known.

The relationship of the Forest flora to that of the extended mountain
regions to the south is still further strengthened by the occurrence of
members of genera which are not found in the Rocky Mountains of
Colorado and northern New Mexico, as Arbutus, Calliandra, Micro-
stylis, Drymaria, Cologania, Stevia, and Tagetes.

To summarize for the mountain as a whole, it may be said that the
floristic relationships of the Desert and Encinal regions are almost
wholly with the Mexican deserts and foothills to the south, while those
of the Forest region are divided between the Mexican Cordillera and
the Rocky Mountains. The Mexican group is the more conspicuous
in the make-up of the vegetation, while the Rocky Mountain contin-
gent is apparently preponderant in number of species.

It will be impossible to summarize the floristic relationships of the Santa
Catalinas in a thorough manner until very much more is known of their
own flora and also of the floras of the many adjacent mountain ranges and
desert valleys, both in the United States and in Mexico. For the explana-
tion of these relationships a closer acquaintance is needed with the
actual mechanisms of transport which are effective in the dispersal of the
seeds of desert and mountain plants. A fuller knowledge is also required
of the fluctuations of climate within recent geological time, and of the
consequent downward and upward movements of the Encinal and Forest
belts of all the southwestern mountains. Such movements would alter-

FLORA OF THE SANTA CATALINA MOUNTAINS.

41

nately establish and break the connections between the vegetations of
the various mountain ranges and elevated plains, thereby permitting the
dispersal and subsequent isolation of species which might find no means
of movement across the desert valleys under existing conditions.

LIST OF CHARACTERISTIC SPECIES.

The lack of a single taxonomic work covering the entire flora of the
Santa Catalina Mountains makes it desirable to bring together here a
list of the plant names which are used throughout this paper, together
with some of the commoner synonyms. The list comprises only those
plants which are common and characteristic components of the vege-
tation of some particular region or habitat of the mountain. The writer
wishes to express here his very great indebtedness to Professor J. J.
Thornber, of the University of Arizona, for determining numerous sets
of plants from the Santa Catalinas and for verifying the following list.

POLYPODIACE.E :

Aspidium filix-mas (L.) Sw.

= Dryopteris filix-mas (L.) Schott.

Asplenium trichomanes L.

Cheilanthes fendleri Hook.

Cheilanthes lindheimeri Hook.

Cheilanthes wrightii Hook.

Cystopteris fragilis (L.) Bernh.
= Filixfragilis (L.) Underw.

Gymnopteris hispida (Mett.) Underw.
= Gymnogramme hispida Mett.

Notholcena ferruginea (Desv.) Hook.

Notholcena hookeri D. C. Eaton.

Notholcena sinuata (Sw.) Kaulf.

Pellcea wrightiana Hook.

Pteris aquilina var. pubescens Underw.

Woodsia mexicana Fee.
EQUISETACE^E:

Equisetum robustum A. Br.
SELAGINELLACE^E :

Selaginella rupincola Underw.

Selaginella sp.

Abies arizonica Merriam.

Abies concolor Lindl. & Gord.

Cupressus arizonica Greene.

Juniperus pachyphlaea Torr.

Pinus arizonica Engclm.

Pinus cembroides Zucc.

Pinus chihuahuana Engelm.

Pinus strobiformis Engelm.

Pseudotsuga mucronala (Raf.) Sudw.

= Pseudotsuga taxifolia (Lam.) Britton.
GNETACE^E:

Ephedra trifurca Torr.
GEAMINE^E:

Agrostis scabra var. subrepens Hitchck.

Andropogon saccharoides Sw.

= Amphilophis saccharoides (Sw.) Nash.

Andropogon scoparium Michx.

Aristida americana var. bromides (H. B. K.)
Scribn. & Merr.

Aristida divergens Vasey.

Aristida scheidiana Trin. & Rupr.

Bouteloua aristidoides (Kunth) Griseb.

Bouteloua curtipendula (Michx.) Torr.

GRAMINE.E — Continued'-

Bouteloua hirsuta Lag.

Bouteloua oligostachya (Nutt.) Torr.

Bouteloua polystachya (Benth.) Torr.

Bouteloua rothrockii Vasey.

Bromus richardsonii Link.

Diplachne dubia (Nees) Benth.

Eragrostis lugens Nees.

Eragrostis neomexicana Vasey.

Eragrostis pilosa (L.) Beauv.

Heteropogon contortus (L.) Beauv.

Hilaria cenchroides H. K. B.

Hilaria mutica (Buckl.) Benth.

K(jeleria cristata (L.) Pers.

Leptochloa mucronala (Michx.) Kunth.

Muhlenbergia affinis Trin.

Muhlenbergia dumosa Scribn.

Muhlenbergia distichophylla (PresI) Munro.

Muhlenbergia gracillima Torr.

Muhlenbergia porteri Scribn.

Muhlenbergia vaseyana Scribn.

Muhlenbergia virescens (H. B. K.) Trin.

Muhlenbergia sp.

Panicularia nervata (Willd.) Kze.

Panicum bulbosum H. B. K.

Panicum bulbosum var. minor Vasey.

Panicum hallii Vasey.

Panicum hirticaulum Presl.

Papphorum wrightii Wats.

Poa fendleriana (Steud.) Vasey.

Sitanion elymoides Raf.

Sporobolus confusus (Fourn.) Vasey.

Stipa neomexicana (Thurb.) Scribn.
CYPERACE^E:

Car ex sp.

Car ex sp.

Cyperus fendlerianus Boeckl.

Cyperus inflexus Muhl.

Cyperus speciosus Vahl.

Eleocharis montana (H. B. K.) R. & S.

Fimbristylis sp.

Hcjnicarpha micrantha (Vahl) Britt.

Stenophyllus capillaris (L.) Britt.

COMMELINACE.E :

Commelina dianthifolia DC.
Tradescantia scopulorum Rose.
Tradescantia pinelorum Greene.

42

VEGETATION OF A DESERT MOUNTAIN RANGE.

List of Characteristic Species — Continued.

JUNCACE-ffi :

Juncus arizonicus Wieg.
Juncus brunnescens Rydb.
Juncus bufonius L.
Juncus interior Wieg.

Anthericum torreyi Baker.

Brodicea capitata var. pauciflora Wats.

Calochortus nuttallii T. & G.

Dasylirion wheeleri Wats.

Disporum trachycarpum (Wats.) B. & H.

Nolina microcarpa Wats.

Smilacina amplexicaulis Nutt.

= Vagnera amplexicaulis (Nutt.) Morong.
Smilacina sessilifolia Nutt.
Yucca macrocarpa (Torr.) Coville.
Yucca schottii Engelm.
AMARYLLIDACE^E :

Agave palmeri Engelm.
Agave parryi Engelm.
Agave schottii Engelm.

IRIDACE.E:

Sisyrinchium arizonicum Rothr.

= Oreolirion arizonicum (Rothr.) Bicknell.
ORCHIDACE.E::
Gyrostachys sp.

Limnorchis sparsiflora (Wats.) Rydb.
Lister a sp.
Microstylis corymbosa Wats.

= Achroanthes corymbosa (Wats.) Greene.
Microstylis monlana Rothr.

= Achroanthes montana (Rothr.) Greene.
Microstylis porphyrea Ridley.

= Achroanthes porphyrea (Ridley) Greene.

SALICACE^E :

Populus angustifolia James.
Populus tremuloides Michx.
Populus sp.

near to Populus wislizeni (Wats.) Sarg.
Salix bonplandiana H. B. K.
Salix exigua Nutt.
Salix scouleriana Barr.
Salix taxifolia H. B. K.
Salix wrightii Anders.
Salix sp.

-

Jugland major (Torr.) Hell.
BETULACE^Q:

Alnus acuminata H. B. K.

= Alnus oblongifolia Torr.
FAGACEJE :

Quercus arizonica Engelm.
Quercus emoryi Torr.
Quercus hypoleuca Engelm.
Quercus oblongifolia Torr.
Quercus reticulata Humb. & Bonpl.
Quercus submollis Rydb.
ULMACE.E :

Momisia pallida (Torr.) Planch.

= Celtis pallida Torr.
Celtis reticulata Torr.

= Celtis occidentalis var. reticulata (Torr.)

Sarg.
HORACES :

Humulus lupulus var. neomexicanus Nels. &

Cockrl.
Morus celtidifolia H. B. K.

SANTALACE-E :

Comandra pallida A. DC.

LORANTHACE^E :

Arceuthobium divaricatum Engelm.

= Razoumofskya divaricata (Engelm.) Kze.
Arceuthobium robustum Engelm.

= Razoumofskya robusta (Engelm.) Kze.
Phoradendron bolleanum Eichl.
Phoradendron californicum Nutt.
Phoradendron flavescens var. villosum

Engelm.
Phoradendron juniperinum Engelm.

POLYGONACE-E :

Chorizanthc bremcornu Torr.
Eriogonum abertianum Torr.
Eriogonum pharnaceoides Torr.
Eriogonum wrightii Torr.
Polygonum douglasii Greene.
Rumex hymenosepalus Torr.
CHENOPODIACE^E :

Chenopodium fremontii Wats.
AMARANTACE^J :

Amaranthus palmeri Wats.
Cladothrix lanuginosa Nutt.
Frcelichia floridana (Nutt.) Moq.
Gomphrena ccespitosa Torr.
Gomphrena nitida Rothr.
NYCTAQINACE.E :

Allionia gracillima Standley.
Boerhaavia pterocarpa Wats.
Boerhaavia watsoni Standley.
Wedelia incarnata (L.) Kze.
PORTULACACE.E :

Calandrinia menziesii (Hook.) T. & G.
Calyptridium monandrum Nutt.
Montia perfoliata (Bonn.) Howell.
Talinum patens var. sarmentosum (Engelm.)

Gray.
CARYOPHYLLACE^E :

Arenaria confusa Rydb.
Cerastium sericeum Wats.
Cerastium texanum Britt.
Drymaria sperguloides Gray.
Dryrnaria tenella Gray.
Silene laciniata var. greggii (Gray) Wats.
RANUNCULACE^E :

Aconitum columbianum Nutt.
Actcea viridiflora Greene.
Aquilegia chrysantha Gray.
Clematis ligusticifolia Nutt.
Myosurus cupulatus Wats.
Thalictrum fendleri var. wrightii Gray.

BERBERIDACE^E:

Berberis wilcoxii Kearney.

PAPAVERACEJE :

Eschscholtzia mexicana Greene.
Platystemon californicus Benth.

CRTJCIFER.E:

Draba helleriana Greene.
Draba spectabilis Greene.
Lepidium lasiocarpum Nutt.
Lesquerella gordoni (Gray) Wats.
Thelypodium linearifolium Gray.

CRASSULACEJB :

Sedum stelliforme Wats.
Tillcea erecta Hook. & Am.

FLORA OF THE SANTA CATALINA MOUNTAINS.

43

List of Characteristic Species — Continued.

SAXIFRAGACE.E :

Fendlera rupicola Engelm. & Gray.
Heuchera rubescens Torr.
Heuchera sanguinea Engelm.
Jamesia americana T. & G.

= Edwinia americana (T. & G.) Hell.
Ribes pinetorum Greene.
Saxifraga eriophora Wats.

= Micranthcs eriophora (Wats.) Small.

PLATANACE.E:

Platanus wrightii Wats.
CROSSOSOMATACE^E :

Crossosoma bigelovii Wats.

ROSACE^E:

Agrimonia brittoniana var. occidentalis Bick-

nell.

Cowania stansburiana Torr.
Fragaria ovalis (Lehm.) Rydb.
Holodiscus dumosus (Nutt.) Hell.
Potentilla thurberi Gray.
Potentilla subviscosa Greene.
Prunus virens (Woot. & Stand.)

= Padus virens Woot. & Stand.
Rosa fendleri Crepin.
Rubus arizonicus Greene.
Rubus neomexicanus Gray.
Rubus oligosperma Thornb.
Sorbus dumosa Greene.
Vauquelinia californica (Torr.) Sarg.

LEGUMINOS^E:

Acacia greggii Gray.

Acacia paucispina Wooton.

Acacia suffrutescens Rose.

Amorpha californica Nutt.

Anisolotus argensis Coville.

Anisolotus puberulus (Benth.) Woot. & Stand.

= Hosackia puberula Benth.
Anisolotus trispermus (Greene) Woot.& Stand.

= Lotus trispermus Greene.
Cassia covesii Gray.
Cassia leptadenia Greenm.

= Chamaecrista leptadenia (Greenm.)

Cockrl.

Cassia leplocarpa Benth.
Calliandra eriophylla Benth.
Calliandra reticulata Gray.
Calliandra humilis Benth.
Cologania longifolia Gray.
Crotolaria lupulina Raf .
Dalea albiflora Gray.

= Parosela albiflora (Gray) Vail.
Dalca parryi T. & G.

= Parosela parryi (T. & G.) Hell.
Dalea polygonoides Gray.

= Parosela polygonoides (Gray) Hell.
Dalea wislizeni Gray.

= Parosela wislizeni (Gray) Vail.
Desmodium arizonicum Wats.

= Meibornia arizonica (Wats.) Vail.
Desmodium bigelovii Gray.

= Meibornia bigelovii (Gray) Kze.
Desmodium grahami Gray.

= Meibomia grahami (Gray) K/e.
Desmodium psilocarpum Gray.

= Meibomia psilocarpa (Gray) Kze.
Erythrina flabelliformis Kearney.
Eysenhardtia orthocarpa (Gray) Wats.
Indigofera sphaerocarpa Gray.

LEGUMINOS.E — Continued:

Lathyrus graminifolius (Wats.) White.

Lupinus sp.

near to Lupinus palmeri Wats.

Mimosa biuncifera Benth.

Nissolia schottii (Torr.) Gray.

Parkinsonia microphylla Torr.

Parkinsonia torreyana Wats.

= Cercidium torreyanum (Wats.) Sarg.

Phaseolus retusus Benth.

Phaseolus wrightii Gray.

Prosopis velutina Wooton.

Robinia neomexicana Gray.

Trifolium pinetorum Greene.

Vicia americana Muhl.

Vicia melilotoides Woot. & Stand.
GERANIACE^E :

Geranium ccespitosum James.

OXALIDACE^E :

Oxalis albicans H. B. K.

= Ionoxalis albicans (H. B. K.) Small.
Oxalis metcalfii (Small).

= Ionoxalis metcalfii Small.
LINAGES :

Linum lewisii Pursh.
Linum neomexicanum Greene.
ZYGOPHYLLACE^E :

Covillea tridentata (DC.) Vail.

= Larrea tridentata (DC.) Coville.
RUTACE.E :

Ptelea cognata Greene.
MALPIGHIACE.E :
Janusia gracilis Gray.

PoLYGALACEjE :

Krameria glandulosa Rose & Painter.
Poly gala alba Nutt.

EUPHORBIACE^E :

Croton texensis (Klotsch) Muell. Arg.

Euphorbia crenulata Engelm.

Euphorbia fiorida Engelm.

Euphorbia heterophylla L.

Euphorbia melanadenia Torr.

Euphorbia pediculifera Engelm.

Jatropha angustidens Muell. Arg.

Jalropha cardiophylla (Torr.) Muel. Arg.

Manihot carthaginensis Muel. Arg.
CALLITRICHACE^E :

Callitriche sp.
BUXACE^E :

Simmondsia californica Nutt.
ANACARDIACE.E :

Rhus aromatica var. mollis (Gray) Ashe.

Rhus elegantula Greene.

Rhus rydbergii Small.

= Toxicodendron rydbergii (Small) Greene.

Rhus trilobata Nutt.
ACERACE^E:

Acer brachypterum Woot. & Stand.

Acer glabrum Torr.

Acer interior Britt.
SAPINDACE^E :

Dodonwa viscosa var. angustifolia. (L. f.)
Benth.

Sapindus drummondii Hook. & Arn.
RHAMNACE^E:

Ceanothus fendleri Gray.

Ceanothus fendleri var. venosus Trel.

Rhamnus crocea var. pilosa Trel.

Rhamnus ursina Greene.

44

VEGETATION OF A DESERT MOUNTAIN RANGE.

List of Characteristic
RHAMNACE.E— Continued :
Zizyphus lycioides var. canescens Gray.
= Condalia lycioides (Gray) Weberbaur.

Parthenocissus dumetorum var. laciniata

Rehder.

= Parthenocissus quinquefolia var. laciniata
Planch.

Vitis arizonica Engelm.
MALVACEAE :

Abutilon incanum (Link) Sweet.

Ingenhousia triloba DC.

= Thurberia thespesioides Gray.

Mahastrum sp.

Sphaeralcea pedata Torr.
STERCULIACE.E :

Ayenia microphylla Gray.
HYPERICACE^E:

Hypericum formosum H. B. K.
FOUQUIERACE.E :

Fouquieria splendens Engelm.
VIOLACE^: :

Viola canadensis var. rydbergii (Greene)
House.

Viola nephrophylla Greene.
LOASACE.E:

Mentzelia albicaulis Dougl.
CACTACE.E :

Carnegiea gigantea (Engelm.) Britt. & Rose.
= Cereus giganteus Engelm.

Echinocactus wislizeni Engelm.

Echinocereus fendleri (Engelm.) Rumpl.

Echinocereus polyacanthus Engelm.

Mamillaria arizonica Engelm.

Mamillaria grahami Engelm.

Opuntia bigelovii Engelm. & Bigel.

Opuntia blakeana Rose.

Opuntia engelmanni Salm Dyck.

Opuntia fulgida Engelm.

Opuntia Icevis Coult.

Opuntia leptocaulis DC.

Opuntia marnillata Schott.

Opuntia santa-rita (Griff. & Hare) Rose.

Opuntia spinosior (Engelm. & Bigel.)
Tourney.

Opuntia toumeyi Rose.

Opuntia versicolor Engelm.

Opuntia sp.

Opuntia sp.
ONAGRACE.E:

Epilobium novomexicanum Hausk.

Gaura suffulta Engelm.

Isnardia palustris L.

= Ludwigia palustris (L.) Ell.

(Enothera hookeri T. & G.

= 0nagra hookeri (T. & G.) Small.

(Enothera mexicana Spach.

Zauschneria californica Presl.
ARALIACE^E :

Aralia humilis Cav.
UMBELLIFER.E :

Daucus pusillus Michx.

Heracleum lanatum Michx.

Hydrocotyle ranunculoides L. f.

Osmorhiza nuda Torr.

= Washingtonia obtusa C. & R.

Pseudocymopterus monta?ius var. purpureus
C.&R.

Species — Continued.

UMBELLIFER^E — Continued:

Pseudocymopterus montanus var. tenuifolius

(Gray) C. & R.
CORNACEJE:

Cornus stolonifera var. riparia (Rydb.)

Garrya wrightii Torr.
ERICACEAE:

Arbutus arizonica (Gray) Sarg.

Arctostaphylos pringlei Parry.

Arctostaphylos pungens H. B. K.

Hypopitys sanguinea Hell.

Pterospora andromedea Nutt.

Pyrola chlorantha Sw.

Pyrola secunda L.
PKIMULACE^B :

Androsace diffusa Small.

Androsace arizonica Gray.

Primula rusbyi Greene.
OLEACE^J:

Fraxinus altenuata Jones.

Fraxinus toumeyi Britt.

Menodora scabra Gray.
APOCYNACE^E:

Apocynum androscemifolium L.

Apocynum scopulorum Greene.

Haplophyton cimicidium A. DC.
ASCLEPIADACE^; :

Asclepias linaria Cav.

Asclepias tuberosa L.

Gomphocarpus hypoleucus Gray.

CONVOLVULACEjE :

Evolvulus arizonicus Gray.

Ipomoea capillacea Don.

Ipomcea coccinea var. hederifolia Gray.

Ipomcea muricata Cav.

POLEMONIACE^E :

Gilia ftoccosa Gray.
Gilia multiflora Nutt.
Gilia thurberi Gray.
Linanthus aureus (Nutt.) Greene.
HYDROPHYLLACE^E :
Ellisia torreyi Gray.
Emmenanthe pendulaeflora Benth.
Nama hispida Gray.
Phacelia distans Benth.

BORRAGINACE.E :

Amsinckia tessellata Gray.

Coldenia canescens DC.

Cryptanthe intermedia (Gray) Greene.

Cryptanthe pterocarpa (Torr.) Greene.

Eremocarya micrantha (Torr.) Greene.

Lithospermum multiflorum Torr.

Onosmodiurn thurberi Gray.

Pectocarya linearis DC.
VERBENACE.E:

Lippia wrightii Gray.

Verbena ciliata Benth.

Verbena wrightii Gray.
LABIATE :

Agastache pallidiflora (Hell.) Rydb.

Hedeoma hyssopifolia Gray.

Hyptis emoryi Torr.

Monarda pectinata Nutt.

Monarda scabra Beck.

Salvia arizonica Gray.

Stachys coccinea Jacq.

Trischostema arizonicum Gray.

FLORA OF THE SANTA CATALINA MOUNTAINS.

45

List of Characteristic Species — Continued.

SOLANACE.E :

Lycium berlandieri Dunal.
Lycium fremontii Gray.
Lycium paniflorum Gray.
Nicotiana trigonophylla Dunal.
Solatium fendleri Gray.

SCROPHULARIACE^E :

Castilleja gloriosa Britt.

Castilleja Integra Gray.

Cordylanthus wrightii Gray.

Linaria canadensis L.

Maurandia antirrhinifolia (Poir.) Willd.

Mecardonia peduncularis (Benth.) Greene.

Mimitanthe pilosa (Benth.) Greene.
= Mimulus pilosus Wats.

Mimulus cardinalis Dougl.

Mimulus guttatus (L.) DC.

Mimulus langsdorfii Sims.

Orthocarpus purpurascens Benth.

Pentstemon barbatus (Cav.) Nutt.

Pentstemon spectabilis Thurber.

Pentstemon torreyi Benth.

Pentstemon wrightii Hook.

Scrophularia sp.

Stemodia durantifolia (L.) Sw.
BIGNONIACE.E:

Chilopsis linearis (Cav.) Sweet.
= Chilopsis saligna Don.

Stenolobium incisum Standley.
ACANTHACE.E:

Anisacanthus thurberi (Torr.) Gray.

Carlowrightia arizonica Gray.
PLANTAGINACE^E :

Plantago fastigiata Morris.

Plantago ignota Morris.
RUBIACE.E:

Bouvardia triphylla Salisb.

Diodia teres Walt.

Galium asperrimum Gray.

Galium rothrockii Gray.

Galium wrightii Gray.

Houstonia wrightii Gray.
CAPRIFOLIACE^E :

Sambucus mexicana Presl.

Sarnbucua vestita Woot. & Stand.

Symphoricarpos oreophilus Gray.
VALERIANACE.B:

Valeriana arizonica Gray.
CAMPANULACE.E :

Lobelia gruina Cav.

Specularia biflora (R. & P.) Fisch. & Mey.
COMPOSITE :

Achillea lanulosa Nutt.

Actinolepis lanosa Gray.

Antennaria marginata Greene.

Artemisia dracunculoides Purah.

Artemisia ludoviciana Nutt.

Artemisia sp.

Artemisia sp.

Baccharis emoryi Gray.

Baccharis glutinosa Pers.

Baccharis pteronoides DC.

Baccharis sarothroides Gray.

Baccharis thesioides H. B. K.

Bceria chrysostoma Fisch. & Mey.

Bahia absinthifolia Benth.

Baileya multiradiata Harv. & Gray.

Bebbia juncea (Benth.) Greene.

COMPOSITE — Continued:
Bidens sp.
Brickellia californica (T. & G.) Gray.

= Coleosanthus californicus (T. & G.) Kze.
Brickellia grandiftora Nutt.

= Coleosanthus grandiflorus (Hook.) Kze.
Carduus rothrockii (Gray) Greene.
Carduus sp.

Carpochcete bigelovii Gray.
Chrysoma laricifolia (Gray) Greene.

= Aplopappus laricifolius Gray.
Crassina pumila (Gray) Kze.

= Zinnia pumila Gray.
Dugaldia hoopesii (Gray) Rydb.
Encelia farinosa Gray.
Erigeron macranthus Nutt.
Erigeron neomexicanus Gray.
Erigeron wootoni Rydb.
Eriocarpum gracile (Nutt.) Greene.

= Aplopappus gracilis (Nutt.) Grey.
Eupatorium arizonicum (Gray).

= Eupatorium occidentale var. arizonicum

Gray.

Eupatorium pauperculum Gray.
Eupatorium rothrockii Gray.
Franseria cordifolia Gray.
Franseria deltoidea Torr.

= Gcertneria deltoidea (Torr.) Kze.
Franseria tenuifolia Gray.

— Gcertneria tenuifolia (Gray) Kze.
Franseria ambrosioides Cav.
Gnaphalium decurrens Ives.
Gnaphalium wrightii Gray.
Guardiola platyphylla Gray.
Gymnolomia mulliflora Rothr.
Gymnosperma corymbosa DC.
Helenium thurberi Gray.
Helianthella arizonica (Gray).

= Helianthella quinquenervia var. arizonica

Gray.

Hieracium discolor.
Hieracium lemmoni Gray.
Hymenoclea monogyra T. & G.
Hymenopappus mexicanus Gray.
Hymenothrix wrightii Gray.
Isocorna hartwegi (Gray) Greene.

=*Bigelovia hartwegi Gray.
Laphamia lemmoni Gray.
Laphamia sp.

near to Laphamia halimifolia Gray.
Machasranthera tanacetifolia (H. B. K.) Neea.
Pectis papposa Gray.
Perityle coronopifolia Gray.
Picradenia biennis (Gray) Greene.

= Actinella biennis Gray.
Pinaropappus foliosus Hell.
Psilostrophe cooperi (Gray) Greene.

= Ridellia cooperi Gray.
Rudbeckia laciniata L.
Senecio neomexicanus Gray.
Solidago bigelovii Gray.
Solidago marshallii Rothr.
Solidago sparsiflora var. subcinerea Gray.
Stephanomeria runcinata Nutt.
Stevia sp.

Tagetes lemmoni Gray.
Trixis angustifolia var. latiuscula Gray.
Verbesina encelioides (Cav.) B. & H.
*=Ximenesia encelioides Cav.

46

VEGETATION OF A DESERT MOUNTAIN RANGE.

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

The latitude of the Santa Catalina Mountains and their position in
the midst of a continental desert give to their lower slopes the climate
which is well known to characterize southern Arizona : a low unequally
distributed rainfall, a short winter with frequent severe frost, and a
long summer with high maximum temperatures and low atmospheric
humidity. The longitudinal position of the Santa Catalinas, between
the Pacific Coast and the southern Great Plains, gives to their climate
also some of the characteristics of both these regions, notably in respect
of the incidence of the rainfall seasons. Both the winter rains of the
Pacific Coast and the summer rains which are prevalent on the Great
Plains extend in attenuated form to Tucson and to the Santa Catalinas,
giving them a short rainy season in July and August, often extending
over into September, and a longer less pronounced rainy season from
December to February or March.* Although the amount of rain in these
seasons increases with altitude, the duration of the seasons themselves
is essentially the same from Tucson to the summit of Mount Lemmon,
and in fact throughout southeastern Arizona.

FEET
9.000

8,000

7.000

6,000

5,000

4000

3,000

JAN. II

JULY 8

DEC 32

JAN. FEB MAR. APR MAY

JUNE

AUG.

NOV. DEC.

FIG. 2. — Schematic representation of rainfall seasons and length of frostless season at Tucson
and in the Santa Catalina Mountains, showing averaged limiting dates of rainfall seasons
for 8 years and averaged limits of the frostless season for 1909, 1910, and 1911 (A A), and
for 1912, 1913, and 1914 (B B).

The long frostless season characteristic of Tucson and the foothills
of the Santa Catalinas naturally decreases in length with altitude until
at 8,000 feet it is only one half as long. The curves of decreasing length
of frostless season and a diagrammatic representation of the incidence
of the rainy seasons are shown in figure 2.

The gentle rains and occasional snowfall of the winter season serve
to replenish the moisture of the soil at all altitudes, but on the desert

* See Shreve, Forrest.
17:9-26, 1914.

Rainfall as a Determinant of Soil Moisture. The Plant World,

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 47

their effect is soon overcome by the desiccating conditions of March
and April. The hot and rainless weeks which precede the mid-summer
have been designated the "arid fore-summer." On the desert this is
a season in which the temperature conditions are conducive to activity
on the part of plants, while the soil moisture conditions are increasingly
deterrent to it. As a result of these conflicting conditions activity may
be observed in the trees which grow near a constant water supply, as
Populus sp. (cottonwood) and Salix sp. (willows), trees which possess
deep-seated root systems, as Prosopis velutina (mesquite), and plants
which contain stores of water, as all species of cacti. The activity of
Populus and Prosopis consists in both flowering and leafing-out, as well
as in shoot growth; in the cacti it consists in flowering and in some
species also in growth. Among all desert plants other than those
indicated the arid fore-summer is a period of drought-rest.

With respect to the water relations of plants the arid fore-summer
is the most trying season of the year, combining low soil moistures with
atmospheric conditions that compel active transpiration. In all respects
in which moisture conditions may be critical for the survival of individuals
or the limitation of the distribution of species it is in the arid fore-summer
that the critical intensity of these conditions must be sought.

The retardation of spring which accompanies increasing altitude
results in a shortening of the arid fore-summer from a length of 15
weeks on the desert to 11 weeks at 6,000 feet and 6 weeks at 8,000 feet
(see fig. 2). Not only does this trying season decrease in length with
altitude, but its physical conditions become ameliorated, as will be shown.

The "humid mid-summer" commences on July 8 and lasts until
September 12, these being the average dates, for 8 years, of the first
and last rains of 0.50 inch or more. In this season the moisture con-
ditions of desert and mountain top are more nearly alike than at any
other time. It is the season of greatest vegetative activity on the
desert and in the forest also. On the desert it is the only season in
which germinations take place among the perennials, and it is the chief
season of growth among all perennial plants, including those that have
been in leaf during the arid fore-summer. In the Encinal region the
evergreen oaks renew their foliage at the advent of spring, but the great
mass of vegetative activity awaits the humid mid-summer. In the
Forest the pines also commence growth with the cessation of frost,
but make their chief growth during July and August. The humid
mid-summer is also the chief period of activity for the herbaceous
perennials and small shrubs of the forested elevations. Heavy snow-
fall during mid-winter or the occurrence of exceptionally late winter
rains may bring about growth among the herbaceous perennials of the
forest during the arid fore-summer. In fact a few species, notably
Frasera speciosa and Dugaldia hoopesii, commence growth before the
last frosts of spring.

48 VEGETATION OF A DESERT MOUNTAIN RANGE.

At the higher altitudes the shortness of the growing season and the
coldness of its nights are inimical to the activity of the herbaceous
perennials. These circumstances make very difficult the introduction
into the Forest region of plants which would seem calculated to flourish
in a region of similar moisture conditions.

After the close of the humid mid-summer the desert is subjected
to a variable period of 6 to 10 weeks of arid conditions, a season known
as the "arid after-summer." Although the temperature, humidity,
soil moisture, and evaporation may reach as extreme values in the
arid after-summer as in the arid fore-summer, nevertheless the total
duration of such extremes is not as great in the former season. A
general cessation of vegetative activity occurs in September and Octo-
ber at the higher elevations and in October and November at the lower
ones. On the desert it sometimes happens that occasional rains during
the arid after-summer prolong the activity of the shrubs and even of
the summer ephemerals to such a late date that they may be seen in
flower side by side with root-perennials which are characteristic of the
winter season.

RAINFALL.

The figures for the monthly average rainfall at Tucson, as determined
from the 38-year record (1876 to 1913), show that the year falls natur-
rally into two humid and two arid seasons (see fig. 4) . Without regard
to the average dates upon which the heavy rains of the humid seasons
commence or terminate, the humid winter may be seen to fall within
December, January, February, and March, and the humid mid-summer
within July, August, and September. Making this artificial division
by months between the rainfall seasons, the percentages of the total
annual precipitation which fall in the four seasons are as follows:
humid winter 31.1 per cent, arid fore-summer 5.9 per cent, humid mid-
summer 50.6 per cent, arid after-summer 12.4 per cent. The two rainy
seasons yield 81.7 per cent of the total annual rainfall, and the light
rains of the two arid seasons (which form the remaining 18.3 per cent)
are of very slight influence upon vegetation. The rains of November
may bring forth some of the winter herbaceous perennials, without any
effect on the large perennials other than the inducing of leaves on Fou-
quieria and Parkinsonia. The rains of the arid fore-summer are usually
too light and too widely separated to bring into activity either the
summer ephemerals or the perennial plants.

SEASONAL DISTRIBUTION OF RAINFALL.

On the Pacific Coast the monthly distribution of rainfall brings over
75 per cent of the annual total within the winter months. On passing
eastward through Arizona this predominance of winter rain is gradually
lost until it becomes less than 20 per cent of the annual total at the
Rio Grande River in New Mexico. Conversely, the precipitation of

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

49

the summer months is almost negligible on the Pacific Coast and grad-
ually increases on passing eastward until it reaches 50 per cent at Tucson.
Between Tucson and the Rio Grande it remains at about 50 per cent, but
from the basin of the Rio Grande eastward the rainfall seasons of the
Tucson region cease to be a natural division of the year (see table 1).

TABLE 1. — Percentages of summer rainfall and of winter rainfall to the annual rainfall for
a series of stations stretching from the Pacific coast to the Rio Grande River, through
southern Arizona.

Station.

Winter.

Summer.

Total.

Los Angeles, Cal

76 0

.7

76 7

Riverside, Cal.

76 8

2 6

79 4

Indio, Cal

73.3

12.0

85.3

Yuma, Ariz

59.0

20 3

79.3

Gila Bend, Ariz

47 5

35 5

83 0

Maricopa, Ariz. . .

45 9

34 0

79 9

Casa Grande, Ariz.

43 7

37 6

81 3

Tucson, Ariz

30.7

50.9

81.6

Benson, Ariz

24 8

57 5

82 3

Bowie, Ariz. . .

34 9

48 5

83 4

Lordsburg, N. Mex

28.3

49.9

78.2

Deming, N. Mex

22.9

64.8

77.7

Agricultural College, N. Mex. . .

17.5

56.8

74.3

In figure 3 are given curves showing the percentages of the annual
rainfall which are formed by summer rains and by winter rains for a
chain of 13 stations stretching from Los Angeles to Mesilla Park, New
Mexico, on the Rio Grande River.

TABLE 2. — The total annual rainfall, the summer rainfall, and the percentage of the latter to
the former for very wet and very dry years at Tucson.

Years.

Eight wet years
(14 inches or over).

Years.

Eleven dry years
(9 inches or less).

An-
nual.

Sum-
mer.

Per-
centage.

An-
nual.

Sum-
mer.

Per-
centage.

1876

14.02
16.66
14.92
15.59
15.07
18.37
15.04
24.17

10.18
10.51
11.98
9.27
1.77
10.84
9.04
4.50

72.6
63.1
80.3
59.5
11.7
59.0
60.1
18.6

1880

6.61
8.48
5.26
8.02
7.78
7.14
8.38
7.79
8.61
8.80
7.85

4.79
3.03
2.88
3.97
3.44
2.61
3.72
2.45
2.31
5.36
5.29

72.5
35.7
54.8
49.5
44.2
36.5
44.3
31.5
26.8
60.9
67.4

1878

1883 . .

1881

1885

1882

1886

1884

1891

1889

1894

1890

i 1899

1905

1900

Average percentage

1902

1903

1904

Average percentage

53.1

47.6

The fall of approximately half the annual precipitation in the humid
mid-summer is by no means a constant occurrence at Tucson. In
1881 the summer rainfall was 80.3 per cent of the annual, and in 1884

50

VEGETATION OF A DESERT MOUNTAIN RANGE.

it fell to 11.7 per cent. Neither does the percentage of summer rain
fluctuate in relation to the occurrence of very wet or very dry years.
In 8 of the wettest years since 1876 (14 inches or above) the summer
rain was 53.1 per cent of the total, and in 11 of the driest years (9 inches
or less) the summer yielded 47.6 per cent of the total (see table 2).

The impossibility of securing figures for the winter precipitation in
the Santa Catalina Mountains makes it necessary to estimate the
annual totals of rainfall at different altitudes from the known figures

20-

10 -

LOS RIVER- INQIO. YUMA. GILA MARI- CASA TUCSON. BENSON. BOWIE. LOROS- DEWING MtSILLA
ANGELES. SIDE. BEND. COPA. GRAT4DE. BURG. PARK.

FIG. 3. — Graphs showing percentage of winter rainfall to annual total
(light line) , and of summer rainfall to annual total (heavy line) , for
a chain of 13 stations from the Pacific to the Rio Grande.

for the summer rain. The average rainfall at the stations at 7,600 feet
and 8,000 feet for the years 1907 to 1914 is 17.45 inches (443 mm.),
from which it may be assumed that the annual average is approximately
35 inches (889 mm.). The summer rain at Tucson during 1907 to 1914
was 54.7 per cent of the annual total. If the seasonal distribution of
rain is the same on the mountain that it is at Tucson, the above esti-
mate of the annual total for the mountain is correct within 1 or 2 inches.
The influence of altitude on the seasonal distribution of rainfall in
Arizona is a matter which can not be determined without further data

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

51

INCHES
2 SOr—

than are now in hand. During the years 1907 to 1912 the percentage
at Benson, Arizona (3,523 feet), was 60 per cent, that at Globe, Arizona
(3,525 feet), was 42.1 per cent, the average of the two 51.0 per cent.
The average of the percentages for Fort Huachuca (5,100 feet), Fort
Apache (5,200 feet), and Bisbee (5,500 feet) is 55.5 per cent. Although
these figures indicate that up to
5,000 feet there is about the same
percentage that holds at 2,400 feet
(at Tucson), nevertheless at Flag-
staff (6,907 feet) the summer rain
was only 40.7 per cent of the total
in the years mentioned. At Chlar-
son's Mill (7,200 feet) an incom-
plete record indicates that in 1907,
1909, and 1910 the summer rain
was far below the percentages for
Tucson for those years. At Greer
(9,200 feet), on the Mogollon Pla-
teau, the summer rain was a much
greater percentage of the annual total in 1905 than it was at Tucson,
while in 1906 and 1908 the percentages were nearly identical. It can
only be said, therefore, that a much larger body of data is necessary to
determine the possible change of seasonal distribution of rain due to
altitude. The evidence at hand indicates that there is little probability
of a marked influence. (See table 3.)

TABLE 3. — The average annual rainfall for 1907 to 1912, the average summer rainfall for the
same years, and the percentage of the latter to the former for stations at different altitudes
in central and southern Arizona.

n

JAN FEB. MAR APR, WAV JUNE JULY AUG. SEPT OCT NOV. DEC.

FIG. 4. — Diagram showing monthly distribution
of rainfall at Tucson. Averages of record for
38 years, 1876 to 1913 inclusive.

Station.

Altitude.

Annual.

Summer.

Percentage.

Tucson

Feet
2390

Inches
11.54

Inches
6 66

Inches
57 7

Benson .

3523

10 51

6 31

60 0 1

Globe

3525

17 01

7 16

42 1 I51'0

Fort Huachuca . .

5100

15 53

9 24

59 5 1

Fort Apache

5200

16.95

7 53

44.4 [55 5

Bisbee

5500

20.68

12 95

62 6 J

Flagstaff

6907

22 60

9 20

40 7

ALTITTJDINAL INCREASE OF RAINFALL.

The measurements of summer rainfall on the Santa Catalina Moun-
tains were begun in 1907 by the installation of a metal gauge at 7,600
feet, where the record was secured until 1911, after which it was re-
moved to a nearby ridge at 8,000 feet. During 1908 and 1909 readings
were secured at the base of the mountain and at 6,000 feet, in 1910 at

52

VEGETATION OF A DESERT MOUNTAIN RANGE.

6,000 feet only. In 1911 a series of stations was selected at vertical
intervals of 1,000 feet, from the base of the mountain, 3,000 feet, to
the station at 8,000 feet, and in 1912 a station was established on
Mount Lemmon, at 9,000 feet. These stations have been continued
in the succeeding summers.

The readings of the gauges have been made at irregular intervals, as
opportunity afforded; the water has been protected from evaporation
by the use of kerosene, and has been measured volumetrically. The
installation of the gauges has been made each spring in tune to secure
the first of the summer rain, and the final readings have been made in
September, closing in 1911 on the 22d to the 25th, in 1912 on the 28th
to 30th, in 1913 on the 25th to the 27th, and in 1914 on October 10th
to llth. The location of the gauges at the various altitudes has been
such as to give them comparable topographic surroundings. Each
station is at the summit of a ridge with a commanding opening to the
south and without nearby trees. A record of rain has also been secured
at the Xero-Montane Garden at 5,300 feet, near the head of Soldier
Canon and just below the 6,000-foot station. A recapitulation of all
the readings of mountain rainfall is given in table 4.

TABLE 4. — Summer Rainfall in the Santa Catalina Mountains.

All readings cover the total precipitation of July, August, and September. Starred figures include
some October rainfall. Figures followed by plus are incomplete, owing to the overflowing of
gauges.

Eleva-
tion,
feet.

1907

1908

1909

1910

1911

1912

1913

1914

Averages of
perfect records.

Inches.

Milli-
meters.

3,000
4,000
5,000
6,300
6,000
7,000
7,600
8,000
9,000

6.65

9.72

6.27
9.45
11.97
12.51
11.07
15.86
21.30

5.61
9.77
8.24
8.67
8.68
14.57

6.46
8.59
10.27
6.09
8.73

10.62*
14.73*
19.13*

7.55
10.63
12.40
8.88
8.05
15.21
18.43
16.47
10.01

192

270
315
226
204
387
468
418
254

9.21*
6.50

10.75
3.42

6.05
5.28

22 . 68*
27.64+

20.92

20.63

17.91

11.40

19.76
20.93 +

13.18
10.01

27.82+
27.17+

The only record of daily rainfall for the Santa Catalinas is one
secured in Marshall Gulch, at 7,600 feet, from June to August 1911,
by Professor J. G. Brown, of the University of Arizona. A comparison
of the daily rainfall at Marshall Gulch and at 8,000 feet with that at
the Desert Laboratory (2,663 feet) for the period of these observations
is given in table 5. The number of rainy days on the desert was greater
than the number on the mountain top — 31 and 19 respectively — owing
to the 16 days with only a trace of rain at the Laboratory. The total
rainfall of the three months was 5.42 inches at the Laboratory (for

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

53

exactly the same days covered by the Marshall Gulch record), and
14.86 inches at the mountain station. The general correspondence
between the dates of heavier rains at these stations, 5,000 vertical
feet apart, indicates the close relationship of the atmospheric factors
which determine the rainfall of all altitudes.

TABLE 5. — Comparative daily incidence of rainfall at the Desert Laboratory (2,663 feet) and
at the Montane Garden in Marshall Gulch (7,600 feet), for June, July, and August 1911.

Day of
month.

June.

July.

August.

Day of
month.

June.

July.

August.

D. L.

M.G.

D.L.

M.G.

D.L.

M.G.

D.L.

M.G.

D.L.

M.G.

D.L.

M.G.

1st. . .

.98
T

4.03

17th

.19
T
T
03

1.96

.52

2d...

18th

3d...

19th. .

.62

.46
.53
.76
.88

4th. .

20th . .

.61
.10

2.60
.64

5th. .

21st. .

.21

.14

6th. .

22d .

7th. .

T

T

23d . . .

8th. .

.08

.15

.87

24th . .

14

9th. .

T
.10

25th . .

.02

10th. .
llth. .
12th. .
13th. .
14th..
15th . .

T
.01
T
T
T

26th.

.33

.42
.20

.25

T

1.27

T

27th. .

28th

.42
.02

T

T
.10

.30

29th . .
30th . .
31st .

T
T

16th. .

.34

Total rainfall: Desert Laboratory, 5.42 in.; Montane Garden, 14.86 in. Total number of
rainy days: Desert Laboratory, 15 (or 31, including days with T); Montane Garden, 19.

Another comparison which it is possible to institute between the
summit of the Santa Catalinas and the desert is the summer rainfall
totals from 1907 to 1914 inclusive (see fig. 9). The directions of the
curves which show the march of the summer precipitation from year
to year indicate an almost complete lack of relationship between the
mountain and the plain. It is obvious that the curve of altitudinal
increase of rainfall determined in such a year as 1910 would be very
unlike the curve determined in 1911.

It has been suggested by Smith * that there may be a relative
increase of rainfall at the higher altitudes as the summer advances,
which is to say that the gradient of increase of rainfall with altitude is
steeper for the late summer than it is for the early summer. In order
to test this possibility the series of ten readings taken in the humid mid-
summer of 1911 and the one set taken in the early arid after-summer
have been grouped into totals for five periods of approximately one
month each (table 6) . An inspection of the table shows that the maxi-
mum rainfall occurred between July 18 and August 24 at 3,000, 4,000,

* Smith, G. E. P. Groundwater Supply and Irrigation in the Rillito Valley. Ariz. Agric.
Exper. Sta. Bull. 64, 1910.

54

VEGETATION OF A DESERT MOUNTAIN RANGE.

5,000, and 8,000 feet, and between June 20 and July 18 at 6,000 and
7,000 feet. In similar manner the less frequent readings of 1912 and
1913 have been divided into the early summer and late summer falls,
by the latest July reading, and the averaged curves for early summer
and late summer rain are of nearly the same shape, but the late summer
curve is not so steep. This short record does not seem, therefore, to
corroborate the suggestion of Smith.

TABLE 6. — Intraseasonal distribution of summer rainfall at the Desert Laboratory and at 6
elevations in the Santa Catalina Mountains for 1911.

Rainfall of the maximum period in heavy type.

Apr. 25-27

June 20-21

July 18-19

Aug. 22-24

Sept. 22-25

Station.

to

to

to

to

to

Totals.

June 20-21.

July 18-19.

Aug. 22-24.

Sept. 22-25.

Oct. 12-14.

Des. Lab. .

0.01

1.37

4.08

2.90

1.61

9.97

3,000 feet..

.00

1.08

3.39

1.80

1.46

7.73

4,000 feet..

.00

2.15

4.65

2.65

1.69

11.14

6,000 feet..

.00

3.62

5.40

2.95

2.75

14.72

6,000 feet..

.00

5.71

3.45

1.91

1.92

12.99

7,000 feet..

.00

7.36

3.68

4.82

2.16

18.02

7,600 feet..

.69

8.30

8.83

3.48

2.56

23.86

The increase of rainfall which accompanies increase of altitude is a
phenomenon of general occurrence throughout the southwestern
United States. The curves by which such increase may be expressed
differ from each other most strikingly, according to the horizontal dis-
tance of the successive stations from each other, according to the
coastal or continental position of the series of stations, or according
to the size of the mountain range on which the successive elevations
are secured. Although it is possible to deduce mathematical formulae
for the vertical increase of rainfall, it is necessary to introduce into all
such formula a constant for the particular region or mountain involved,
and the figures thus secured are merely in the nature of hypothetical
means near which the normal conditions may fall. It would be of very
great interest in the extension of plant geography to possess data on
the actual amounts of rainfall at successive elevations in a large number
of mountains and shelving plains throughout the southwest. The mean
rainfall conditions which are expressed in a gradient based on a long
climatological record are of great importance in connection with vege-
tation, but only when consideration is also given to the extremes of
rainfall, and particularly to the lower extremes, if a semi-arid country
is under consideration. The securing of typical normal gradients of
altitudinal increase of rainfall is not of so much importance in plant
geography, therefore, as a knowledge of the actual oscillations of the
rainfall conditions from year to year throughout the series of stations
or localities involved.

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

55

Smith* has deduced two curves of altitudinal increase of rain, one
applicable to Pima and Final Counties, Arizona (the counties in which
the Santa Catalinas lie), the second to Graham and Cochise Counties.
These curves are based on records of various lengths, chiefly from sta-
tions located in the valleys of these mountainous counties. Smith's
curves are reproduced in figure 5, in which they have been brought
half way down toward the base line in order to make them comparable
with the curve expressing the average summer rainfall of the Santa
Catalina Mountains for 1911, 1912, and 1913. The portion of Smith's
Graham-Cochise curve extending above 5,500 feet is based on a single
short record at 6,000 feet.

The curve of altitudinal rise of rain for 1911, 1912, and 1913 in the
Santa Catalinas is merely a simple average of the actual readings for
the three summers, without any attempt to correct in accordance with

INCHES
20 |—

6.000

7,000

8.000 9.000

8.000 9.000

FIG. 5. — Graph showing altitudinal increase of summer rainfall on the Santa Catalina Mountains
in 1911, 1912, and 1913 (solid line) ; together with Smith's curves for Pima and Final Counties,
Arizona (dotted line), and for Cochise and Graham Counties (broken line).

FIG. 6. — Graphs showing vertical increase of summer rainfall in the Santa Catalina Mountains in
1911 (solid line), 1912 (broken line), and 1913 (dotted line).

the departure of the neighboring lowland rainfall from the normal
during these years, without the application of any rainfall formula,
and without the smoothing of the lines. Reference to table 4 will show
that the record for 7,000 feet is based on two years only, and the record
for 9,000 feet on one correct summer's reading and the reading of one
summer in which the gauge overflowed.

A comparison of Smith's curves with the curve for the Santa Cata-
linas shows the latter to have a sharper rise from 3,000 to 4,000 feet,
and to have a relatively level stretch from 4,000 to 6,000 feet, where
the former curves have their sharpest ascent.

* Smith, G. E. P., loc. cit.

56

VEGETATION OF A DESERT MOUNTAIN RANGE.

Figure 6 gives the actual curves for the three summers for the Santa
Catalinas. It will be noted that in each curve there is a sharp rise
from 3,000 to 4,000 feet, a rise which continued at the same gradient
to 5,000 feet in 1911 and 1913. From these submaxima, reached at
5,000 feet in 1911 and 1913 and at 4,000 feet in 1912, there is a fall to
a subminimum at 6,000 feet in the two former years and at 5,000 feet
in the latter year. There is then a pronounced rise in the curve to
7,000 and 8,000 feet. The rainfall at 9,000 feet in 1912 was probably
an inch or more greater than indicated by the curve, in any case was
greater than that at 8,000 feet; whereas in 1913 the precipitation at
9,000 feet was less than that at 8,000 feet, in fact less than that at
5,000 feet.

The horizontal distances between the rainfall stations were unequal
(see plate A), the angle of rise from 3,000 to 4,000 feet being very

1,000

3.000

5,000

FIG. 7. — Graph showing vertical increase of summer rainfall in the Santa Catalina Mountains

in 1911 (solid line), together with averaged vertical increase in a series of 13 Weather Bureau

stations in Arizona (broken line).
FIG. 8. — Graph showing vertical increase of summer rainfall in the Santa Catalina Mountains

in 1912 (solid line), together with averaged vertical increase in a series of 21 Weather Bureau

stations in Arizona (broken line).

sharp, that from 4,000 to 5 000 slightly less sharp, and that from 5,000
to 6,000 still less sharp and exactly equal to the angle of rise from 6,000
to 7,000 feet. The stations at 8,000 and 9,000 feet are located at the
west end of the main ridge and are consequently not in line with the
lower stations. The sharp rise in elevation between the 3,000 and 4,000
foot stations is doubtless partially accountable for the rapid increase
of rainfall between them. The steep rise of the rainfall graphs between
6,000 and 7,000 feet may indicate an influence due to the position of
the 7,000-foot station on the north rim of Bear Canon, with a very
abrupt wall immediately below it. There is no topographic cause,
however, to which it is possible to attribute the dip in the rainfall
curves for 6,000 feet in 1911 and 5,000 feet in 1912.

In order to institute a comparison between the mountain gradients
of rainfall and those of the valley stations of the Weather Bureau the
data have been collated which are expressed in the curves of figures 7

SHREVE

Plate A

Adapted from Tucson ard Winkelmen

-

TOPOGRAPHIC MAP Of THE SANTA CATALTNA MOUNTAINS

Scale 625,000
101334 5 "Miles

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 57

and 8. These figures compare the summer rainfall curves of the Santa
Catalinas and those of selected stations for the same summers. In
figure 7 the rainfall of July, August, and September 1911 has been used,
for 13 stations located in southeastern Arizona, east of Phcenix and
south of Fort Apache. The rain has been averaged for each group of
stations lying within the same thousand-foot interval of altitude.
Figure 8 shows the curve for the Santa Catalinas for 1912 and the curve
for 21 stations in the same area. A single record above 5,000 feet has
been available for this curve, that at Chlarson's Mill, in the Pinaleno
Mountains.

The significance of the comparison of these rainfall records for a
single season is entirely different from that of averages for long series
of years. Such a comparison as this makes possible the contrasting of
records which are strictly contemporaneous and serves to show the
way in which the same complex of meteorological conditions affected
the precipitation at various altitudes, and how these conditions affected
the rainfall of a single mountain range in comparison with that of an
extended adjacent area lying at different levels. The extremely small
number of rainfall records secured at localities above 5,500 feet in
southern Arizona does much to vitiate such a comparison. In 1911 the
gradient of rise was greater between 3,000 feet and 5,000 feet in the
Santa Catalinas than it was in the Weather Bureau stations. In 1912
the rise was sharper in the mountains from 3,000 to 4,000 feet than
it was in the valleys, but the fall from 4,000 to 5,000 feet was paralleled
by a rise in the curve of the valley stations. The fall at 7,200 feet at
Chlarson's Mill was far below that at 8,000 feet in the Santa Catalinas
for the same period.

The shape of the averaged curve of rainfall in the Santa Catalinas
for the three summers is correlated with the nature and movement
of the convectional storms to which the summer precipitation is due.
It would appear that certain rains are derived from low-lying clouds
which form over the desert and are then driven against the mountain
wall by the prevailing southwest winds of summer. These rains in-
crease in intensity as they pass up the mountain slopes and yield their
maximum downpour at about 4,000 or 5,000 feet, according to the
conditions. The rainfall at the Xero-Montane Garden was greater
than that at the 6,000-foot station (700 feet above it and only half a
mile distant) for four of the six summers in which records have been
kept in the two localities (see table 4). The Garden is located at the
head of Soldier Canon, and just above it there is a sharp increase in
the gradient of the mountain slopes. It is probable that the head of
the canon is the terminating point in the course of many of the desert
rain storms. The rapid increase of rainfall between 6,000 and 7,000
feet may be due to a similar topographic cause, as mentioned in a
preceding paragraph, or it may give indication that the rains of the

58

VEGETATION OF A DESERT MOUNTAIN RANGE.

INCHES
30 i

higher elevations are derived from a higher cloud level, probably from
convectional clouds which form at times when the atmospheric con-
ditions cause condensation at a greater distance from the earth. When
a long series of records shall have been secured from the 9,000-foot
station it will probably show that its average rainfall is greater than
that at 8,000 feet, but the 9,000-foot record for 1913 indicates that
there will be occasional years, at least, in which the maximum for the
mountain is recorded at 8,000 feet. This probably means that at
10,000 feet on adjacent mountains there is a constantly lower rainfall
than at 8,000 or 9,000 feet.

The check in the vertical increase of rainfall which has been described
as occurring between 4,000 and 6,000 feet appears to be absent from
all curves derived from widely separated valley stations. The writer
has seen no such plateau in any
curves derived from southwest-
ern data, but there is always the
possibility that a plateau has
been smoothed out of the curves
or that the data have been sub-
jected to the influence of a
straight-line equation. The
character of the increase of pre-
cipitation with altitude in a sin-
gle small range of mountains is
no more a special case than is the
increase in a widely separated
series of stations in any locations
whatsoever. In so far as con-
cerns the study of meteorological
dynamics, such a mountain
range as the Santa Catalinas offers exceptional opportunities for investi-
gation, and much more might be learned in a single summer of intensive
meteorological study on its slopes than could be ascertained by an exami-
nation of records of rainfall covering a period of a thousand years.

As regards vegetation, the most important feature of the study of
rainfall conditions is the determination of the extremes of variation in
the amount and seasonal distribution of rain, and the ascertaining of
the effect of these extremes upon the conditions of soil moisture. Years
of heavy precipitation are important for the maintenance of the forest
which clothes the higher mountain slopes and for the general restora-
tion of the supplies of soil moisture and ground water. The years of
low precipitation, and especially the series of consecutive years with
deficient rainfall, are of first importance to the vegetation which occu-
pies the Encinal region of the mountains. During such years, and
particularly during the arid fore-summer of such years, the lowest

^- — Graph showing lack of relation between
summer rainfall at Marshall Gulch (7,600 feet)
and at Desert Laboratory (2,663 feet) from 1907
to 1914.

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 59

individuals of all Encinal and Forest species are subjected to conditions
of water supply which are perhaps below any conditions that have
previously occurred during their lives, or are surely — in the case of
perennials — the most trying conditions when considered in the light
of the plants having grown to greater size and heavier water-demand
than during the dry periods of their earlier existence.

SOIL MOISTURE.

It is obvious, from a consideration of the monthly distribution of
rainfall in the Santa Catalinas, that at all elevations there are annually
two periods of high soil moisture, coinciding with the humid mid-
summer and the humid winter, and two periods of decreasing soil mois-
ture content, coinciding with the arid fore-summer and arid after-
summer. The influence of the earliest rains of summer and winter is
quickly exerted in an elevation of the soil moisture, but at the close
of these seasons it is with relative slowness that the soil falls to low
percentages of moisture, particularly at the highest altitudes. The
minimum moisture content of the year is usually to be detected just
before the first heavy rain of the humid mid-summer, but the content
in September or October may sometimes be quite as low.

At low elevations in the Santa Catalinas the annual march of soil
moisture may be expected to be analogous to that which has been
described by the writer for Tumamoc Hill, the site of the Desert
Laboratory.* Marked differences will result from a comparison of the
two localities, however, owing to the difference in the character of
the soil. The very fine clay of Tumamoc Hill is conservative in its
changes of moisture content, both with respect to increases and de-
creases of moisture, while the coarse loam found at the lower elevation
in the Santa Catalinas possesses a greater permeability and a lesser
holding power. The soils of elevations of 7,000 feet and more are
richer in organic matter than those of the Desert and Encinal regions
of the mountain, and are doubtless more like the clay of Tumamoc
Hill in the smoothness of their curves of change in moisture content.

The few readings of soil moisture content that have been made were
directed toward a determination of the soil conditions in the most arid
portion of the year. It is obvious that it is these annual minima which
are of the greatest importance to plants, particularly to such plants
as are near the lowest limit of their vertical occurrence. Much less
interest attaches to the high moisture contents which might be found
in the midst of the rainy seasons. It is true that these high moistures
are the ones which call forth general vegetative activity and condition
the appearance of ephemeral plants at the lower altitudes. It is like-
wise possible that high and protracted soil moistures may be of some
importance as a limiting factor for desert species at the upper edges

* Shreve, Forrest. Rainfall as a Determinant of Soil Moisture. The Plant World, 17 : 9-26, 1914.

60 VEGETATION OF A DESERT MOUNTAIN RANGE.

of their ranges. It was impossible, nevertheless, to secure a set of
soil samples in the humid mid-summer which would be representative
of the maximum moisture conditions and at the same time comparable
for the various altitudes. A set of samples taken at the same interval
after a rain of the same amount, at each of the several elevations,
would comply with the requirements.

All samples of soil for moisture content were taken from a depth
of 15 cm. The conditions at this depth are of importance for ephemeral
herbaceous plants and for some shrubs, but the trees and larger shrubs
are, of course, dependent for their supplies on much more deep-seated
bodies of soil. The rocky character of the substratum means that the
largest perennial plants are dependent to a great extent upon the
moisture contained in the soil which occupies the crevices of the rock
in situ. It is particularly noticeable that the lowest trees of the Encinal
region grow in the uppermost part of talus slopes or along the bottoms
of cliffs. In such situations it is doubtless possible for the roots of
these trees to reach soil-filled crevices which are fed by gravity with
the water of large veins of soil above.

The samples of soil were secured by digging with a hand trowel and
transferring quickly to bottles, which were tightly stopped, and then
coated over the stopper with vaseline. The soils were dried in the
original bottles by heating to 100° C. until they showed constant
weight. The percentages of moisture have been calculated on the dry
weight as unity. The physical texture of all samples taken was very
similar, but there was a greater amount of humus in those from the
higher elevations.

Three series of soil samples were taken at various times to determine
the conditions prevailing in the arid fore-summer. These samples
were taken at 1,000-foot intervals, from the vicinity of the rainfall
stations, and were secured in pairs, one sample being from a south
slope and one from a north slope. The localities chosen for sampling
were typical of the slopes at the several elevations, and in every case
the pair of samples was secured in the midst of the dissimilar vegeta-
tions which occupy the opposed slopes.

On April 27 to 29, 1911, a series was secured from 3,000 feet to 7,000
feet (see table 7) . For the three months preceding the taking of these
samples there had been only light and infrequent rains over the sur-
rounding region, the rainfall of the mountains themselves for this period
being unknown. At Tucson there was a rainfall of 0.28 inch on April
2, and there was no appearance of rain on the mountains after that
date. On June 9 to 11 another series of samples was secured at the
same stations, together with a pair from the station at 8,000 feet.
There had been no rain between the securing of the two sets of samples.

A comparison of the percentages of moisture in April and in June
shows them to be of about the same order of magnitude. The relative

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

61

dryness of the three months preceding the taking of the first set of
samples, together with the 25 days of rainless weather just preceding
the taking of the samples, had reduced the moisture of the superficial
soil to an amount which was near the minimum for the year, as repre-
sented by the percentages for June. The percentages for June were
all slightly lower than those for April. The reading of 5.2 per cent for
the south slope at 5,000 feet in April is undoubtedly too high. The
fall in moisture on the north slope at 7,000 feet from 9.2 per cent in
April to 3.2 per cent in June is doubtless significant of the long reten-

TABLE 7. — Soil moisture in the arid fore-summer at a depth of 15 cm. on north and south
exposures at seven altitudes on the Santa Catalina Mountains.

Elevation
and
exposure.

Apr. 27 to 29,
1911.

June 9 to 11,
1911.

May 15 to 20,
1914.

Average of
the three de-
terminations.

3,000 south

2.5

2 3

1 2

2 0

4,000 south

2.7

2.4

1.0

2.0

4,000 north

3.2

2.8

1.5

2.5

5,000 south

5.2

1.8

2 4

3 1

5,000 north

3.8

1.3

5 5

3 5

6,000 south

2.3

1 3

1 9

1 8

6,000 north

5.1

2 3

3 0

3 5

7,000 south

3.1

2.6

2.1

2.6

7,000 north

9.2

3.2

4.1

5 5

8,000 south

6.1

8.8

7.4

8,000 north . . .

8.1

14 5

11 3

9 000 south

9 4

9 4

9.000 north

27.9

27.9

The percentages are based on dry weight.

tion of moisture derived from winter rains, characteristic of the forested
elevations. In June the north slope at 8,000 feet had fallen to a slightly
lower percentage of moisture than the north slope at 7,000 feet in April.

Between May 15 and 20, 1914, another series of moisture samples
was secured at the same localities and extended to the 9,000-foot
station. The preceding winter had been slightly below the average in
precipitation, but the rainfall for March had been above the average.
At the time of the taking of the samples there had been no rains for
six weeks. This series of percentages is similar to those secured in
1911, and the three may be taken together as indicating the average
soil moisture conditions of the arid fore-summer.

A significant feature of all three of the series of moisture determina-
tions is the fact that there is no appreciable increase of soil moisture
up to an elevation of 7,000 feet, beyond which elevation there is a
sharp rise in the percentages, particularly those for the north slopes.
In other words, so far as the superficial soil moisture conditions are
concerned, the arid fore-summer carries the desert up to the lower
limit of the Pine Forest.

One of the underlying causes of the importance of slope exposure
for vegetation is revealed in a comparison of the percentages of soil

62

VEGETATION OF A DESERT MOUNTAIN RANGE.

moisture for north and south slopes. For the stations at 4,000, 5,000,

and 6,000 feet the percentages for the two slopes scarcely differ by

more than the error which may be attributed to the inequalities of the

moisture in adjacent bodies of soil. Nevertheless, in all but two cases

there is a slightly greater moisture

content on the north slope than on

the south. At 7,000 feet the difference

between the two exposures becomes

greater, and is still greater at 8,000

and at 9,000 feet. An inspection of

the averages (see fig. 10) shows that

the south slope at 7,000 feet has a

lower soil moisture than the north

slope at 6,000 feet. The south slope

at 8,000 feet, however, has a higher

moisture than the north slope at

7,000 feet.

The fact that there is a very slight
difference between the soil moisture
of north and south slopes at lower
elevations and a greater difference
with increasing altitude would sug-
gest that there might be a more pronounced set of vegetational phe-
nomena resulting from slope exposure at higher elevations than at lower
ones. This is, indeed, the case, and will be discussed under a later
heading (see p. 98).

TABLE 8. — Soil moisture in the arid fore-summer and arid after-summer at a depth of 15 cm.
on north and south exposures, in shade and open, at various altitudes on the Santa Catalina
Mountains.

9.000

FIG. 10. — Graph showing vertical increase
of soil moisture at 15 cm. in the Santa
Catalina Mountains on north slopes
(heavy line) and south slopes (light line) .
Average of three determinations in arid
fore- summer.

Date.

Elevation.

Slope,
exposure.

Location.

Moisture.

May 27, 1910

Feet
5,300

South

Open

2.1

Do ...

5,300

South

Shade

2.4

Do

5,300

North

Shade

3.8

May 29, 1910

7,600

South

Open, base

Do

7,600

South

of slope
Open, crest

4.7

May 30, 1910

5,030

South

of ridge
Open

3.9

5.7

Sept. 24, 1910

4,600

South

Open

2.2

Do

4,600

North

Open

4.0

Do

5,300

South

Open

2.9

Do

6,300

North

Shaded

2.6

In the summer of 1910 several samples of soil were taken for the
determination of the moisture conditions on opposed slopes and in
shaded and open soil, as well as on the top and at the base of a slope.
These data are shown in table 8. The September readings, when taken

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 63

in comparison with the April, May, and June readings, show moistures
of about the same amount, indicating that the after-summer is often
a season of as great soil aridity as the fore-summer.

The data for shaded and unshaded soil, both in May and September,
corroborate similar determinations made on Tumamoc Hill and go to
show that in the arid seasons the influence of shade in sustaining the
moisture of soil is so slight as to be negligible. The influence of shade
in retarding the desiccation of the soil just after a rain is not without
its importance, but in the Desert and Encinal regions the soil in the
shade of trees will soon reach as low a percentage as that in the full sun.

EVAPORATION.

It has been frequently pointed out, in recent botanical literature,
that the measurement of the evaporative power of the air affords a
concise expression of the combined effects of temperature, humidity,
and air movement in so far as these factors affect the loss of water by
plants. The obvious importance of these factors — and consequently
of evaporation — in the environmental complex of the Santa Catalinas
led to the early installation of a series of atmometers (or evaporimeters)
at several elevations in these mountains. In the summer of 1906 Dr.
B. E. Livingston secured data from three porous cup atmometers at
elevations of 6,000, 7,500, and 8,000 feet.* In 1908 and 1910 the
writer installed series of atmometers at five elevations, from which
readings were secured which are not sufficiently complete and reliable
to be worthy of publication. In 1911 a new series of atmometers was
installed at the six rainfall stations, from 3,000 to 8,000 feet inclusive,
at 1,000-foot intervals. These instruments were exposed in pairs, on
north and south exposures, and were operated in the most careful
manner, in accordance with the experience of the two preceding years.
The atmometers were read at fortnightly intervals, or nearly so, and
at each reading fresh cups were installed. The actual readings were
reduced to standard by the use of an average between the original and
the final coefficients of correction. Only good distilled water was used,
and it was conveyed in tin canteens (rather than galvanized iron ones)
from which the resin remaining from the soldering had been removed
with carbon bisulphide. The bottles used for the atmometers had a
capacity of 1 gallon at the lower stations and of 2 quarts at the higher
stations, such ample amounts of water providing against the possible
danger of the atmometers going dry. The stoppers in the mouths of
the bottles were made very tight, to prevent the cups from blowing
loose, but were provided with grooves to admit air. These grooves
were stopped with loose cotton, to prevent the entrance of ants, and
the stoppers were covered by aprons to exclude rain. The atmometers
were all exposed in situations such that they received full insolation

* Livingston, B. E. Evaporation and Plant Habitats. The Plant World, 11: 1-9, 1908.

64

VEGETATION OF A DESERT MOUNTAIN RANGE.

throughout the day, except in the case of the instrument on the north
slope at 8,000 feet, which it was impossible to place in such a way as
to avoid a slight amount of light shade in the mid-morning and in the
mid-afternoon.

Readings were secured at the stations from 3,000 to 7,000 feet from
April 25 to 27 until September 5 to 6, and at 8,000 feet from June 7
until September 5. The actual amounts of the readings are given in
table 9, in terms of the average loss per day in cubic centimeters from
a standard cup.

TABLE 9. — The average daily evaporation (in cubic centimeters), for the periods indicated, on
north and south exposures, at 6 elevations in the Santa Catalina Mountains.

3,000
feet.

4,000 feet.

5,000 feet.

6,000 feet.

7,000 feet.

8.000 feet.

Dates.

S

S

N

S

N

S

N

S

N

S

N

Apr. 25-27 to May 18-19 .

97.2

75.2

68.6

50.1

61.2

57.7

52.4

60.8

48.0

May 18-19 to June 6- 7. 120.5

84.8

91.2

74.2

83.1

67.7

68.6

72.5

57.5

June 6- 7 to June 20-22 . 85 . 7

81.3

88.4

60.8

88.8

52.8

47.4

55.2

44.3

29.3

29.4

June 20-22 to July 6- 7 . 61 . 1

67.6

76.5

50.9

46.1

50.4

43.3

46.8

43.3

27.5

15.9

July 6- 7 to July 18-21 . 43 . 2

41.0

43.7

32.5

25.5

22.6

24.5

22.8

19.5

10.7

5.9

July 18-21 to Aug. 8-9. 59.8

54.7

53.5

44.2

37.2

34.2

33.6

37.3

34.1

23.3

19.0

Aug. 8- 9 to Aug. 22-23 .

61.0

56.0

59.3

45.2

44.2

37.6

36.1

39.9

27.0

13.1

7.4

Aug. 22-23 to Sept. 5- 6 .

55.6

42.8

56.9

50.8

33.4

39.5

28.3

39.3

24.0

18.9

12.5

Sept. 5- 6 to Sept. 22-25 .

45.4

56.9

39.9

31.6

20.2

22.5

20.8

28.0

30.4

11.2

6.8

In order to ascertain the altitudinal gradient of evaporation rate
the readings from the north and south slopes at each altitude have
been averaged. The averaged total evaporation of the summer for
each station has been subdivided, so as to show the amount for the
arid fore-summer as shown by the first three series of readings, and
for the humid mid-summer as shown by the last six series. The curves
in figure 12 show the altitudinal fall in evaporation rate during the
two seasons, in terms of the average daily loss from the atmometer.
These curves bring out in striking manner the low evaporation rates
of the humid mid-summer as contrasted with the arid fore-summer,
the latter being nearly twice as great as the former. There is a strong
parallelism between the two curves, but the one for the humid season
is slightly flatter than the one for the arid season. This means that,
so far as concerns the evaporation conditions alone, there is a less
differentiation between Desert and Forest in the summer rainy season
than there is in the arid portion of the summer. The pronounced drop
in evaporation from 7,000 to 8,000 feet is particularly significant, as
the former elevation marks the lower edge of the Forest, while the
latter is in the midst of the best stands of pine. It is possible that the
forest itself interferes with air movements near the ground in such a
way as to be responsible for the sharp fall in evaporation.

In order to exhibit the seasonal march of evaporation rate at the
several altitudes the curves of figure 11 have been drawn. These

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

65

curves show the averages between the south and north slopes for each
station at each reading, being expressed in cubic centimeters of evapo-
ration per day. Here again is brought out the pronounced fall in

c.c.>

110

100
90

80
70

eo

50
40
30
20

10
0

I

APR.

MAY

JUNE

JULY

AUG.

SEPT.

c.c.

100

60

FIG. 11. — Graphs showing seasonal march of evaporation rate at 6 altitudes in the Santa Cata-
linas in 1911. Amounts are average daily losses from the atmometer, and each reading is
the average of one on a north slope and one on a south slope.

rate which follows the advent of the summer rains and the cloudy

and relatively humid weather by which they are accompanied. After

the period of heavy rains by which the

humid mid-summer was opened in the last

days of June and early days of July there

was a slight rise in evaporation, followed

by a slight fall in late August and early

September. The curves for 3,000 and

4,000 feet accompany each other closely

after the first two readings, and the curves

for 6,000 and 7,000 feet accompany each

other closely throughout the summer.

The curve for 8,000 feet stands always

well apart from that for 7,000 feet. The

grouping of these curves is, therefore,

analogous to the natural subdivisions of

the vegetation. The readings taken in the

Desert region at 3,000 and 4,000 feet, those taken in the Encinal and the

lower edge of the Forest at 5,000, 6,000, and 7,000 feet, and the one series

3,000 4.000 5.000 6.000 7.000 &.QQO

Fia. 12. — Graphs showing altitudinal
decrease in rate of evaporation in
the Santa Catalinas in arid fore-
summer (heavy line) and humid
mid-summer (light line) of 1911.

66

VEGETATION OF A DESERT MOUNTAIN RANGE.

taken in the heart of the Forest stand apart in three loosely defined
groups in close parallelism to the zonation of the vegetation itself.

During the arid fore-summer the evaporation at 5,000 feet is similar
to that at 4,000 feet, while during the humid mid-summer it is more
nearly like that of the 6,000-foot station. In other words, the advent
of the rains causes the evaporation conditions of the Upper Encinal
and lower Forest region to extend downward into the Lower Encinal.

The significance of slope exposure in determining evaporation rate
is indicated in figures 13 and 14. In these graphs the vertical gradients
of evaporation at the six elevations are shown separately for the instru-
ments on the south slopes and the north slopes at each station. The
gradients for the arid fore-summer and for the humid mid-summer
are shown, as well as the curves for the entire summer. In the arid
season there is even a slightly greater evaporation on north slopes at
4,000 and 5,000 feet than there is on south slopes, but this condition

c.c

100

60

40

CO

fcO

ZO
3,000

FIG. 13. — Graphs showing altitudinal decrease in rate of evaporation in the Santa Catalinas on

south-facing slopes (heavy line) and on north-facing slopes (light line) in arid fore-summer

of 1911.
FIG. 14. — Graphs showing altitudinal decrease in rate of evaporation in the Santa Catalinas on

south-facing slopes (heavy line) and on north-facing slopes (light line) in humid mid-summer

of 1911.

is reversed at the higher elevations. In the humid season there is also
a slightly greater rate of evaporation on the north slope at 4,000 feet,
while all of the higher stations show an almost uniformly greater rate
on the south slopes. It is impossible to explain the cases in which the
evaporation was greater on north slopes than on south ones. It is
possible, of course, that they require no explanation but are typical of
the extremely arid conditions of the lowest elevations at the driest
time of the year. They are at least accordant with the fact that the
soil moisture is sometimes greater on the south slopes.

The summer averages show a difference of from 5 to 10 c.c. per day
between the evaporation on opposed slopes, in readings of 35 to 45 c.c.
per day or less. As the actual amounts of evaporation fall with increas-
ing altitude, the difference between the opposed slopes becomes pro-
portionately greater.

As in the case of all climatological data, it would be impossible to
state the normal conditions of evaporation at the various altitudes

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 67

and on opposed slopes without instrumental data for several series of
stations and covering several years. It is possible, nevertheless, to
state from the data presented that (a) the rate of evaporation through
the arid and humid summer seasons is about 3J^ times as great on the
desert as it is at 8,000 feet; (6) the rates of evaporation are approxi-
mately half as great in the humid mid-summer as they are in the arid
fore-summer; (c) at the middle and higher altitudes the evaporation
on north slopes is less than on south slopes; (d) the difference between
the amounts of evaporation on north and south slopes becomes greater
with increase of altitude, in proportion to the amounts of each.

HUMIDITY.

The prevalence of low atmospheric humidities is one of the most
pronounced features of desert climate and is an extremely potent factor
in causing the high rates of evaporation that have been shown to occur
at the lowest stations in the Santa Catalina Mountains. The relative
humidity is lowest in the arid fore-summer, although it is sometimes
nearly as low for brief periods in the arid after-summer. During the
two rainy seasons the humidity is extremely variable and may fluctuate
through a daily range of as much as 70 per cent. The daily curve of
humidity is extremely uniform during the cloudless days of April, May,
and June, falling rapidly during the early forenoon to mid-day values
as low as 5 and 10 per cent, and rising slowly through the late afternoon
and more rapidly during the night to a daily maximum of 20 to 30 per
cent just before sunrise. Cloudy days in the arid seasons cause a higher
minimum but seldom raise the maximum above 40 per cent unless
there is a trace of rainfall.

The humidities of the mountain, varying with altitude and with the
seasons, possess their greatest importance for vegetation in their role
as joint determinants of the rate of evaporation. The altitudinal
gradient of humidity has, therefore, been most satisfactorily investi-
gated when it has been measured together with temperature and wind
in the collective effect of these climatic factors upon the evaporative
power of the air. It is not without interest, nevertheless, to know
something of the relative humidities which are prevalent at the moun-
tain altitudes and are partially responsible for the rates of evaporation
encountered there.

In spite of the pronounced altitudinal changes of vegetation and of
climatic conditions which have been discussed (or are yet to be treated),
there are so many features of the Encinal and Forest vegetation that
strongly suggest the Desert (see p. 36) that it seemed particularly
desirable to secure readings of relative humidity at the forested alti-
tudes in the arid fore-summer. The few figures to be given here were
secured with a sling psychrometer and converted to percentages by
the use of Marvin's tables.

68 VEGETATION OF A DESERT MOUNTAIN RANGE.

On May 22, 1911, in the east fork of Sabino Basin, at 3,400 feet
elevation, the humidity at 3h 30m p. m. was 6 per cent, at 6 p. m. it
was 8 per cent, and at 8h 30m p. m. it was 12 per cent. At 4h 30m a. m.,
on the following day, the humidity had risen to 24 per cent. These
figures show the prevalence of desert humidities at a locality which is
low in elevation but is well in the heart of the mountain mass as a whole.
At Marshall Gulch, at 7,600 feet in the Forest region, on May 20, 1911,
the humidity at llh 30m a. m. was 10 per cent, and it was the same at
4h 30m p. m. At 9h 15m a. m. on the following day the humidity was
16 per cent, and at 3 p. m. it was 11 per cent. Although these figures
are roughly twice those of the readings in Sabino Basin, they are never-
theless indicative of a low humidity for a forested locality and show
that in the arid fore-summer there are days on which the humidity is
nearly as low as it is on the Desert.

A number of humidity readings were taken in June 1911, but none
of them showed as low values as those just mentioned. In Bear Canon,
at 6,100 feet elevation, on June 21 (a dull and intermittently cloudy
day), the humidity was 46 per cent at 1 p. m. and 42 per cent at 3 p. m.,
falling to 41 per cent at 7 p. m. In Marshall Gulch on June 23 (a nearly
cloudless day), the humidity at 5 a. m. was 33 per cent and it fell
steadily to 22 per cent at 12 noon, with a temporary rise during an
interval of cloudiness at 10 a. m. In the afternoon the percentages
rose from 25 per cent at 3 p. m. to 29 per cent at 6 p. m., but fell again
to 26 per cent at 8 p. m. The highest humidity observed at Marshall
Gulch was 48 per cent at 4h 30m p. m. on June 8, 1911, after the summit
of the mountain had been covered several hours with cumulus clouds.

Continuous records of relative humidity have been secured in yellow
pine forest at the Fort Valley Experiment Station at 7,200 feet eleva-
tion, near Flagstaff, Arizona, by Pearson.* The monthly mean values
for May and June (1909-1912) are 38 and 34.9 per cent respectively.
Some of the lowest extremes involved in the composition of these means
have been kindly communicated to the writer by Pearson. The number
of days in June on which the humidity fell to 16 per cent or less was
as follows: 8 days in 1909; 11 days in 1910; 6 days in 1911. The lowest
humidities were a single occurrence of 10 per cent and two occurrences
of 11 per cent. Humidities as low as 11 per cent also occur in July,
and values as low as 16 per cent occur between May and October.

These figures for the Coconino Plateau are in agreement with the
lowest figures secured at Marshall Gulch in May 1911, and show that
desert humidities are of frequent occurrence in the arid fore-summer,
both on isolated mountains, such as the Santa Catalinas, and on ex-
tended plateaus in the midst of the arid region. On the days that
exhibit such low humidities at higher elevations there is practically a

* Pearson, G. A. A Meteorological Study of Parks and Timbered Areas in the Western
Yellow Pine Forests of Arizona and New Mexico. Mo. Weather Rev., 41: 1615-1629, 1914.

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 69

flat altitudinal gradient of humidity. The difference between the
observed humidities of 6 per cent in Sabino Basin and 12 per cent at
Marshall Gulch is a very small one, and would doubtless register very
small differences on the rate of evaporation under otherwise identical
conditions. The differences in evaporation actually found to exist
between the base and summit of the mountain are to be ascribed to the
nocturnal humidities, which are higher in the Forest than on the Desert,
to the greater frequency of cloudiness at higher elevations in the arid
fore-summer, to the lower temperatures at higher altitudes (especially
at night), and to the wind protection afforded by the forest cover itself.

TEMPERATURE.

The investigation of temperature on the Santa Catalinas has been
carried on with a view to determining the decrease in length of the
frostless season which accompanies increase of altitude, the normal
decrease of temperature with increasing altitude, and the departures
from the normal gradient of decrease which are due to the nature of
the topographic relief and to other causes. The results secured afford
an outline of the major temperature features which are capable of
influencing the distribution or seasonal activities of the plants of the
Desert, Encinal, and Forest regions.

The character of the temperature conditions, and their relation to
altitude and topography, in an isolated desert mountain is not without
complexities which make it impossible to predict the conditions for
vegetation in a given locality through a knowledge of its altitude
and of general meteorological theory. The relative smallness of the
entire mountain mass and its position in the midst of arid plains make
its temperature conditions very different from those of extensive
plateaus of the same elevation. The currents of warm air which ascend
by day and the streams of cold air which descend by night serve to
increase the diurnal amplitude of temperature in certain situations and
to give striking differences within very short distances. Differences
of slope exposure bring about differences of diurnal warming and
nocturnal cooling of the soil, and these differences affect the general
temperature conditions and also directly influence the vegetation.
The differences of diurnal warming and nocturnal cooling which exist
between the relatively bare soils of the Desert and Encinal regions
and those of the Forest, with their heavy cover of vegetation, their
litter of leaves and high humus content, are also considerable and tend
to lessen the importance of topography at the higher elevations.

Temperature readings have been secured at two localities, at ele-
vations of 5,300 and 7,600 feet, respectively, since the early summer of
1908. Since 1911 a series of thermometers has been exposed at the
rainfall situations at 4,000, 6,000, and 7,000 feet, and during 1913 and
1914 a complete series of thermometers was maintained at 1,000-foot

70 VEGETATION OF A DESERT MOUNTAIN RANGE.

intervals from 4,000 to 9,000 feet. All of the instruments in this series
were located on the summits of ridges, so as to give comparable readings
from similar topographic situations. In addition to the six instruments
in this series there were also thermometers in the bottoms of canons at
5,000, 6,000, and 7,600 feet; a thermometer was exposed on the top
of the fire tower of the Forest Service on Mount Lemmon, the actual
elevation of the instrument being 9,225 feet, and thermometers were
buried in the topmost layer of soil at 6,000 and 8,000 feet.

Alcoholic minimum thermometers were used in the earlier years of
these observations, but were replaced by mercuric Six's thermometers
in 1913 and 1914. Various types of thermometer have been used at
the station at 7,600 feet, and as many as three instruments have been
exposed simultaneously at that place. All thermometers have been
calibrated before use and have been verified in place from time to time
by comparison with a portable thermometer of known error. The
readings of the thermometers have been taken at irregular intervals,
as opportunity afforded, and most of the figures secured are for periods
of several weeks, or for the several months which elapse between the
last visit in the autumn and the first in the spring. Only at the 7,600-
and 9,000-foot stations has it been possible to expose the thermometers
in such a manner as to secure reliable maxima; at all other stations
the only data secured have been the absolute minima for the intervals
between visits. The placing of the thermometers in small boxes, with
numerous perforations, has made possible the securing of good minima,
but no record has been made of the maxima secured under such con-
ditions of exposure. The conspicuousness of adequate instrument
shelters would have invited human interference with the thermometers
which would have been productive of errors.

A few records of temperature from the same locality for a number
of consecutive days have been secured by Professor J. G. Brown and by
Dr. H. A. Spoehr, as well as by the writer. A large number of single
observations of minima and of current temperatures have been made
by the writer at various localities, and it has been possible to use these
in connection with data from the regular stations in determining the
normal gradient of temperature decrease and in ascertaining the verti-
cal shortening of the f restless season.

LENGTH OF FROSTLESS SEASON.

It has been impossible, for the most part, to make direct observations
of the dates of last vernal and first autumnal occurrence of a tempera-
ture of 32° F. at the several stations on the Santa Catalinas. The
dates at which visits were made to the mountain were occasionally
such as to establish the dates exactly for one of the stations, and in
several cases visits were made at such frequent intervals as to place
the date within a week or two. In the majority of cases, however,

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

71

the frost dates which limited the growing season fell between the last
visit of autumn and the first one of spring. This has particularly been
the case with all of the lower stations. These circumstances have
made it necessary to resort to an indirect method of determining the
dates, which is as follows: A series of graphs was drawn showing the
march of the weekly absolute minimum temperatures at the Desert
Laboratory, as registered by thermograph, for the years covered by
the mountain records. Each reading of minimum temperature for a
given station was then compared with the minimum for the same
period at the Desert Laboratory, and the total number of such differ-
ences was averaged. In this manner it was possible to secure the
figures given in tables 10 and 11.

TABLE 10. — Altitudinal shortening of the frostless season in the Santa Catalina Mountains,
as shown by the dates of the last spring occurrence and the first autumn occurrence of a
temperature of 32° at 3 altitudes in 1908, 1909, and 1910.

Station, etc.

Year.

Last 32°
in spring.

First 32°
in autumn.

Desert Laboratory; elevation, 2,663
feet; length of frostless season, 285
days.

(1908
J1909
[1910

Feb. 20
Mar. 15
Feb. 25

Nov. 30
Nov. 30
Dec. 31

Average dates

Mar. 1

Dec. 10

Xero-Montane Garden ; elevation,
5,300 feet; length of frostless season,
195 days.

[1908
J1909
[1910

May 10
Apr. 12
Apr. 20

Oct. 10
Nov. 10
Nov. 26

Average dates

Apr. 24

Nov. 5

Montane Garden, Marshall Gulch;
elevation, 7,600 feet; length of frost-
less season, 126 days.

(1908
U909

[1910

June 15
May 31
May 9

Sept. 25
Sept. 26
Oct. 17(1)

Average dates

May 29

Oct. 2

With a knowledge of the average difference between the minimum
temperature at the Desert Laboratory and at a given station, and with
the graph showing the march of minima at the Laboratory, it was
possible to locate the approximate date of the last and first occurrence
of 32° at the mountain station. Such a graph for 1911 is given in
figure 17, together with the graph of march of minimum temperatures
at the Montane Plantation in Marshall Gulch, at 7,600 feet. It will
be noted that the graph for the Laboratory rises by several pronounced
jumps during March, April, and May, and falls by precipitate stages
during September, October, and November. The relatively sudden
advent of summer and of winter is an invariable annual occurrence,
and it has helped to make more accurate the estimation of the limiting
dates for the mountain stations. In the cases in which a minimum

72

VEGETATION OF A DESERT MOUNTAIN RANGE.

temperature of 32° or less was registered at a station during an interval
of two or three weeks, the date of such minimum could be determined
by finding the exact date of the lowest temperature for the same period
at the Desert Laboratory, and such determinations undoubtedly have
a very slight possibility of error. Taking into account the number of
direct observations and the larger number of estimations, the limiting
dates of the frostless season, given in tables 10 and 11, may contain
errors of as much as 7 to 10 days. The swamping of these errors by
averaging the dates for the 6 years of observation reduces the probable
error to about 5 days.

TABLE 11. — The altitudinal shortening of the frostless season in the Santa Catalina Mountains,
as shown by the dates of the last spring occurrence and the first autumn occurrence of a
temperature of 32° at 5 altitudes in 1911, 1912, and 1913.

[Data for 9,000 feet are partially interpolated.]

Station.

Year.

Last 32°
in spring.

First 32°
in autumn.

Desert Laboratory; elevation, 2,663
feet; length of frostless season, 282
days.

(1911
U912
[1913

Feb. 20
Feb. 26
Mar. 31

Dec. 25
Dec. 9
Dec. 8

Average dates

Mar. 7

Dec. 14

Climatological Station; elevation, 5,000
feet; length of frostless season, 248
days.

(1911
•U912
(1913

Feb. 27
Apr. 15
Apr. 14

Dec. 4
Dec. 9

Nov. 21

Average dates

Mar. 30

Dec. 1

Climatological Station; elevation, 7,000
feet; length of frostless season, 187
days.

[1911
•U912
[1913

Apr. 3
May 13
May 5

Oct. 30
Nov. 18
Oct. 13

Average dates

Apr. 27

Oct. 31

Marshall Gulch; elevation, 7,600 feet;
length of frostless season, 148 days.

(1911
0912
[1913

May 16
May 18
May 9

Oct. 30
Oct. 2
Sept. 26

Average dates

May 14

Oct. 9

Mount Lemmon; elevation, 9,000 feet;
length of frostless season, 122 days.

(1911
•U912
[1913

May 29
May 20
June 16

Oct. 16

Oct. 7
Sept. 8

Average dates

June 1

Sept. 30

In figure 2 are shown separately the curves for altitudinal shortening
of the frostless season for the years 1908, 1909, and 1910, and the years
1911, 1912, and 1913. In the latter group of years the advent of spring
was nearly a month earlier at the middle altitudes than it was during
the former years, and the advent of winter was correspondingly later,
in spite of the fact that the frostless season was of approximately the
same length at the Desert Laboratory during the two periods. A con-

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

73

sideration of the two sets of curves is more fruitful than the possession
of their average, as it shows the extent of a fluctuation which must
be a common and normal feature of the climatology of the mountain,
just as it is of every locality, regardless of its topographic location.

The average length of the frostless season at the Desert Laboratory
is about 38 weeks, at 5,000 feet in the Santa Catalinas it is about
30 weeks, while at 8,000 feet it is about 17 weeks. The period of safe
plant activity is, therefore, less than half as long in the Forest region
of the mountain as it is in the Desert of the Santa Cruz Valley. The
altitudinal abbreviation of the frostless season is of primary importance
to vegetation, especially as it is accompanied by a series of inseparable
features of temperature, such as the lower range of the entire daily
curve of temperature, the attainment of lower minima, the more

7,000
6,000
5,000
4,000
3.000
2,000
1,000

/

MARSHALL

GULCH

\

CHLARSON

J MILL

/

FL*

G S T A F F

\

/

FT HUACMUC

A

\

.

/

,

OCMISE

/

BENSON

\

/

7

TUCSON

\

PHOENIX

YUMA

0

JAN.

FEB.

MAR.

APR.

MAY

JUNG

JULY

AUG.

SEPT.

OCT.'

NOV.

DEC.

Fm. 15. — Schematic representation of length of frostless season at different altitudes in Arizona,
together with curves for limits of growing season at successive altitudes in the Santa Cata-
linas for 1909 to 1914 inclusive.

frequent occurrence, and the longer duration of freezing temperatures.
One of the cardinal features of importance in the altitudinal shortening
of the growing season is the concomitant shortening of the arid fore-
summer. The rising temperatures of spring call the vegetation of the
Desert into activity at a time of the year when extremely arid condi-
tions are bound to prevail for from 14 to 16 weeks. At the summit of
the mountain, however, the advent of spring is only 5 or 6 weeks in
advance of the earliest of the mid-summer rains. In other words, the
most trying season of the year is shortened on the mountain by the
inhibitory effects of low temperatures, so that the arid fore-summer is
only one-third as long in the Forest as in the Desert. These relations
are graphically represented in figure 2, which shows both the curves
of the frost dates and the incidence of the rainy seasons. More will
be said in regard to this subject in a succeeding section (see p. 93).

74

VEGETATION OF A DESERT MOUNTAIN RANGE.

In order to compare the altitudinal shortening of the frostless season
on the Santa Catalinas with the same datum for a series of valley
stations located at progressive altitudes, figures were collected which
are shown in table 12. These figures are based on the last vernal and
first autumnal occurrence of a temperature of 32°, for 1903 to 1912
inclusive, without regard to the reports of frost made by the voluntary
observers at these stations. In figure 15 the length of the frostless
season at the several stations is graphically shown by horizontal lines,
and the limits of the frostless season for the Santa Catalinas (for 1909
to 1914) are shown by oblique lines.

Spring opens at an earlier date at 3,000 and 4,000 feet on the Santa
Catalinas than it does at Tucson and Benson, but at 4,000 and 5,000
feet it does not open at so early a date as it does at Cochise and Fort
Huachuca. At all four of the elevations mentioned the close of the

TABLE 12. — The altitudinal shortening of the frostless season in southeastern Arizona, as
shown by the dates of the last spring occurrence and the first autumn occurrence of a tempera-
ture of 32° at eight stations at graduated altitudes, in the decade of 1903 to 1912,

Station.

Last 32°
in spring.

First 32°
in autumn.

Length of
frostless
season (days).

Yuma, 141 feet

Jan. 28

Dec. 18

325

Phoenix, 1,108 feet

Feb. 1

Dec. 18

321

Tucson, 2,390 feet

Mar. 15

Nov. 19

248

Benson, 3,523 feet

Mar. 26

Nov. 7

226

Cochise, 4,219 feet

Mar. 10

Nov. 6

241

Fort Huachuca, 5,100 feet

Mar. 26

Nov. 8

227

Flagstaff, 6,907 feet

June 11

Sept. 24

105

Chlarson's Mill, 7,200 feet *

May 13 f

Oct. 19 f

159

* The elevation of Chlarson's Mill is reputed by the proprietor to be 8,000 feet, and it is so
stated in the publications of the Weather Bureau. Several aneroid determinations by the writer
indicate that it is approximately 7,200 feet.

t These dates are based on an incomplete record.

growing season comes sooner at the valley stations than it does on the
mountains. The length of the frostless season at Flagstaff is notably
shorter than it is at the same elevation in the mountains.

The advent of spring is retarded at Tucson and Benson by the cold-
air drainage of the Santa Cruz and San Pedro rivers respectively.
Cochise is situated in the middle of the eastern bajada of the Dragoon
Mountains and Fort Huachuca at the top of the northern bajada of
the Huachuca Mountains. Each of these stations is therefore removed
from the operation of cold-air drainage, as is manifested by the failure
of their greater altitude to impose upon them shorter frostless seasons
than those of Tucson and Benson.

The length of the frostless season at Marshall Gulch and at the
similarly situated mountain station at Chlarson's Mill (Pinaleno Moun-
tains) is greater than at Flagstaff, which is at a slightly lower altitude.
This is to be attributed partly to the higher latitude of Flagstaff, but

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

75

chiefly to its location on an extensive elevated plateau and to its
proximity to the cold-air flow from the San Francisco Peaks and other
neighboring elevations.

It may be said, in general, that the frostless season is longer on the
ridges of an isolated mountain than it is in adjacent valleys at the
same elevations. Although the advent of spring at Cochise and Fort
Huachuca is earlier than on the mountains, the arrival of autumn is
earlier also, so that these two stations show an equality in length of
frostless season with the mountain ridges without a correspondence
with them in the dates of commencement and close.

NORMAL ALTITUDINAL TEMPERATURE GRADIENT.

The temperatures which have been secured on the Santa Catalinas
do not form an altogether satisfactory basis for the determination of

TABLE 13. — Average differences between all observed minimum temperatures at stations situ-
ated on ridges in the Santa Catalina Mountains and the minima for the same days or
periods at the Desert Laboratory.

Station.

1911.

1912.

1913.

1914.

Average.

4,000 feet...

+7.2

+ .1

1.6

+ 1.9

5,000 feet...

8.1

8.1

6,000 feet . . .

12.5

7.5

7 6

9.2

7,000 feet . . .

16 1

11 1

14 0

13.7

8,000 feet...

20 1

20 1

9,000 feet...

31 1

20 8

25 9

[The plus sign indicates a higher temperature at the mountain station.]

the normal altitudinal gradient, a datum which should be derived from
extended series of mean temperatures. However, in the absence of an
ideal collection of records from the several altitudes these imperfect

3.000

10

4,000

5,000

6,000

7,000

8,000

9.000

10,000

I

I

I

30

FIG. 16. — Graph showing altitudinal fall in temperature in the Santa Catalina Mountains
(A), and lines showing rate of fall in free air (E) on Pike's Peak (C), on the Sierra
Nevada (D), and average rate for 17 extra-tropical mountains (B).

data have been used as the basis of an approximate determination of
the gradient of fall of temperature with increase of altitude. The

76

VEGETATION OF A DESERT MOUNTAIN RANGE.

minimum temperatures at the several stations are the only ones that
have been used, and they have been in all cases compared with the
minimum for the same period at the Desert Laboratory. The average
differences between the minima of the mountain stations and those of
the Laboratory are shown in table 13, and it is these differences that

TABLE 14. — Daily maximum and minimum temperatures at Marshall Gulch (7,600 feet) and
at the Desert Laboratory for June, July, and August 1911.

Day

of
month.

June.

July.

August.

Minima.

Maxima.

Minima.

Maxima.

Minima.

Maxima.

M.G.

D. L.

M.G.

D. L.

M.G.

D. L.

M.G.

D. L.

M.G.

D. L.

M.G.

D. L.

1st

65
65
69
73

78

80

81
83
95
100
103

98

51
52
48
45
48

76
81

77
76
74

81

77
80
83
81

2d. .

48
46

74
64
70

67

67

3d

106
103

4th

5th

6th ..

46

103

7th

48
50
48
50

48
50
50
49
54

52
52
49
50
49

52
51
46
48

80
80
74
76

77
79
78
76
81

82
73
73
79
71

68
69
67
74

84
74
78
74

73
70
75
76
81

74
71
74
76
75

63
61
62

8th

48
48
44

46
48
49
41
41

41

75
76

73

76
73

77
74
74

70
69
71
81
73

75

78
73
76
73

73
75

78
78

74

77

106
103
104

97
96
85
98
97

98
102
106
101
99

106
104
104
106
105

108
108
108
100
86

9th

91
94

89
102
105
107
110

109
107
105
107
100

88
93
94

10th. ..

79

67
66
59
70

74

76

llth
12th
1 3th

52
49

73
73

62
73

92

98

14th.. ..

15th.

16th
17th

52
51

80

71

73
65
63
64
64

65

70
66
63
65

61
67
73
67
78

82

102
92
95
97
94

94
99
101
100
94

96

89

18th

39
40
41

41
41
40
46
49

43
48
52
46

79
74
72

81
81
83
83
85

85

84
84
76

19th
20th

21st
22d
23d

24th

54
52

52

48
49
50
49

53

47
53
53
52

50

76
79

74
71
76
78
76

78
67
72
76
78

77

25th

26th

27th

28th

29th

100
104

30th

31st ....

Average j 21 d 30.5o

difference] J

20 days, 25.0°

20 days, 24.1°

21 days, 29.4°

23 days, 26.1°

17 days, 27.3°

have been used in the construction of the graph in figure 16, which
shows the gradient for the Santa Catalinas, the average gradient for
17 mountain ranges in different portions of the world (according to
Hann), the gradient for Pike's Peak, the Sierra Nevada, and also the
gradient in the free air, as determined at the Blue Hill Observatory.

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

77

Some estimate of the error involved in basing the gradient only on
minimum readings may be made by the figures presented in tables
14 and 15. These tables exhibit the only daily records of maximum and
minimum temperatures for the Catalinas for any period longer than
a few days. The average maxima and minima for Marshall Gulch for
the months of June, July, and August have been contrasted in table 14,
with the average maxima and minima for the Laboratory. During
June the apartness of the minima was 30.5°, of the maxima 25.0°. In
July this relation was reversed, the apartness of the minima being
24.1°, that of the maxima 29.4°, while for August the two were more
nearly the same, the apartness of maxima being 26.1°, of maxima 27.3°.
The facts that the minimum temperatures of valley and mountain
were further apart than the maxima were during June, and not so
far apart in July and August, may be connected in some manner with
the clear dry weather of June and the rainy, cloudy character of July
and August. However, the data in table 15, showing the maximum

TABLE 15. — Daily record of maximum and minimum temperatures for a portion of June
1912, at summit of Mount Lemmon, with corresponding data for the Desert Laboratory.

Day of month.

Minimum.

Maximum.

Day of month.

Minimum.

Maximum.

M. L.

D. L.

M. L.

D. L.

M. L.

D. L.

M. L.

D. L.

June 7

op

46
46
45
40.5
42
41.5

°F.

°F.
73
77
72
67
69
69.5

°F.

June 13

o p

43
43.5
45
44
41
41

o p

74
71
76

78

77
72

op

68
66
69
71
70
69

op

101
104
105
102
102
106

8

14

9

15

10

102
101
100

16

11

70
71

17

12

18

Average difference of maxima, 33.9; of minima, 31.1.

and minimum temperatures on Mount Lemmon in June 1912, indi-
cate that there was a greater apartness of maxima than of minima
when these data are averaged and contrasted with those for the Desert
Laboratory.

It would require a much fuller mass of figures than are in hand to
make any attempt at an explanation of the differences that exist in
the apartness of desert and mountain maxima and minima in different
localities and different months. For the present purpose it is instruc-
tive to average the entire set of apartnesses for Marshall Gulch for
June, July, and August 1911, which gives an average difference of
minima of 80.7°, and of maxima of 81.7°. In other words, throughout
a series of several months there is doubtless a swamping of the irregu-
larity of the apartnesses for individual months. If, then, there is an
average equality between the apartness of maxima and minima — which
is to say that there is an equality of daily mean range of temperature
between desert and mountain — it would indicate that the minima are

78

VEGETATION OF A DESERT MOUNTAIN RANGE.

just as good data on which to base an estimation of the gradient as
mean temperatures would be. Since the figures used in constructing
the gradient were secured in all months from April to October, in
several years, and in a wide diversity of localities — all outside the
influence of cold-air drainage — it is probable that the gradient here
presented, figure 16, is within one or two degrees of the same measure
of accuracy that could be secured by a long series of consecutive read-
ings of maximum and minimum.

Using the elevation of the Desert Laboratory (2,663 feet) as a base,
the actual fall of temperature between that locality and the 9,000-foot
station on the Santa Catalinas is at the rate of 4.11° per 1,000 feet.
The gradients between the several mountain stations are indicated in

60

SO

30

10 v*

FEB. MAR.

MAY JUNE JULY AUG. SEPT. OCT.

FIG. 17. — Graphs showing march of weekly minimum temperature at
Desert Laboratory (upper) and weekly or other minimum tempera-
ture at Marshall Gulch for 1911 (lower).

figure 16, on which it will be seen that there is a negative gradient
between the Desert Laboratory and the 4,000-foot station, and that
the gradient between 4,000 and 5,000 feet is at the rate of 10° per 1,000
feet. From 5,000 feet upward the gradients are more uniform and
more nearly equal between the successive stations.

Figures are given by Hann * for the gradients of temperature on a
number of mountains in different parts of the world. The average
values for 17 extra-tropical mountains is 3.13° F. per 1,000 feet (0.57° C.
per 100 m.). The only mountains in the western United States for
which Hann gives figures are Pike's Peak and Sierra Nevada (Coif ax,
Placer County). The gradient of the former is 3.46° per 1,000 feet

* Hann, J. Handbook of Climatology. Translation by R. DeC. Ward, pp., 243-246. New
York, 1903.

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 79

(0.63° C. per 100 m.), and for the latter 4.12° per 1,000 feet (0.75° C.
per 100 m.). The gradient for Coif ax is seen to be almost exactly
coincident with the entire gradient for the Santa Catalinas. It is
interesting to note, in this connection, that the fall of temperature in
the free air has been determined at the Blue Hill Observatory to be
2.5° F. per 1,000 feet, which is far more gradual than any of the moun-
tain gradients that have been cited. The Blue Hill data apply only
to low elevations, but are in substantial agreement with figures more
recently secured in the free air at Avalon, California.* Seven balloon
ascensions from Avalon to elevations of 18 km. and higher showed a
mean gradient of fall in the first 3 km. (9,842 ft.) of 2.2° per 1,000 feet.
These two determinations of the free-air gradient indicate a conserva-
tism of temperature change in the lower atmosphere as compared
with the changes on the slopes of mountains.

While the normal temperature gradient is of profound interest from
the standpoint of pure climatology, it is nevertheless of subsidiary
value in the study of climate in relation to vegetation. Its chief value
is as a basis with which to compare the differentiation of temperature
conditions originating in the irregularities of topography and other
causes. In later pages the subsidiary influences upon the temperature
gradient will be discussed.

THE ABSOLUTE MINIMUM OF WINTER.

The absolute minimum temperature of the winter was secured at
5,300 feet and at 7,600 feet for four winters, and during the winter of
1912 and 1913 was secured at four stations, and during the succeeding
winter at 10 stations, differing both in altitude and in topographic
location.

The winter of 1912 and 1913 was one of exceptional severity at
Tucson — in fact throughout the extreme southwestern United States —
while the winter of 1913 and 1914 was one of the customary modera-
tion. The data for these two winters are calculated, therefore, to
exhibit the extreme and the average conditions of winter temperature
for stations in Arizona.

The minimum temperature readings at the mountain stations are
given in table 16; and in table 17 are given the minima for December,
January, and February of the same years for a selected series of stations
in Arizona. The lowest temperature recorded on the Santa Catalinas
in 1912-13 was —6° at 6,000 feet, while the lowest temperature at
the highest station, at 7,600 feet, was —2°. This figure should be con-
trasted both with the absolute minimum at the Arizona Experiment
Station, 6°, and with that at the office of the Desert Laboratory, 1°,
as well as with that for Flagstaff,— 23°, situated in northern central

* Blair, William R. Free-Air Data in Southern California, July and August, 1913. Mo.
Weather Rev., 42:410-426, 1914.

80

VEGETATION OF A DESERT MOUNTAIN RANGE.

Arizona at an elevation 700 feet lower than that of the 7,600-foot
Station in Marshall Gulch. The extremely low temperatures at the
Arizona Experiment Station and the office of the Desert Laboratory,
as contrasted with the minimum temperature of 17° at the Desert

TABLE 16. — Absolute minimum temperatures of two winters at stations in the Santa Catalina

Mountains and at Tucson.

Station.

Location of station.

1912-1913.

1913-1914

Arizona Experiment Station
Office of Desert Laboratory

Near bottom of Santa Cruz Valley .
Edge of flood-plain of Santa Cruz
Valley

6
1

26
22

Desert Laboratory

Slope of hill, 335 feet above Santa

Cruz Valley

17

29

4 000 feet .

Ridge, in Lower Desert Region. . . .

13

27

5 000 R

Slope, 125 feet above floor of Soldier

Canon

18

5 000 V

Floor of Soldier Canon, in Lower

Encinal ...

18

6 000 R

Ridge in Upper Encinal

— 6

18

6 000 V

Floor of Bear Canon

6

7 000

Ridge, north rim of Bear Canon, in

Pine Forest

.5

12

7 600 .

Bottom of Marshall Gulch, in Fir

Forest

— 2

15.5

8 000

Ridge, north rim of Marshall Gulch,

Pine Forest

5

9 000

North slope, below summit of Mount

9 225 .

Lemmon in heavy Fir Forest ....
Top of fire-tower on summit of

5

Mount Lemmon

3

Laboratory, are to be accounted for through the operation of cold-air
drainage. On the coldest night in January 1913 there was a difference
of only 3° between the temperatures of the Santa Cruz Valley and
Marshall Gulch, 5,400 vertical feet apart. It is desired here not so

TABLE 17. — Absolute minimum temperatures of winter months for two winters at selected
stations in Arizona, together with the absolute winter minima at 7,600 feet in the Santa
Catalina Mountains.

Station.

1912-1913.

1913-1914.

Dec.

Jan.

Feb.

Dec.

Jan.

Feb.

Santa Catalinas 7,600 feet . . .

- 2
6

15.5
26

Tucson 2 390 feet

18
15
0
- 7
- 2
21
- 4

28

28

26

Chlarson's Mill 7 200 feet

Flagstaff 6 907 feet .

-23
-27
- 5

1
- 8

- 7
— 14
5
21
-11

- 5

-17
3
26
10

- 1
-13
- 1
23

7

2
- 4
14
29
9

Fort Valley Experiment Station, 7,300 feet

Fort Apache 5 200 feet

Fort Huachuca 5 100 feet

Snowflake 5 644 feet

much to lay emphasis on the exceptional coldness of the Santa Cruz
valley, but on the exceptional warmness of the summit of the Santa
Catalinas, as contrasted with similar elevations in Arizona which are

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 81

very dissimilar in their topographic location. In table 17 it will be
seen that the winter minimum at Flagstaff, in northern central Arizona,
was 21° below the minimum for Marshall Gulch, although Flagstaff is
located 700 feet lower. This relation is similar to that which exists
between the length of f restless season at Marshall Gulch and at Flag-
staff, and is due to the facts mentioned in that connection on page 74.
At the Fort Valley Experiment Station, located in the vicinity of
Flagstaff and nearer to the San Francisco Peaks, the winter minimum
was 4° lower than at Flagstaff. At Snowflake, on the extensive Mogol-
lon Plateau, and at Fort Apache, in the cold-air drainage of one of the
main forks of the Salt River, there were also minima which were
respectively 9° and 3° lower than at Marshall Gulch, although these
stations are respectively 2,000 and 2,400 feet lower than Marshall
Gulch. Fort Huachuca is located at the base of the Huachuca Moun-
tains in such a manner as to escape cold-air drainage from any of the
large canons of that range of mountains, and its absolute minimum was
6° higher than that of Fort Apache, which is of approximately the same
elevation. Chlarson's Mill is situated in Frye Canon in the Pinaleno
(Graham) Mountains, surrounded by heavily forested slopes. Its loca-
tion is analogous to that of Marshall Gulch, being similarly situated
in an isolated desert mountain and surrounded by heavily forested
slopes. The single monthly minimum available for Chlarson's Mill
is 15°, for a month in which the minimum at Tucson was 18°, while
it was 7° for the Fort Valley Experiment Station, at almost the same
elevation as Chlarson's Mill.

An inspection of the absolute minima for 1913-14, in table 17, will
show that the same relations hold true between the several stations
that have just been discussed. The winter minimum for Tucson, 26°,
was much higher than in the preceding winter, and so was that for
Marshall Gulch, 15.5°, although the absolute minimum in the new
station on the fire tower at Mount Lemmon was 3°, and in the heavy
timber on the north face of Mount Lemmon was 5°.

The data just discussed amply bear out the statement that the lowest
temperatures of winter are less severe on the Santa Catalinas than
they are at the same elevation on the plateau of north-central Arizona,
and even less severe than they are in many situations of lower altitude.
The fragmentary records of earlier years at Chlarson's Mill, which
are not given here but are available in publications of the Weather
Bureau, show that it is likewise favored by lower winter temperatures
than are the plateau stations of the same elevation in Arizona. The
length of the frostless season has just been shown to be less in the Santa
Catalinas and in the Pinaleno Mountains than at Flagstaff. In short,
the indications are very strong that all phases of winter temperature
conditions are less severe on the small and isolated desert mountains
than on the plateaus and highlands of the same elevations and of nearly

82 VEGETATION OF A DESERT MOUNTAIN RANGE.

the same latitude. The radiation of the desert valleys and the diurnal
convection currents of warm air are not without a strong ameliorating
influence on the climate of elevated but small masses of land.

DEPARTURES FROM THE NORMAL TEMPERATURE GRADIENT DUE TO

COLD-AIR DRAINAGE.

The ideal conditions of uniform decrease of temperature with increase
of altitude are seldom actually encountered in nature, at least not in
mountains of small size and rugged topography. The principal factor
which brings about departures from the normal or ideal gradient of
fall is the operation of cold-air drainage, or inversion of temperatures.
This is a phenomenon which has long been known and has frequently
been discussed with respect to its influence on vegetation. In an
earlier paper * the writer has described some observations of tempera-
ture inversions at the Desert Laboratory and in the Santa Catalinas,
and has pointed out the causes involved in making cold-air drainage
much more pronounced in deserts than it is in humid and forested regions.

The scanty vegetation of the desert subjects its soil, rocks, and sands
to full insolation and to a pronounced heating throughout the day.
The dark rocks of Tumamoc and other volcanic hills in its vicinity
become so hot during the long clear days of May and June that it is
impossible to hold one's hand on them without pain. During the day
there is a constant and active radiation of heat from the rocks and soil,
which warms the lowest layers of air and causes a convectional heating
of the lowest portion of the atmosphere. Immediately after sunset
the warmed surfaces become rapidly cooler and the rate of radiation
is quickly reduced. The air nearest the cooling rocks and soil becomes
cooler than the air above it, and consequently begins to fall by gravity
before there is opportunity for it to mix with the warmer air above.
This cooled air descends from hillsides and even from gentle slopes
and soon collects in valleys and depressions, where it results in a slowly
or rapidly moving mass of air which is appreciably cooler to the senses
than is the air of the slopes or hillsides. The inversion of temperature
thus caused usually reaches its maximum during the first half of the
night, although this is determined in great measure by the size of the
drainage area.

It is only on clear and still nights that cold-air drainage operates
in the most pronounced manner. A high wind will disturb the flow
or completely eliminate it. Heavy cloudiness will cause the rate of
radiation to lag so that there is time for an admixture of cool and warm
air, thereby preventing the flow of cold air or greatly reducing it.

On clear nights which follow heavy rains the inversion of temperature
will be reduced to a negligible amount, because of the increase of the
specific heat of the soil brought about by its becoming wet.

* Shreve, Forrest. Cold Air Drainage. The Plant World, 15: 110-115, 1912.

CLIMATE OF THE SANTA CATALINA MOUNTAINS.

83

It has been shown in the paper to which reference has been made
that the Desert Laboratory is situated well above the level of the
cold-air flow of the Santa Cruz Valley, 335 feet below it. The greatest
observed difference of minimum temperature in a single night between
the Laboratory and the Valley was 24°, and the greatest difference
between the mean monthly minima of the two localities was 17.8° for
May. During the humid mid-summer the mean monthly difference
falls to 8° and 9° for these stations.

TABLE 18. — Minimum temperature records to show the operation of cold-air drainage in the
open vegetation of Soldier Canon and Bear Canon and its abeyance in the heavy forest of
Marshall Gulch.

In each case one record is from the floor of the canon and the other from its slopes or rim. The
minus differences indicate a higher temperature on the floor and the absence of cold-air
drainage.

Dates.

Slope or
rim.

Floor.

Difference.

Soldier Canon, floor at 4,900 feet, slope at 5,025 feet:
Sept. 27, 1913, to May 16, 1914

18

18

0

May 17 to 19, 1914

49

42.5

6.5

May 20, 1914 . . . . . .

52

44

8

May 21 to July 22, 1914

45

39

6

July 23 to 28, 1914

60

60

0

July 29 to Aug. 8, 1914

63

62

1

Aug. 9 to Oct. 10, 1914. . .

40

40

0

Bear Canon, floor at 6,000 feet, rim at 7,000 feet:
Sept. 24 to Sept. 26, 1913

38

33

5

Sept. 27, 1913, to May 17, 1914

12

6

6

May 18 to 19, 1914

47

38

9

May 20 to July 23, 1914

38

34

4

July 24 to 27, 1914

52

54

— 2

Marshall Gulch, bottom at 7,600 feet, rim at 8,000 feet:
Sept. 25, 1913

34.5

33.5

1

Sept. 26, 1913

30.6

31.5

- 1

Sept. 27, 1913, to May 17, 1914

5

15.5

— 10.5

May 18 to July 24, 1914

33.5

51.5

-18

July 25, 1914

50.5

49.5

1

July 26, 1914

48.5

50.5

— 2

July 27, 1914

51.5

51.5

0

July 28 to Oct. 11, 1914

29.5

32.5

— 3

The vigor of cold-air drainage is determined not only by the condi-
tions of cloudiness and wind but also by the size and nature of the area
from which the cold air is derived and by the character of the valley
bottom through which it moves. In the Santa Cruz Valley cold air
is derived from an area of more than 1,000 square miles, resulting in
the pronounced low temperatures shown in tables 16 and 18. The
broad level trough of the valley is conducive to a slow movement of
the air, and the nocturnal minimum is usually reached during the last
hours of darkness. The valleys of the Salt and Gila Rivers are larger
than the valley of the Santa Cruz, and they have their sources in still
higher mountains, but they do not seem to possess a well-marked

84 VEGETATION OF A DESERT MOUNTAIN RANGE.

cold-air drainage, to judge by the minimum temperature records for
the towns on the lower courses of these rivers, as Florence and Phcenix.
It is possible that in traveling long distances at nearly the same altitude
the cold air is gradually warmed by mixture with the warm air above it.

During the winter of 1913-14 and the summer of 1914 thermometers
were exposed in the Santa Catalinas so as to give a basis for comparing
the cold-air drainages of the mountain with the drainage of the Santa
Cruz Valley as investigated at the Desert Laboratory. Table 18 shows
the readings of instruments placed so as to reveal the differences of
temperature due to cold-air drainage, at three localities of different
elevation. The first locality is in Soldier Canon, in the open Encinal,
where readings were taken on the floor of the canon, at 4,900 feet and
on its slope at 5,025 feet. The lowest minima of the winter were
identical at the two stations, which can be accounted for only on the
possibility of the stream of cold air having become so deep as to reach
the upper thermometer, or else on the possibility that the lowest mini-
mum of the winter occurred on a cloudy or very windy night. During
three intervals in the arid fore-summer the depression of the temperature
in the floor of the canon was 6.5°, 8°, and 6° respectively, whereas through-
out the humid mid-summer the depression was absent or negligible.

The regular 7,000-foot station is located on the rim of Bear Canon,
and the data from it may be compared with those from an instrument
placed in the floor of the canon 1,000 feet below. This station may
further be compared with the 6,000-foot station located on the summit
of Manzanita Ridge. In spite of the difference of 1,000 feet in the
elevation of the two thermometers in Bear Canon, the lowest tempera-
ture of the winter was 6° in the Canon and 12° on the rim. A two-night
interval in the autumn of 1913 gave a difference of 5° between these
stations, due to air drainage, and during the arid fore-summer of 1914
differences of 9° and 4° were obtained. During the three particularly
cloudy and rainy nights in July there was an actual reversal of the
conditions of cold-air drainage and a manifestation of the true tempera-
ture conditions to be expected from altitude alone, the rim having a
minimum 2° lower than the floor. Although the winter minimum of
1913-14 was 6° in the floor of Bear Canon at 6,000 feet, it was 18° on
Manzanita Ridge at 6,000 feet. This difference of 12° between two
stations at the same altitude is as great as should be expected, under
the operation of the normal gradient of temperature, between localities
3,468 vertical feet apart (12° -f- 3.46°, the rate of fall per 1,000 feet).

Although there are some small bodies of forest on the walls of Bear
Canon and many scattered trees, nevertheless the surface of its sides
is largely occupied by cliffs and boulders (see plate 21), and these are
responsible for the acute operation of the drainage phenomenon.

Several preliminary tests had shown an extremely weak manifesta-
tion of cold-air drainage at the heavily forested elevations of the Santa

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 85

Catalinas. The data presented in table 18 for the rim of Marshall
Gulch at 8,000 feet, and for the Montane Garden at 7,600 feet, bear
out the results of the preliminary tests. The rim was 10° colder than
the bottom of the gulch in the over- winter period and 18° colder in
the period from May 18 to July 24. The 5 one-night readings and the
readings for the period from July 28 to October 11 all show an equality
or a slight difference, more often a difference, with the temperature
more commonly higher in the bottom of the Gulch than on the rim.
In other words, cold-air drainage is in abeyance at this locality.
Whether this is invariably the case can only be stated after further
instrumentation and after complete assurance that the readings have
not been influenced by the character of the weather during the nights
of lowest temperature. The heavy vegetation of the Forest region,
together with the high humus content of the soil and the litter of leaves
by which it is covered, all militate against the rapid terrestrial radia-
tion in which cold-air drainage has its origin. It will not be surprising,
therefore, to find that the phenomenon is either very weak or absent
above the elevation of 7,000 feet in such portions of the mountain as
are forested. The elimination of cold-air drainage by a forest cover
can take place only in small mountains which are forested to the sum-
mit. A large mountain mass, an extremely steep mountain side, or
an extensive area lying above timber line will cause a flow of cold air
down through forested areas below. This is exemplified at the San
Francisco Peaks, Arizona.

The case mentioned in which cold-air drainage occasioned a differ-
ence of temperature at the same altitude, which was the equivalent of
nearly 3,500 feet, probably represents its maximum effect. The differ-
ence of 8° in Soldier Canon is the equivalent of an altitudinal difference
of about 2,200 feet.

The influence which cold-air drainage exerts on vegetation is regis-
tered chiefly in the shortening of the season of vegetative activity on
the floor of a canon as contrasted with its sides. This effect has been
repeatedly observed in Bear Canon, where the oaks on the floor of
the canon are always far behind the individuals on the canon wall in
the advancement of their new foliage in the spring. Likewise in the
autumn the frost-killing of herbaceous perennials and of the leaves of
Prunus, Rhus, and Populus jamesii takes place in the floor of the canon,
while the herbaceous plants of the slopes are still green and active.
The plants on the canon floor are, in other words, subjected to a grow-
ing season similar to that usually found at a much higher altitude.

The influence of cold-air drainage in determining the distribution of
plants is likewise marked. It is not wholly responsible for the fact
that mountain species extend down the canons to lower altitudes than
they assume on slopes or ridges, for the influence of ground water and
soil moisture is very potent in this connection. The occurrence of the

86

VEGETATION OF A DESERT MOUNTAIN RANGE.

highest individuals of every species — other than those of aquatic or
streamside habitat' — on or near the summits of ridges, and their invari-
able absence from the bottoms of canons at these higher elevations,
are to be attributed to the absence of cold-air drainage from the ridges
and higher slopes, together with the influences grouped in the "factor"
of slope exposure.

SOIL TEMPERATURE.

In the autumnof 191 3 instruments were placed at the 6,000- and 8,000-
foot stations to secure the absolute winter minimum temperature of
the soil at a depth of 3 cm., and the thermometers were maintained in

TABLE 19. — Minimum temperatures of the soil and of the air at 3 elevations in the
Santa Catalina Mountains for irregular periods.

Station.

Air

temperature.

Soil
temperature.

Difference.

At 6,000 feet:
Sept. 23 to Sept. 27, 1913, . ...

42

47

+ 5

Sept. 28, 1913, to May 16, 1914
May 17 to 19

18
53

28
51

+ 10
— 2

May 20 to July 22

44

45

+ 1

July 23 to 27

58

60

+ 2

July 28 to Oct. 10

40

42

+ 2

At 8,000 feet:
Sept. 25, 1913

34.5

41

+ 6.5

Sept. 26, 1913

30.5

37

+ 65

Sept. 27, 1913, to May 17, 1914
May 18

5

39.5

30
43

+25.0
+ 3.5

May 19 to July 24

33.5

41

+ 7.5

July 25

50.5

69

+ 18 5

July 26

48.5

58

+ 96

July 27

51 5

57

+ 55

July 28 to Oct. 11

29.5

39

+ 9.5

At 7,600 feet:
Sept. 25, 1913

33.5

39

+ 9

place and read at irregular intervals during the summer of 1914. The
object in placing the thermometers at so slight a depth was to obtain
a measure of the activity of terrestrial radiation by a comparison of the
superficial minima of the soil and the atmospheric minima. The ordi-
nary type of Six's thermometer was used, buried in a wooden box and
covered with earth. The readings secured in this manner and the
readings of atmospheric minima for the corresponding periods at the
same stations are given in table 19.

At the 6,000-foot station the soil minima are higher than the air
minima in every case except one, the over-winter difference being 10°.
At the 8,000-foot station all of the 9 readings secured show a higher
minimum for the soil. The over- winter period shows a difference of
25°, and the night of July 25 shows a difference of 18.5°. The readings

CLIMATE OF THE SANTA CATALINA MOUNTAINS. 87

show that the soil temperatures at 6,000 feet were cooler in general,
in terms of the air temperature, than were those of the 8,000-foot
station. This difference is not to be attributed to the difference of
elevation so much as to the naked and stony character of the soil at
the 6,000-foot station and the relatively abundant humus and litter in
the surface soil at 8,000 feet. In short, the radiation from the soil
surfaces in the Encinal is greater than it is in the Forest, as has been
already discovered from the difference in the behavior of cold-air
drainage in these two regions. There is also a slight indication that
the differences between the air and soil minima are least in the dry
seasons of May and September, which is again in keeping with the
greater radiation exhibited in dry soils as compared with wet ones.

On the night of September 25, 1913, the difference between the air
and soil temperatures was simultaneously determined on the rim of
Marshall Gulch at the 8,000-foot station and in a thicket of young fir
trees in the bottom of the gulch. The soil remained 6.5° warmer than
the air on the rim of the gulch and 9° warmer in the fir thicket, showing
the degree to which a heavy cover of vegetation retards radiation and
conserves the warmth of the soil. On this night the air temperature
in the bottom of the gulch was 1° lower than that on the rim (see
table 18).

One of the most striking features of the soil minima is the fact that
although the air temperature at 8,000 feet fell to 5° in the winter of
1913-14, the soil temperature fell only to 30°. This means that in the open
forest on the rim of Marshall Gulch the soil must have been very slightly
if at all frozen in the winter in question, which was apparently a winter
of about average severity. During the same winter a lower absolute
minimum of the soil was recorded at 6,000 feet than at 8,000 feet. In
shaded situations and on north slopes in the Fir Forest the soil undoubt-
edly freezes to a slight depth. Inasmuch as no soil temperatures have
been secured with the bulb of the thermometer in contact with the
soil, and no readings have been secured at a greater depth than 3 cm.,
the further discussion of soil temperature conditions in these moun-
tains should await further investigation.

88 VEGETATION OF A DESERT MOUNTAIN RANGE.

CORRELATION OF VEGETATION AND CLIMATE IN THE
SANTA CATALINA MOUNTANS.

The earlier chapters of the present paper have described the salient
features of the vertical distribution of vegetation in the Santa Catalinas,
and also some of the principal gradients of climatic change. Both the
vegetation and the climate have been shown to exhibit progressive
changes with increase of altitude, and these changes have been found
to undergo hastening or retardation under the influence of topographic
irregularities. It will be the object of the following pages to correlate,
in so far as possible, the altitudinal changes of vegetation and climate,
in an effort to determine roughly some of the physical factors which
are of critical importance in limiting the vertical ranges of the types
of vegetation and of their characteristic species.

THE NORMAL ALTITUDINAL GRADIENT OF VEGETATION.

Any attempt to ascribe vertical limits to the Desert, Encinal, and
Forest, or to state the vertical limits of individual species, is met at
once by the omnipresent importance of slope exposure in determining
these limits. The altitudinal range of vegetations and species may be
determined by examining only slopes of south exposure, or only those
of north exposure, and the two examinations would agree closely as
respects the vertical ranges, but would disagree by approximately
1,000 feet with respect to the upper and lower limits of the vegetations
or species. It is impossible to determine the normal character of vege-
tation at a given altitude by seeking level ground, for it will be found
only in the flood-plains, subject to the influence of a high soil moisture,
or on a ridge, subject to equally special conditions. It is also impos-
sible to visit adjacent valleys or plateaus lying at the same elevation
and to find on them vegetation which is subject to the same climatic
and soil conditions. For some purposes it is desirable to consider the
vertical stages of vegetation under ideal conditions, as affected by
altitude without the complications due to topographic features. It
is then possible to hypothecate a norm of vertical stages of vegetation
by averaging the altitude of any given limit as separately determined on
north and south slopes, or it is possible to take into consideration only
the altitudinal changes of south slopes or of north slopes, taken alone.

It has been shown that the influence of topography on the vegetation
is chiefly (sometimes solely) to carry the common types of vegetation
above or below the elevations at which they are universal. The
influence of topography on the gradients of climate is of the same char-
acter; the topographic relief causes no wholly new factors to come into
play, but serves merely to carry the physical conditions of the Desert
into the Encinal, for example, or to bring the conditions of the Forest

CORRELATION OF VEGETATION AND CLIMATE. 89

down into the Encinal. The physical factors which underlie the effects
of topography are, then, to be considered simply as special cases of the
same influences that are grouped in the effects of altitude itself. It
is desirable, nevertheless, in studying the correlation of climate and
vegetation to consider separately the normal gradient of vegetation
and the departures from the normal gradient.

THE VERTICAL DISTRIBUTION OF INDIVIDUAL SPECIES.

The student of vegetation too often loses sight of the fact that vege-
tation is composed of individual species of plants and that the behavior
of the vegetation is a function of the behaviors of these species. After
our review of the vegetation of the Santa Catalinas, and in connection
with the discussion of its control by climatic factors, it is necessary
to consider the vertical distribution of the individual species in relation
to the physical conditions of the mountain.

There are no species of plants which grow spontaneously both at
the base and the summit of the Santa Catalina Mountains, except a
few palustrine forms of Carex and Juncus. The total range of physical
conditions through the 6,000 feet of elevation here involved is so great
that no native plant possesses the power of accommodation to the
complete gamut of Desert, Encinal, and Forest. Indeed, very few
plants range through half of the entire gradient of conditions, in any
portion of it. The species which exhibit the widest belts of vertical
distribution are to be found in the most dissimilar habitats at the lower
and upper edges of their ranges, which indicates that these species are
not really capable of existence through 2,000 or 3,000 vertical feet of
the climatic gradient under the same conditions of topographic loca-
tion, slope exposure, and insolation. In fact, a close analysis of the
habitats occupied by characteristic plants, in connection with their
vertical ranges, indicates that, below 6,000 or 7,000 feet, no plants
outside the desert succulents and semi-succulents range through more
than 1,000 to 1,500 feet in habitats of the same topographic character.
At higher elevations a number of common plants extend more than
1,500 feet in situations of the same character, as for example Pinus
arizonica, which ranges through nearly twice that altitude on dry
southern slopes.

A vertical range of 4,700 feet is exhibited by Vitis arizonica, which
occurs in several arroyos and canons at 3,000 feet and is found in the
same habitat throughout the Desert and Encinal regions of the moun-
tain, reaching its highest observed station at 7,700 feet in a steep dry
arroyo in the Pine Forest. Although the habitat of Vitis is superficially
identical throughout its range, it is found at 3,000 to 5,000 feet only
in the largest arroyos, in which it is able to draw upon much greater
and more constant supplies of soil moisture than are available in the
small arroyos to which it is confined at the upper edge of its range.

90 VEGETATION OF A DESERT MOUNTAIN RANGE.

Robinia neomexicana ranges through 3,800 feet, from its lowest occur-
rence on flood-plains near constant water at 5,300 feet, to its highest
occurrence on dry ridges near the summit of Mount Lemmon at 9,100
feet. On the higher mountains of southern Arizona this species ascends
to over 10,000 feet. Amorpha californica, after the manner of Vitis,
ranges 3,500 feet from moist arroyos at 4,200 feet to dry ones at 7,700
feet. Agave palmeri ranges from dry slopes of east or west exposure at
3,200 feet to open ridges and crevices of rock at 7,400 feet, a belt of
4,200 feet. Among the species which reach neither the Bajada nor
the top of the mountain there are no others with vertical extensions
of more than 4,000 feet. An extreme range of 3,700 feet is possessed
by Juniperus pachyphlcea, from northern slopes at 4,200 feet to ridges
at 7,900 feet. Nolina microcarpa extends from 3,750 feet on north
slopes to 7,200 feet in open pine forest, a range of 3,450 feet, and Dasy-
lirion wheeleri from 3,600 feet to 6,600 feet, on opposing slopes, a range
of 3,000 feet.

Among other plants which occur chiefly in the Encinal Region there
are none with vertical ranges in excess of 3,000 feet, few in fact approach
that range. Pinus cembroides extends from north slopes at 5,000 feet
to open rocky ridges at 7,800 feet, a range of 2,800 feet; Agave schottii
ranges from 3,700 to 6,000 feet, a belt of 2,300 feet; Garry a wrightii
ranges from 4,300 to 6,500 feet, an extent of 2,200 feet; and Quercus
emoryi extends from north slopes at 4,300 feet to south slopes at 6,200
feet, a vertical range of 1,900 feet.

Among the plants which have their lowest occurrence in the flood-
plains of the Encinal and their principal range through the Forest
Region, a number have vertical belts of occurrence of more than 3,000
feet. Pinus arizonica itself is found through 3,300 feet and its upper
limit is determined only by the height of the mountain. Pseudotsuga
mucronata is found through 3,100 feet, and is also terminated by the
summit of the mountain. Quercus hypoleuca and Quercus reticulata are
found through nearly 3,000 feet, and this extent of vertical range is
attained by a large number of herbaceous perennials of the Upper
Encinal and Forest.

The Desert species which are encountered at the foot of the mountain
do not begin their vertical ranges at that point, and statements of the
elevations which they reach on the mountain are not to be compared
with figures for the ranges of Encinal and Forest plants. Mamillaria
grahami reaches the attenuated limit of its occurrence at 7,000 feet,
Echinocactus wislizeni at 5,600 feet, Fouquieria splendens at 5,600 feet,
Opuntia versicolor at 5,500 feet, and Carnegiea gigantea at 5,100 feet.
Very few other species of the bajadas and desert hills are found above
4,700 feet.

Among the most restricted vertical ranges of any plants which reach
neither the foot nor the summit of the mountain are those of Quercus

CORRELATION OF VEGETATION AND CLIMATE. 91

oblongifolia and Vauquelinia californica. If we except the occurrence
of the former in the beds of Sabino and Ventana canons at 3,000 to
3,200 feet, its lowest occurrence on slopes is at 3,900 feet and its highest
at 5,600, a range of 1,700 feet; while Vauquelinia ranges from 3,900 to
5,500 feet, a vertical range of only 1,600 feet. These limits also apply
very nearly for Erythrina flabelliformis, Ingenhousia triloba, and several
shrubs and shrublets, and are only slightly exceeded by the range of
Quercus emoryi, which has already been stated to be 1,900 feet.

Certain species of plants are confined to arroyos throughout their
vertical ranges, as are Vitis arizonica, Amorpha californica, Platanus
wrightii, and Juglans rupestris; or are found chiefly in arroyos, as
Cupressus arizonica and Acer interior. The great majority of trees,
shrubs, and shrublets, as well as the semi-succulents (such as Agave,
Yucca, Nolina, and Dasyliriori), are found on slopes and ridges in at
least some portions of their ranges, or are chiefly found there. The
oaks, the deciduous trees, and most of the shrubs may be found along
arroyos, or in flood-plains at elevations from 500 to 1,000 feet below
the level at which they become common components of the slope vege-
tation. The semi-succulents, like the succulents, are rarely found in
arroyos, although they may grow very close to them or may be found
in dry flood-plains. Of all species not confined to arroyos, their lowest
occurrences are generally to be sought on north slopes or in arroyos
at even lower elevations, and their highest occurrences are to be sought
on south or southwestern slopes or (particularly in the case of cacti)
on rocky ridges. At the vertical center of the distributional range of
these species they may be found, as a rule, on slopes of every exposure,
and perhaps in flood-plains as well, particularly in the case of the ever-
green oaks. The exceptions to the rule are Quercus oblongifolia, which
is commoner on south slopes than on north ones at all parts of its
vertical range except the very lowest, and Pinus chihuahuana, which
is rarely found on north slopes at any part of its range, even its lowest
occurrences being on south slopes or on an approximate level.

PHYSICAL FACTORS INVOLVED IN THE DETERMINATION OF THE NORMAL
ALTITUD1NAL GRADIENT OF VEGETATION.

In order to overlook for the moment all of the subsidiary influences
which cause local disturbance of the vegetistic gradient let us consider
that the southern slopes at all elevations are representative of the
normal altitudinal changes of vegetation, and let us then consider
some of the differences of physical conditions that accompany the
ascent from 3,000 to 9,000 feet. The differentiations of vegetation
which we are accustomed to designate as "due to altitude" are actually
due to three groups of physical factors: (a) moisture factors, (6)
temperature factors, (c) light factors. It has been customary to regard
atmospheric pressure as a negligible agency in relation to plants, but

92 VEGETATION OF A DESERT MOUNTAIN RANGE.

there is no work known to the writer which proves or disproves this
view. The light factors have been but little investigated and may
prove to occupy a role of great importance. In spite of the funda-
mental physiological role of light, it is more than probable, however,
that this factor plays a less important part in influencing the distri-
bution of plants than do moisture and temperature.

MOISTURE FACTORS.

In the case of a mountain which arises from an arid region to a
considerable height, the moisture factors are of critical importance in
controlling the vertical distribution of plants. This group of factors
may be defined as those which have to do with the maintenance of a
close degree of equality between the daily intake and outgo of water
through the plant. The description of rainfall and soil moisture con-
ditions in the Santa Catalinas has indicated the nature of the water
supply for plants, and the data on atmospheric evaporation have shown
the collective force of the chief of those ultimate external factors
which determine the water loss of plants. During the humid seasons
the ratio of water available to water lost is such as to make conditions
favorable for all plants. During the most acute periods of aridity the
value of this ratio becomes an item of the first moment.

The soil moisture data given in an earlier section are from too slight
a depth to indicate the possible supplies for trees and the largest shrubs.
They are nevertheless from a depth which is freely exploited by the
roots of perennial plants, and it is more than likely that they bear a
rather definite ratio to the moisture at greater depths.

Since the arid fore-summer is the portion of the year in which the
maintenance of an equilibrium between intake and outgo of water is
most difficult, it is instructive to determine their relation for this season
at the different altitudes. This may be done by determining the ratio
of evaporation (in terms of cubic centimeters per day) to soil moisture
(in percentage of dry weight), using the average daily evaporation of
the arid fore-summer, and the average soil moisture of the arid fore-
summer at 15 cm. These ratios are exhibited in table 20. The approx-
imate ranges of the ratios are 1 to 25 for the Forest, 20 to 35 for the
Encinal, and 35 to 50 for the Desert. If evaporation data had been
secured at 9,000 feet the value of the ratio for that elevation would
have been less than unity.

The values of the ratio of evaporation to soil moisture afford a
concise expression of the major conditions which affect the water
relations of plants, and they demonstrate the wide divergence of these
conditions in the desert valleys and on the forested mountain summits
during the arid fore-summer. The average daily evaporation rate
has been shown (fig. 12) to fall during the humid mid-summer to half
the amount during the arid fore-summer. The soil moisture is like-

CORRELATION OF VEGETATION AND CLIMATE.

93

wise increased in the former season to an amount that would greatly
reduce the values of the ratio if determined for the humid mid-summer.
The ratio of evaporation to soil moisture is not in itself a full index
of the comparative aridity of Desert, Encinal, and Forest, for the con-
ditions expressed by the ratio are of much longer duration at 3,000
feet than at 8,000 feet. The shortening of the arid fore-summer from
16 weeks at 3,000 feet to 7 weeks at 8,000 feet (see fig. 2) signifies that
the most severe drought conditions of the year are more than twice
as prolonged at the lowest elevation as compared with the uppermost.
It is necessary here to bear in mind that the effects of drought on plants
are cumulative, and that, for example, a period with a given set of
conditions of increasing aridity which endures for 16 weeks may be
twice as fatal or deleterious as a period that lasts for 14 weeks. For
purposes of general climatic description, however, the values of the
ratio of evaporation to soil moisture multiplied by the duration of the
arid fore-summer may be taken as an index of the aridity of the several
elevations (see table 20).

TABLE 20. — Average daily evaporation (E) and the moisture of the soil (SM), together with

/ E1 \

the ratio of evaporation to soil moisture ( ~ -, I for north and south exposures at six eleva-
tions in the Santa Catalina Mountains for the arid fore-summer of 1911.

Expo-

V

0 Tlf

E

Duration of

Elevation.

sure.

Vegetation.

S M

fore-summer.

3 000 feet .

s

Desert

101.1

2.0

50.5

16

4 000 feet

s

Desert

80.4

2.0

40.2

4 000 feet

N

Encinal . ...

82.7

2.5

33.1

14

5 000 feet .

s

Encinal

61.7

3.1

19.9

5 000 feet .

N

Encinal

74.4

3.5

21.3

13

6 000 feet .

s

Encinal

59.4

1.8

33.0

6 000 feet

N

Forest

56.1

3.5

16.0

11

7 000 feet

s

Forest

62.8

2.6

24.1

7 000 feet

N

Forest

49.9

5.5

9.1

9

8 000 feet

s

Forest

29.3

7.4

3.9

8 000 feet .

N

Forest

29.4

11.3

2.6

7

The ratio of evaporation to soil moisture comprises a measurement
of all the external factors which affect the water relations of plants,
except the influence of radiant energy on transpiration and the possible
effects of soil temperature on this function. It is accordingly unneces-
sary to give further consideration to rainfall, which is not in itself a
factor for vegetation, at least in such a region as Arizona. If any
differences existed between the seasonal distribution of rainfall at
different elevations in the Santa Catalinas the fact would be of great
importance to the vegetation, but only in the effect it would have on
the annual march of the soil moisture conditions. The evidences of
observation and instrumentation have shown that the major drought

94 VEGETATION OF A DESERT MOUNTAIN RANGE.

periods of the Desert are rarely broken on the mountain by rainfall
of significant amount.

The records of rainfall for 8 years show that the summer rain at
8,000 feet may be from 1.9 to 3.5 times as great as that at 3,000 feet
(see table 4). The average of the 8 years shows the summer rain at
8,000 feet to be about 2.4 times that on the Desert. The average of
a longer series of years will probably approximate this amount and
the securing of the winter precipitation would probably make little
difference in the proportion.

The very conditions of low evaporation which favor the water rela-
tions of the plants in the Forest region are also favorable to the preser-
vation of the moisture of the soil. The effect of the winter rains upon
the soil moisture of the Forest is accordingly carried forward many
weeks (see table 7 for soil moisture at 9,000 feet after 6 weeks without
rain) . The slow melting of snow on north slopes still further prolongs
the effect of winter precipitation. These causes underlie the vernal
activity of herbaceous plants in the Forest region (see p. 29) and the
growth by trees of the Forest region during the arid fore-summer.

TEMPERATURE FACTORS.

The role played by temperature in differentiating the vegetation of
the various altitudes of the Santa Catalinas is by no means so simple
a matter as that played by moisture conditions, and is far from being
capable of expression in a concise mathematical form. It has already
been shown that the frostless season decreases from a length of 40
weeks on the Desert to a length of 19 weeks in the Forest at 7,600 feet
(see tables 10 and 11, and fig. 2), and that the temperature has an
average apartness of 26° between Desert and Forest (see table 13).
No instrumentation has been carried on which would establish the
quantitative nature of the difference between other phases of the tem-
perature conditions. The shortening of the growing season with in-
crease of altitude, and the concomitant lowering of the temperatures
of this season, are factors of great moment to the vegetation, but no
work has been done to establish the precise temperature and tempera-
ture-duration requirements of any species of plants. The shortness of
the growing season in the Forest and the coldness of the nights of
mid-summer (40° to 50°, see table 14) are both hostile to growth
activity and may well be limiting factors in the upward distribution
of many species of the Encinal.

The temperature conditions of winter are equally important with
those of summer in underlying the limitation of species and vegetations,
and their role may be played independently from that of the summer
temperature conditions, or the two may play conjointly upon the
same species at the same altitude. Among the various phases of winter
temperature conditions are the length of the period subject to frost,

CORRELATION OF VEGETATION AND CLIMATE. 95

the number of days with freezing temperature, the number of consecu-
tive days or hours of freezing, and the absolute minimum reached.
Any two or more of these phases may operate conjointly to influence
a plant, and the temperature preceding a particular constellation of
conditions may enhance the harmful effects of those conditions. In
fact the general weather conditions accompanying or following a given
phase of temperature may determine the full effect of the temperature.

In those cases in which plants are killed by the action of low tem-
peratures the most important factor to be considered is the actual
duration of the period during which the plant is subjected to tempera-
tures below 32°. Secondary to this are the considerations of the amount
of precooling received by the plant, the actual minimum temperature
to which it was taken, the condition of the soil and the atmosphere
during the freezing, and the nature of the weather subsequent to it.
In a previous paper* the writer has called attention to the manner in
which the most critical phase of low temperature conditions increases
in severity with increase of altitude. Even on the coldest winter days
the temperature on the Desert never fails to rise above 32° during the
mid-day. The lowest temperatures of winter are invariably accom-
panied by a clear sky, and the days preceding and following very cold
nights are clear. A cloudy or rainy period is always accompanied by
more moderate temperatures, as is shown by the rarity of snow on the
Desert itself. The longest duration of a shade temperature below
freezing, in the ten-year records of the Desert Laboratory, is 19 hours.
It frequently happens that a duration of 6 hours is the greatest for
an entire winter. On ascending the mountains the length of the most
prolonged period of freezing becomes greater until an altitude is
reached at which there are occasional winter days when the air
temperature does not rise above freezing. At this altitude there is
a sudden increase in the maximum number of hours of frost from
22 or 23 hours to a length of 40 to 45 hours. Such a sudden in-
tensification in the duration of a critical climatic condition causes
this condition to operate more sharply in the limitation of plant
distribution than is the case with conditions that exhibit the usual
form of slowly graduated change.

No winter thermograph records have been secured in the Santa
Catalinas, and it is therefore impossible to state the exact altitude at
which this sudden intensification of the frost factor becomes manifest.
It is probable that it lies at about 4,500 feet. The exposure of plants
to insolation may often save them from the effects of an air tempera-
ture of less than 32°, and in the case of succulents the temperature to
which their tissue is raised during the insolation of the preceding day
will shorten the period of freezing for them. These subsidiary matters

* Shreve, Forrest. The Influence of Low Temperatures on the Distribution of the Giant
Cactus. The Plant World, 14: 136-146, 1911.

96 VEGETATION OF A DESERT MOUNTAIN RANGE.

affect the exact altitude at which the maximum number of freezing
hours may operate in the limitation of a given species; and the exact
topographic character of the location of an individual plant may also
affect the operation of this factor.

The experimental work of the writer has shown that a duration of
more than 18 hours of freezing temperature is fatal to Carnegiea and
that Opuntia versicolor and Echinocereus polyacanthos are capable of
withstanding durations of 66 hours. The limitation of Carnegiea is
apparently due to the operation of this factor. Its occurrence becomes
confined to south slopes at 4,000 feet and it becomes less and less
abundant from that elevation up to 4,500 feet. One of the highest
individuals at the latter elevation is protected by a rock on its north
side, above the summit of which the cactus now projects for 8 inches.
This projecting top was badly frosted on its north side in the severe
winter of 1912-13, while the north side of the plant below the summit
of the rock was uninjured. A small Carnegiea (18 inches high) has
been discovered in Soldier Canon at 5,100 feet. It grows on the south
side of a low rock, and its location is on the steep south slope which
terminates a long ridge between two main branches of the canon. The
plant is here well protected from the cold-air flow of the canon and is
subjected to the full insolation of the short winter days. It showed
some slight effects from the exceptionally cold winter just referred to,
but succeeded in recovering from them. In the early arid fore-summer
of 1911 the writer transplanted a young Carnegiea 3 inches high from
the base of the mountain to the vicinity of the 6,000-foot station on
Manzanita Ridge. The cactus was placed on the southwest side of a
rock, with a large plant of Arctostaphylos northeast of it, and occupied
a location near the summit of the ridge. The plant was watered
several times in order to help it to become established, but was not
assisted after the commencement of the summer rains. It success-
fully passed the winter of 1911-12; it made gains in turgidity in the
summer of 1912, but no measurable growth; in the spring of 1913,
after the winter in which the minimum temperature at that locality
was —6°, the plant was found to be dead. Although the rainfall
at Manzanita Ridge in the summer of 1912 was 8.68 inches as com-
pared with 5.61 inches at the location from which the cactus was
taken (near the 3,000-foot station), it was not able to seize the
advantage. This fact itself involves the factor of summer temperature,
which doubtless determines the rate of growth of the roots and their
power for the intake of water.

The evidence which shows Carnegiea to be limited in its upward
distribution by the greatest number of freezing hours is probably
applicable to a large number of desert plants, non-succulent as well
as succulent forms, which find their limitation at about the same
elevation.

CORRELATION OF VEGETATION AND CLIMATE. 97

THE ROLE OF TOPOGRAPHIC FEATURES IN DETERMINING DEPARTURES
FROM THE NORMAL ALTITUDINAL GRADIENT OF VEGETATION.

The normal or ideal gradient of vegetation is disturbed by three
sets of topographic influences: (a) that of slope exposure, (6) that of
the surface flow or underflow of streams and arroyos and the high
soil moisture of flood-plains, and (c) that of location with respect to
ridges, slopes, or valley bottoms, which may be designated briefly as
the influence of topographic relief. These three sets of topographic
features do not bring into operation any factors, nor any intensities
of the common factors, which are not involved in the normal vertical
gradients of physical factors, although in some cases they bring about
new combinations of factors not exactly duplicated at other elevations
under the conditions of the hypothetical normal gradient. In the
description of the vegetation there have been frequent allusions to
these three sets of departures from the ideal gradient of vegetation.
Instrumentation has also been described which throws light upon the
operation of slope exposure and of topographic relief. The influence
of streams has been very obvious in its nature and has not been investi-
gated instrumentally.

THE ROLE OF SLOPE EXPOSURE.

The importance of slope exposure in determining the vertical limits
of species, and in thereby determining the vertical range of types of
vegetation, has been a matter of observation and comment among
almost all writers on the vegetation of the western United States.
Although the phenomenon is of universal occurrence throughout the
extra-tropical portions of the globe it is rendered particularly striking
in regions where there are transitions from desert or grassland into
forested country. In any region like the Santa Catalina Mountains,
with their steep climatic gradient and varied topography, the opera-
tion of the factors involved in slope exposure is such as to present an
alternation of vegetistic regions, causing constant departures of the
vegetation from the theoretical norm to the norm of higher or lower
portions of the mountain.

Slope exposure is a " factor" in differentiating the vegetation of
opposed slopes at all elevations. Even at altitudes between 2,000 and
3,000 feet among the volcanic hills of the Tucson region, there are
conspicuous differences between the south slopes, with their heavy
stands of Carnegiea, Encelia farinosa, and Opuntia bigelovii, and the
north slopes with their abundant individuals of Parkinsonia microphylla
and Lippia wrightii and their heavier growth of perennial grasses.*
The difference between northern and southern exposures is most con-
spicuous between 4,000 and 5,000 feet, where the former have orchard-

* See Spalding, V. M. Distribution and Movements of Desert Plants. Carnegie Inst., Wash.,
Pub. 113, 1909.

98 VEGETATION OF A DESERT MOUNTAIN RANGE.

like stands of evergreen oaks and the latter are treeless (see plate 9s).
Almost equally striking, however, is the contrast between the open
pine forests on south slopes at 9,000 feet and the heavy, deep-shaded
stands of fir on north slopes at the same elevation (compare plate
29A and plate 35).

Although the influences of slope exposure are operative at all eleva-
tions they acquire added power with increase of altitude.* In the
Desert region and the Lower Encinal the uppermost limits of species
on north slopes and on south slopes are from 600 to 1,000 feet apart
(Carnegiea, Echinocactus, Quercus emoryi), while in the Forest region
the upper limits on opposed slopes differ by 1,000 to 2,000 feet (Quercus
hypoleuca, Juniperus pachyphlcea, Arbutus arizonica). Another test
of the same fact may be had by comparing a north slope at 3,000 feet
with a south slope at 6,000, and by then carrying the comparison up
3,000 feet. Between the north slope at 3,000 feet and the south slope
at 6,000 feet is the strong contrast of Desert and closed Encinal, with
only a few xerophilous ferns and one small cactus in common. Between
the north slope at 6,000 and the south slope at 9,000 feet is the very
close resemblance of two stands of Pine Forest, in one of which are still
to be seen a few Encinal forms that have disappeared from the other and
higher one.

The increased influence of slope exposure at higher elevations is not
to be attributed to the fact that the species of the Upper Encinal and
Forest range through greater elevations than do the species of the
Desert and the Lower Encinal. The number of feet through which
a species ranges on south slopes or on north slopes has no necessary
connection with the difference between its upper limits on north and
on south slopes. The ability of a large number of plants to range
through a greater vertical distance in the Upper Encinal and Forest
than it is possible for the plants of the lower vegetations to do may be
owing to the ability of the plants of the upper portion of the mountain
to withstand a greater gamut of conditions than the plants of the basal
vegetations can. It would, in any event, not be due to the existence
of more gradual gradients of climatic change at the higher elevations,
since in every case of the measurement of these gradients they have
been shown to grow steeper between 6,000 and 9,000 feet than below
6,000 feet. The increase in the effects of slope exposure with increase
of altitude can only be ascribed to an increasing differentiation of the
climatic conditions between north and south slopes at higher elevations.
An examination of the curves of evaporation and of soil moisture (figs.
10 and 14) will show that the readings for the highest stations exhibit
the greatest apartness, at least with respect to the intensities involved.

* Merriam has illustrated this fact in a diagrammatic profile of San Francisco Peak, but has
not mentioned it in the text of his paper. See Merriam, C. Hart. Biological Survey of the San
Francisco Mountain Region, Arizona. U. S. Dept. Agric., North Amer., Fauna No. 3, 1890, pi. 1.

CORRELATION OF VEGETATION AND CLIMATE. 99

It is obvious that the importance of slope exposure lies in the topo-
graphic control of the physical factors which form the environment of
the plants concerned. It is possible to know, on purely a priori grounds,
that two slopes of the same inclination, which lie in opposed positions
so that one faces north and the other south, will present to plants two
environments differing in almost every essential physical feature. The
temperature of the air on two such slopes might be identical as deter-
mined by the thermometers of a carefully established meteorological
station, but they are distinctly different as they affect the vegetation,
for the plants not only receive the direct rays of the sun but receive
very different amounts of heat through diurnal terrestrial radiation.
This circumstance is of small importance to full-grown trees and large
plants, but is of great importance to young plants and seedlings. The
soil temperatures of opposed slopes are also widely unlike, even in the
presence of the undisturbed cover of natural vegetation. The two
opposed slopes would in all likelihood receive the same rainfall, al-
though this is not necessarily the case. An equal amount of rain might
effect an equal elevation of the soil moisture on the two slopes, and to
the same depth, but the soil evaporation of the south slope would
greatly exceed that of the north slope, and a lower moisture would soon
prevail in the soil of the former. Greater or less differences may thus
be shown to obtain between the opposed slopes with respect to the
most vital features of plant environment. Any attempt to explain the
importance of slope exposure in determining plant distribution is there-
fore incomplete unless it takes into account every possible environ-
mental difference between the slopes. Some of these differences are
undoubtedly of far greater importance than others, but the question
of their relative importance is always one that must be asked with
respect to a particular species of plant. To make a thoroughgoing
answer as to the importance of slope exposure for a single species is in
itself a very great undertaking.

The universal occurrence of a large number of species of plants in
the vegetation of the Santa Catalinas, their commonness within their
ranges, and the consistency of their distribution with respect to slope
exposure, all indicate that there has been ample time in the history of
the mountain for all of these species to attain as wide a distribution as
it is possible for them to have under existing climatic conditions. It
is difficult to conceive of any upward or downward movement being
possible for any of the common species of plants, inasmuch as thousands
of years have already given an opportunity for such extensions of
range. In view of the steep climatic gradient of the mountain it is
easy to believe that all of the common species have reached upper and
lower limits beyond which their survival is prevented by definite fea-
tures of the physical environment. The present vertical limit of a
species, whether upper or lower, must be looked upon as the average

100 VEGETATION OF A DESERT MOUNTAIN RANGE.

point at which some particular feature of its physiological activities
is met by some particular environmental condition that is preventive
or unduly inhibitory to it. The minor fluctuations of climate, which
have their minimal and maximal values within periods that are as
brief as the normal life of a perennial plant, are registered in the infre-
quency of every species as it approaches its distributional limit and
in the scattered individuals which lie farthest out from the main area
of occurrence. The secular changes of climate which have their maxi-
mal and minimal points many centuries apart are registered in slight
movements of the limits of species, the marginal region of scattered
occurrence being, of course, the first affected by such movements.

The writer has seen no evidence indicating that competition between
plants is at any place in the Santa Catalinas responsible for the limi-
tation of any species. There is, of course, competition such as that
between seedling pines in heavy stands of 10 to 40 years in age, and
such competition as occurs between individuals of the same or different
species of herbaceous plants in small areas of moist flood-plain. While
competition may thus determine the surviving individuals of a stand
of young trees, or may determine the composition of a small community
of ephemeral or root perennial plants, it is not responsible for the find-
ing of a plant in one habitat rather than in another, and is not respon-
sible for the exclusion of a species from an area in which it might find
favorable conditions.

It is only consistent with our knowledge of the diversified physical
requirements of plants that there should be such great diversity in
the location of the belts of altitude occupied by different species, and
it accords with our knowledge of the distribution of plants in general
that these belts should be wider in some cases than in others. It is
possible, however, to pick out groups of plants the limits of which
correspond roughly with the limits of the Desert, Encinal, and Forest
types of vegetation respectively. Even the plants of equatorial regions,
in which there is a notable constancy of climate, both daily and annual,
are able to endure small ranges of climate, or occasionally to endure
changes in individual factors which are many times as great as the
normal fluctuations of their native climates. In addition to the
fluctuations of climate from month to month or from year to year which
must be endured by any plant, there are often even greater differences
which must be simultaneously endured by the most remotely separated
individuals of the same specific stock. For example, Asclepias tuberosa
is found from Maine to Minnesota in the north and from Florida to
Texas in the south, and thence sporadically in the mountains westward
to Arizona. In view of the prodigious range of this somewhat poly-
morphous plant it is surprising not to find it reaching a greater elevation
than 6,500 feet in the Santa Catalinas. With regard to possible differ-
ences in the physiological behavior of the most remotely separated
individuals of a plant stock of such wide range, we know little.

CORRELATION OF VEGETATION AND CLIMATE. 101

To find a plant growing only on a north slope at 5,000 feet, only on
a south slope at 7,000 feet, and on both at 6,000 feet, as is the case with
Pinus cembroides, for example, means that there is much in common
between the physical conditions on the north slope at 5,000 feet and
the south one at 7,000. If such reversals of habitat in relation to slope
were rare it would only be warrantable to state that there were some
physical features in common between the opposed slopes, but, as
already stated, there are only two common plants (aside from those
of the stream ways) regarding which a similar statement could not be
made. It is obvious, therefore, that if we compare separately the alti-
tudinal gradients of climatic change for the south slopes and for the
north slopes of these mountains, the two gradients will be similar in
character and will be closely related. Their relationship will consist
in the fact that a given intensity or value on one of the gradients will
be found on the other at a lower or higher elevation, unless barred by
the base or summit of the mountain. In the curve showing the alti-
tudinal gradients of evaporation on north and on south slopes (fig. 14)
it will be seen that the rate on the south slope at 6,000 feet is exactly
the same as the rate on the north slope at 5,000 feet. The rate on the
south slope at 8,000 feet, however, is far less than the rate on the north
slope at 7,000 feet. The latter rate is found on south slopes at about
7,300 feet, according to the evidence of the curve. In spite of differ-
ences in the pitch of the climatic gradients at different elevations, it
is always possible to find a slope which exhibits the same intensity of
a given factor as that which has already been found on an opposed
slope, but it is necessary to go up or down the mountain from 500 to
1,500 feet to do so. It might be possible to find two spots on opposed
slopes in which there was very nearly the same complex of all environ-
mental conditions, although the finding of two slopes with identical
ones would be rendered almost impossible by the necessity of seeking
these spots at different altitudes.

Even if a series of considerable differences were found between the
north slope at 5,000 feet and the south slope at 7,000 feet, on both of
which Pinus cembroides is growing, nevertheless such differences would
be little greater than those which are met by this species as it grows
on both north and south slopes at 6,000 feet or those that exist between
the north slopes at 5,000 and 6,000, or the south slopes at 6,000 and
7,000 feet.

The physical factors which underlie the influence of slope exposure
are simply a special case, for the most part, of the same factors which
cause the altitudinal differentiation of the vegetation of the entire
mountain. The only instrumentation carried out with a view to secur-
ing a measure of the influence of slope exposure is comprised in the data
on soil moisture and on evaporation for north and south exposures at 7
and 5 elevations respectively (see tables 7 and 9 and figs. 10, 13, and 14).

102

VEGETATION OF A DESERT MOUNTAIN RANGE.

c.c.

100

- -5.000 N

— 6.000 N andS

7.000 5.

Neither the data for soil moisture nor those for evaporation show
the exact alternation exhibited by the vegetation itself, by virtue of
which a given north slope is similar in vegetation to a south slope
about 1,000 feet above it, and a given south slope is similar to a north
slope about 1,000 feet below it (with the exception of the highest
altitudes). The conditions of evaporation found through the range
of Pinus cembroides, which has been used as an example of the effects
of slope exposure, are indicated in figure 18, where curves are given
showing the seasonal march of evaporation on a north slope at 5,000
feet, the average of the evaporation on north and south slopes at 6,000
feet, and the amounts on the south slope at 7,000 feet. These curves
follow a course which is parallel and indicate evaporation conditions
which are remarkably
similar for the lower, cen-
tral, and upper individ-
uals of this pine, except
for the higher rate at
5,000 feet during the arid
fore-summer.

The ratios of soil mois-
ture to evaporation at
different altitudes have
been worked out sepa-
rately for the north and
south exposures at the 6
stations (see table 20).
Since these ratios are an
expression of the conditions of the arid fore-summer they must be
taken as elucidating only those phases of slope exposure which are
themselves due to the climate of that season. Any subsidiary in-
fluence of temperature in affecting the slope exposure phenomena of
vegetation in the Santa Catalinas still awaits a full investigation.

The comparative conditions of the lower, central, and upper habitats
of Pinus cembroides may be again investigated in the light of the ratios,
which are as follows: North slope at 5,000 feet 21.3, average of north
and south slopes at 6,000 feet 24.5, south slope at 7,000 feet 24.1. These
figures indicate a still more remarkable similarity of the water con-
ditions in the three habitats than the evaporation figures do. To
compare a plant of lower range we may take A gave schottii, which
encounters conditions expressed by the following ratios : north slope at
4,000 feet 33.1, average of north and south slopes at 5,000 feet 20.6,
south slope at 6,000 feet 33.0. These figures fail to show as close
agreement, but indicate a close similarity of the water conditions at
the lowest and uppermost habitats, and a more favorable set of con-
ditions in the central habitat. To make a similar comparison for a

FIG. 18. — Graphs showing seasonal march of rate of evap-
oration for a north slope at 5,000 feet (dotted line), the
average for a north and a south slope at 6,000 feet (solid
line), and for a south slope at 7,000 feet (broken line).

CORRELATION OF VEGETATION AND CLIMATE. 103

plant of higher range than Pinus cembroides we may take Quercus
hypoleuca, which ranges through 3,000 feet, with the usual alternation
at the top and bottom of its range. The ratios for its habitats are as
follows: North slope at 6,000 feet 16.0, average of north and south
slopes at 7,000 feet 16.6, south slope at 8,000 feet 3.9. Here is close
agreement of the ratios for the lower and central portions of the range,
with a much lower value for the top, indicating that in spite of the
ability of Quercus hypoleuca to withstand the conditions expressed by
the ratio of 16, it is likewise capable of withstanding the more favorable
water conditions indicated by the ratio of 3.9. Here, in other words,
is a typical Encinal plant, accompanied throughout its range by many
others, which is able to extend up to an elevation at which the water
conditions are much more favorable than they are in the lower part
of its range. This is a thing which the Desert plants do not do, and
the reason is undoubtedly that the plants of the Desert encounter
unfavorable temperature conditions at the same elevations at which
they begin to encounter more favorable water conditions, while such
a plant as Quercus hypoleuca is capable of withstanding the rigorous
temperatures of 8,000 feet and is thereby enabled to range upward
into a region of more favorable water conditions.

Allusion has been made to the more pronounced character of the
effects of slope exposure at higher elevations. It is of interest in that
connection to contrast the ratios of evaporation to soil moisture for
similarly located pairs of habitats at low and at high altitudes. For
example, the north slope at 4,000 feet has a ratio of 33.1, the south
slope at 6,000 has a value of 33.0. To carry the comparison up 2,000
feet: the north slope at 6,000 feet has a ratio of 16.0, the south slope
at 8,000 feet has one of 3.9. The greater similarity of the radios for
the two lower habitats is in accord with the evidences from the vege-
tation (see p. 98), which indicate an altitudinal increase in the potency
of slope exposure in the determining of the vegetation.

The fundamental causes differentiating the conditions on opposed
slopes are only partly comprised in the evaporation — soil-moisture ratios.
The differences of evaporation rate on north and south slopes are
largely due to the dry, warm winds which ascend the mountain during
the day and partly to the differences of air temperature. The humidity
of the air shows only slight differences on opposed slopes. The soil
moisture on north slopes is higher than on south ones because of the
more direct insolation on south slopes, and because of the higher soil
temperature and increased soil evaporation which are due to this. In
addition to the differentiating features which are expressed in the
ratios, we have the differences of soil temperature, due to the direction
of slope, and the differences of air temperature, which are only partially
registered in their effect upon the evaporation rate. The increased
insolation on slopes as compared with level ground has been worked

104 VEGETATION OF A DESERT MOUNTAIN RANGE.

out by Hall,* who shows that the amount of radiant energy reaching
a south slope of 45°, with the sun 45° above the horizon, is 1.4 times
as great as that reaching a level piece of ground through an aperture
of the same size. It is through this difference, which is still greater
between south and north slopes, that the soil is given a higher tempera-
ture, that the air is warmed to a higher degree through radiation, and
the soil dried more rapidly on all south-facing exposures.

THE ROLE OF STREAMS AND FLOOD-PLAINS.

In the description of the vegetation of the Santa Catalinas constant
allusion has been made to the distinctive plant communities of springs,
streams, flood-plains, and arroyos. The contrast between the vege-
tation of these moist or relatively moist situations and that of the
mountain slopes is very striking at the mouths of the larger canons,
and throughout the Desert and Encinal regions. At the higher alti-
tudes, and particularly in the Fir Forest, the moist habitats are not
only less striking to the casual observer, but their vegetation actually
comprises a great many species which are frequently found away from
proximity to streams.

The influence of streams and flood-plains consists, in brief, in bring-
ing components of the upland vegetation of each altitude down along
the streamways of the altitudes just below. In this manner the Encinal
is traversed by bands of Forest, and the Desert slopes are traversed
by bands of Encinal. Furthermore, the streams and springs of the
mountain afford the sole habitats for a number of species of aquatic
and palustrine plants which do not appear on the upland at any
elevation.

The mechanical agencies of gravity, sheet-floods, and stream flow
are all capable of aiding in the downward dissemination of the seeds
of all mountain plants and these mechanical agencies should assure
the occurrence of all mountain plants in all situations at lower altitudes
in which they are capable of survival. The number of seeds which
are borne down by streams is, of course, enormous, and the number of
resulting germinations is probably very large. The number of sur-
vivals, however, is controlled by the physical conditions of the new
low-altitude habitat, and in a manner to be further considered.

In the discussion of slope exposure no account has been taken of the
occurrence of plants along streamways at elevations below their lowest
upland occurrence, since the individuals along the arroyos and streams
are subjected to a very different set of environic controls from those
that determine the location of the upland individuals. In the earlier
discussion of the vertical limits of species the streamway occurrences
were taken into account.

* Hall, H. M. A Botanical Survey of San Jacinto Mountain. Univ. Cal. Pubn. Bot., vol. i,
pp. 1-140, 1902.

CORRELATION OF VEGETATION AND CLIMATE. 105

Among the palustrine plants which occur along streams at 7,000 to
8,000 feet are two species of Juncus and two of Carex which also occur
in Sabino Canon, under the most favorable conditions of moisture
supply, at 3,000 to 3,200 feet elevation. The perennial composite
Tagetes lemmoni grows along the drier arroyos of the Pine Forest down
to 6,000 feet, and is found in lower Sabino Canon growing along the
margin of the stream at 3,200 feet. Other palustrine plants of the
Forest region are found from time to time at low elevations along the
largest streams, but no others have been observed to become thoroughly
established there.

The well-known cosmopolitanism of many aquatic plants would
cause us to expect such behavior as is exhibited by Juncus, Carex, and
Tagetes in the Santa Catalinas. There are several species of Scirpus
and Eryngium, and at least one woody plant (Cephalanthus occidentalis)
which range from the Gulf of Mexico across the southwestern boundary
of the United States to California. The individuals of these species
are subjected to a wide diversity of atmospheric humidities, but are
all found under conditions of closely equivalent high soil moisture.

A greater interest attaches, in the present connection, to the cases
of low streamside occurrence of plants which grow typically in upland
situations. Mention has already been made of the trees of Quercus
arizonica and Quercus oblongifolia which grow along the Sabino Creek
at 2,800 feet, about 1,200 feet below their lowest occurrence on north
slopes. Small plants of Quercus hypoleuca have been found growing in
deep shade in the bed of Sabino Canon at 3,200 feet, which is 2,700
feet below the lowest north slope occurrence of this tree. The first-
named oaks have descended no further than many other upland plants
have done, but the last-named oak shows the most pronounced de-
pression of range that has been detected.

At the mouth of Soldier Canon, at 3,000 feet, the writer has found
one or two individuals each of Dasylirion wheeleri, Mimosa biundfera,
Erythrina coralloides, and Asclepias linifolia. At an elevation of 4,500
feet Dasylirion and Asclepias have begun to appear on slopes of south
exposure, and at 5,000 feet Erythrina and Mimosa have also left the
arroyos.

At 4,900 feet Ceanothus fendleri is found in the shade of oaks on the
flood-plain of Soldier Canon. It occurs also at 5,300 feet in similar
situations at the head of Soldier Canon, and becomes frequent in the
Upper Encinal at 6,000 feet. Similarly Quercus submollis occurs near
the constant water at Horse Camp, in Bear Canon, at 6,100 feet and
is of increasing frequence along streams up to 7,200 feet. At that
elevation and up to the uppermost limit of Pine Forest it is common
on slopes as well as near streams. Robinia neomexicana is found in
the flood-plain of Soldier Canon, near a spring, at 5,300 feet, and first
becomes a frequent upland shrub of the Pine Forest at about 7,500 feet.

106 VEGETATION OF A DESERT MOUNTAIN RANGE.

It is possible to say, in brief, that the conditions presented by stream-
sides and flood-plains are such as to depress the ranges of very many
plants by as much as 1,000 feet, and of a few plants by amounts as
great as 2,000 feet. A depression of as much as 2,700 feet, found in
the case of Quercus hypoleuca, does not represent the lowest occurrence
of established plants, but rather a chance survival at an elevation in
which it would doubtless be impossible for the tree to reach maturity.
It can at least be said that throughout the entire length of Sabino
Canon, from the mouth to the Basin, there are no known occurrences
of full-grown trees or even shrubs of Quercus hypoleuca.

The extent to which the types of vegetation are depressed in their
ranges by the influence of streams and flood-plains is about the same
as the average depression of the individual species, that is to say about
1,000 feet. In the case of the occurrence of a closed Encinal in the
Basin of Sabino Canon there has been a depression of 1,500 feet in
the limit of this type of vegetation — from 5,500 to 4,000 feet.

Some evidence has already been given leading to the view that the
lower limits of all Encinal and Forest plants are determined by those
features of the environment which in turn determine the water relation
of plants. The facts of the depression of vertical ranges by streams
form an additional evidence of this view. So far as concerns atmos-
pheric water-demand the plants growing beside streams are subjected
to the same conditions as plants of the nearby upland, but the conditions

of water supply are infinitely better for them. In other words, in the

v
ratio of evaporation to soil moisture, SM, the numerator is the same

for stream-side and upland plants and the denominator is greatly
increased for the latter, thereby lowering the values for the ratio. In
the cases alluded to in which the lowest individuals of a species not
only grow in a flood-plain but in the shade of larger vegetation, the
plants are under ameliorated conditions with respect to the numerator
as well as the denominator in the ratio.

A number of mountain plants are able to survive when taken down
to the Desert provided they are placed under conditions in which one
or both of the sets of conditions indicated by the above-mentioned
ratio are ameliorated. Parthenocissus from 6,000 feet survives with
irrigation and partial shade; Echinocereus polyacanthos, from 5,000 to
7,000 feet, survives with occasional irrigations during the arid fore-
summer; Zauschneria calif ornica, Aquilegia chrysantha, and Sedum slelli-
forme, all ranging from 5,500 to 7,500 feet, are capable of survival at
Tucson from year to year when grown in complete shade with frequent
irrigation during the arid fore-summer. These facts point to the ability
of such plants to withstand at least the shade temperatures of the
Desert, provided the moisture supply of the soil and the moisture
requirement of the air are made more nearly like those conditions in
the mountain habitats of the plants. Many introduced plants have

CORRELATION OF VEGETATION AND CLIMATE. 107

shown themselves incapable of withstanding the atmospheric aridity
at Tucson even when grown under the most liberal irrigation. The
inability of a plant to pass water on to its transpiring tissues as rapidly
as it is withdrawn by a desert atmosphere is undoubtedly a feature
common to very many mesophilous plants, and it is apparently the
cause which prevents a greater number of palustrine mountain plants
from descending the large streamways to the Desert, and it doubtless
prevents a lower descent upon the part of many Forest species which
reach the flood-plains of the Lower Encinal.

It might be argued that the low occurrence of Forest along the
streams of the Encinal and the descent of the Encinal into the Desert
Slopes are due to the influence of cold-air drainage rather than to the
effects of soil moisture, or that cold-air drainage is at least an important
contributory factor. It is difficult to believe that low temperatures,
especially those of the winter months, should be a favoring factor for
plants which are subjected during the day to just as high temperatures as
are the plants of the upland. During the summer months the low noctur-
nal temperatures might be of some slight importance, but such importance
would reside solely in aiding the plant to recover from the excessive
transpiration of the preceding day and to build up a reserve of water
against the transpiration of the following day, as has been shown by
Edith B. Shreve to occur in Parkinsonia microphylla.* The facts that
it is the highest diurnal temperatures that are apt to be deleterious
to low-ranging mountain plants and that their effect can be only
indirectly and slightly offset by the lowest nocturnal temperatures
make it appear that cold-air drainage has at least a very minor role
in this connection as compared with the moisture conditions.

THE ROLE OF TOPOGRAPHIC RELIEF.

Each of the leading types of vegetation in the Santa Catalinas
reaches the uppermost limit of its occurrence on ridges and high south-
facing slopes. This carries the Desert upward into the Encinal and
carries the Encinal up into the Forest in such a manner that there is
an interdigitation of the vegetistic regions throughout the portions of
the mountain in which the topography is mature enough for it to be
manifest. This appearance of interdigitation is partly brought about by
the influence of streams (which has just been discussed) and is some-
times merged with the influence of slope exposure. These facts do not
in the least obscure the high range of each type of vegetation on the
narrow ridges which point due south or north and are therefore free
from the influence of slope exposure.

On the ridges which lie between the tributaries of Soldier Canon
have been found the highest individuals of all of the characteristic

* Shreve, Edith B. The Daily March of Transpiration in a Desert Perennial. Carnegie Inst.
Wash., Pub. 194, 1914.

108 VEGETATION OF A DESERT MOUNTAIN RANGE.

species of the Desert (for elevations see page 37). On a high ridge
tributary to Bear Canon have been found the highest individuals of
Opuntia sp., the highest species of that genus on the mountain, and
Mamillaria grahami, the highest-ranging plant of the Desert. The
individuals which most nearly approach these highest stations for
Opuntia and Mamillaria have been found on south exposures about
600 feet lower, in the Bear Canon drainage.

On an exposed ridge, with a considerable inclination to the south,
at 7,800 feet are found the highest individuals of Pinus cembroides,
Juniperus pachyphloea (with one known exception), Yucca schottii,
Echinocereus polyacanthus, and Arctostaphylos pungens. In this station
the influences of slope exposure and of topographic relief are combined,
thereby bringing about the pronounced conditions that are expressed
in the highest occurrence of 5 species of the Upper Encinal. On the
ridges above Marshall Gulch are found the highest occurrences of
Quercus hypoleuca and Quercus reticulata, both of which forms extend
further down the south faces of these east-and-west ridges than they
do down the north faces.

When Desert plants are found on the ridges of the Encinal region
they fail to appear on the south-facing slopes just below these ridges.
When the plants of the Encinal are found at their highest locations
on ridges of the Forest Region they are also absent on the south-facing
slopes just below the ridges. This does not appear to be the case with
respect to the highest occurrences of plants which are believed to have
their true climatic limit just below the summit of Mount Lemmon,
such as Quercus hypoleuca and Quercus reticulata.

The extent by which the highest individuals on ridges exceed the
highest individuals on south slopes is never more than 500 to 600 feet,
except in the case of Pinus cembroides, in which it is about 700 feet.
Opuntia sp. and Mamillaria grahami, which have their upper limit in
the vicinity of 7,000 feet, agree in this respect with Opuntia versicolor,
Echinocactus wislizeni, and Fouquieria splendens, which have their
limit in the vicinity of 5,500 feet.

Perhaps the most common explanation of the highest occurrence of
species on ridges is that the soil is driest in such situations and therefore
offers to plants from lower elevations a habitat more like that in which
they are abundant. The principal objection to such an explanation
is the unquestionable fact that a somewhat more moist soil is not
inimical to the plants of the Desert nor to the plants of the Encinal.
Neither is there a sufficient difference between the soil moisture at
the bottom of a slope and on the ridge at the top of the slope, in the
arid seasons, to cause a differentiation of the vegetation.

The explanation of the phenomenon may be sought partly in the
existence of cold-air drainage, which is at least responsible for the
absence of Desert and Encinal plants from the bottoms of canons at

GENERAL CONCLUSIONS. 109

the highest elevations to which they attain. The streams of cold air
are not more than 75 to 100 feet deep, however, and can not, therefore
be functional in preventing the occurrence of plants on the middle
and upper slopes of canons. An apparently valid explanation of the
high occurrences on ridges is in accordance with the theory already
mentioned, that the upper limits of the Desert species, and possibly
of the Encinal species also, are set by winter temperature conditions.
The ridges are obviously the localities which receive the fullest and
longest insolation on the short winter days with low sun. This cir-
cumstance would not only warm the plants themselves but would
warm the soil and rocks in a manner such as to lessen the severity of
the coldest nights. With the pronounced low temperatures in the
canons, due to cold-air drainage, and with the favorable conditions
of the ridges for a pre-warming of both plant and habitat, it may be
expected that there will be great differences between the vertical
limits of species in canon bottoms and on ridges.

GENERAL CONCLUSIONS.

The desert mountain ranges of the southwestern United States stand
in the midst of a region which presents severe conditions for plants.
The relative richness of the vegetation in this region is due chiefly to
the occurrence of two yearly seasons of rainfall. The entire annual
vegetational behavior is related primarily to the moisture seasons and
much less pronouncedly to the thermal seasons. The perennial plants
lead an existence which permits of rapid growth during the warm
humid season, together with an extremely low ebb of activity during
the arid seasons, and with the possible loss through drought-death of
much of the growth that has just taken place.

The severe conditions of the desert environment cause the vegetation
to exhibit a high degree of sensitiveness to slight topographic and
edaphic differences. Wherever the character of the soil or the topo-
graphic location is such as to present a degree of soil moisture slightly
above that of the general surroundings, or as to maintain it for a longer
time in the periods of extreme aridity; or in whatever locations plants
are protected from the most extreme conditions of transpiration — in
such places are to be found heavier stands of vegetation or else particular
species of plants.

The higher mountains of the desert region exhibit strong gradients
of change in climate and in vegetation. Both of these gradients are
much more pronounced than those of mountains of equal elevation
in more humid regions. They lead from arid to humid, or at least
semi-humid, conditions of moisture, and from sub-tropical to tem-
perate conditions of temperature; from low, open microphyllous and
succulent desert, through a sclerophyllous semi-forest to heavy conif-
erous forest.

110 VEGETATION OF A DESERT MOUNTAIN RANGE.

The sensitiveness which desert vegetation exhibits to slight environ-
mental differences is even more pronounced with respect to the climatic
gradients of the mountains. Throughout a vertical range of 6,000
feet there is not only a very striking gradient of vegetation, but a
very nice adjustment of vegetation to the physical conditions. In the
Desert and Encinal regions, and to a great extent in the Forest as well,
this is chiefly an adjustment of plant to environment and scarcely at
all an adjustment of plant to plant. Every juvenile individual in the
open Desert and Encinal regions is a pioneer, and on reaching maturity
this individual is part of an ultimate stable community.

The principal features of altitudinal climatic change are : the short-
ening of the frostless season, the lowering of the daily curve of tempera-
ture throughout the frostless season, the increasing of the intensity
and duration of all critical phases of low temperature during the frost
season, the shortening of the arid fore-summer (the critical season of
aridity), the increasing of precipitation and therefore of soil moisture,
and the decreasing of evaporation.

On a mountain having the form of a smooth cone it would be possible
to observe the ideal manner in which these climatic gradients would
collectively control the vertical distribution of the vegetation. The
occurrence in nature of irregularities of relief is responsible, however,
for local departures from the ideal vertical gradients of climate and
also from the ideal altitudinal distribution of vegetation which would
be anticipated on a geometrically constructed mountain. It is possible,
nevertheless, to correlate the climatic and vegetational gradients in
spite of the local irregularities of each of them, and in fact the study
of these departures from the ideal has aided in the interpretation of
the correlations.

The vertical distribution of vegetation on the Santa Catalina Moun-
tains has been found to be due to the interaction of two sets of controls
wilich are nearly distinct. One of these controls has its seat in the
moisture conditions, the other in the temperature conditions. The
temperature control has been studied experimentally only with respect
to three species of plants, but it is believed on this evidence (as well
as the evidence of the departures from the normal gradient of vegeta-
tion, correlated with instrumentation) to be the control which limits
the upward distribution of the Desert species and perhaps of some
species of the Encinal. The moisture control has not been studied
experimentally in connection with the present investigation, but its
operation is well known, and the instrumental study of soil moisture
and evaporation at successive altitudes, with due attention to the
departures from the normal gradient of vegetation, has indicated that
the ratio of the latter factor to the former affords a concise expression
of the control which limits the downward distribution of Forest and
Encinal plants.

GENERAL CONCLUSIONS. Ill

The principal departures of the vegetation from the ideal gradient
that would be found on a geometrical cone are expressed in the irregu-
larity of the upper or lower limits of vegetations or of individual species
as observed in different habitats. The chief departure is that due to
slope exposure, by virtue of which the vegetation of north-facing and
south-facing slopes at the same elevation shows striking differences.
A second departure is that due to the influence of streams and the
high moisture content of the soil of arroyos and flood-plains, by reason
of which the plants of all altitudes are carried below their normal
lowest occurrences on slopes. Another departure is due to the influence
of ridges, on which the plants of all elevations (and particularly those
of the Desert) find their highest occurrences. These departures seldom
result in the occurrence of distinctive plant communities, but are
operative rather in the carrying of the usual and widespread communi-
ties into elevations at which they are exceptional. The effect of slope
exposure is to carry the normal vegetation of a given elevation both
up and down the mountain, so that its lowest occurrences are on north
slopes and its highest on south slopes. The effect of stream ways is
to carry either the normal or the streamside vegetation down the moun-
tain, so that the extreme lowest occurrences of almost all Encinal and
Forest plants may be sought along the stream ways. The effect of
ridges is to carry the vegetation (or more particularly individual
species and small groups of species) up the mountain, so that all highest
occurrences of Desert and Encinal species are to be found on narrow
ridges — the highest occurrences of Forest plants are not reached on
the Santa Catalina Mountains, and they are controlled by a very
dissimilar group of factors.

It is impossible to study the distribution of vegetation in a region
where pronounced differences may be found within short distances
without being impressed with the independence which each species
exhibits in its allocation. Plants which are associated on the Lower
Desert Slopes, for example, range to very different maximum altitudes,
and plants which are associated in the Upper Encinai are found to be
in part at the upper edges of their ranges, in part at the lower edges,
and also in part rather closely restricted to that region. It is nowhere
possible to pick out a group of plants which may be thought of as
associates without being able to find other localities in which the asso-
ciation has been dissolved. Certain plants may be thought of as having
closely identical physical requirements because of their associated
occurrence in the same spot. Nevertheless, the fact that the vertical
ranges and habitat characteristics of these species will reveal more
or less pronounced differences goes to show that each of them has
survived in a particular section of the climatic gradient. It is true in
the Santa Catalina Mountains, as it is true in all other places, that the
associated members of a plant community are not able to follow each

112 VEGETATION OF A DESERT MOUNTAIN RANGE.

other to a common geographical and habital limit. The physical
requirements of plants are so varied and so elastic that the composition
of a series of communities occupying similar habitats in widely sepa-
rated places shows the constant overlapping of the ranges of individual
species which is due to the physiological inequivalence of these species.

It is particularly true of the plant communities of arid and semi-
arid regions that the most closely associated individuals are not alike
in their life requirements, and this is true to a less pronounced extent
in all plant communities. The members of the many diverse biological
types, or growth forms, which are found together in Desert and Encinal
find their soil water at different levels, procure it at different seasons,
and lose it through dissimilar foliar organs, at the same time that they
react differently to the same temperature conditions. In brief, these
associated plants are not living in the same climate but are living in
different sections of the same climate, the demarcation of these sections
being either temporal or spatial.

The use of the physical characteristics of the habitat as a criterion
in the definition of a plant community does something to give a greater
rigidity and a wider applicability to the definition. On the other hand
it confuses cause with effect and makes it impossible to investigate
the relation of physical conditions to a community defined in that
manner without reopening the whole question as to the nature and
identity of the community. There is much strong logic to support
the view that all necessary definitions and classifications of vegetation
should be made on the basis of the vegetation alone. When units of
vegetation are thus defined they lend themselves to the further study
of their life requirements, and it is such study — applied to individual
species as well as to vegetation — that affords the most promising and
important field for ecological activity.

The distribution of vegetation in the Santa Catalina Mountains is
strongly controlled by a steep climatic gradient; the vegetation itself
is diversified in its display of growth forms; and the secular changes
of vegetation due to physiographic phenomena, and to the reaction
of the plant upon its habitat, are in almost complete abeyance. These
circumstances have made it possible to give a delineation of the vege-
tation upon purely vegetational characteristics, without regard to the
secular changes which are taking place in very restricted areas, and
with particular emphasis upon the individualism of behavior among the
characteristic species. The same circumstances have also made it
possible to lay side by side the facts respecting the vegetational gradient
and those respecting the climatic gradient in such manner as to reveal
the correlations between the two and to indicate some of the physical
controls which operate in the limitation of the activities and of the
ranges of species and of vegetations.

SHREVE

Plate 1

SHREVE

Plate 2

A. South face of Santa Catalina Mountains viewed 7 miles from their base. Mount Lemmon is on right center.
In foreground is bajada vegetation of Covillea tridentata, Opuntia spinosior, and Isocoma hartwegi.

*f - rtJ*5B»5*'fc* •" •'rjRr" ••^••,-m^'^f *vpv.:^v

B. Extreme southwestern ridge of Santa Catalinas viewed from the north. In foreground is the bed of the Canada
del Oro, with individuals of Hymenoclea monogyra and a marginal fringe of Prosopis velutina and Chil-
opsis linearis.

SHREVE

Plate 3

A. Typical Low Bajada, with pure stand of Covillea tridentata and summer carpet of Bouteloua aristidoides.

B. Looking southwest from Upper Bajada near mouth of Soldier Canon. In the arroyo in foreground are
Carnegiea gigantea, Parkinsonia microphylla, Acacia greggii, and Fouquicria yplendens. Distant
hills are relict toes of ancient bajadas.

SHREVE

Plate 4

SHREVE

Plate 5

A. Desert Slopes near mouth of Pirn a Canon. In foreground, from the left are Parkinsonia microphylla,
Simmondsia calif arnica, Carnegiea gigantea, Opuntia tourney i, and Lycium berlandieri.

B. Rocky Desert Slopes in same vicinity as A. Against the sky are FOI/I/U/I ri/i splendens, Prosopis
velutina, and Momitsia pallid a; in foreground Spfiirralcea pedata, Lippia wn-jhtii, and Opuntia
engdmanni.

SHREVE

Plate 6

'•*•-• • '•*
' *

'

.. . y*

SHREVE

Plate 7

SHREVE

Plate 8

SHREVE

Plate 9

A. Looking northeast along Lowest Slopes of the Lower Encinal at 4,300 feet. At left Arctostaphylos pungens, at
right Quercus oblonrji folia, below it Dasylirion wheeleri, to left of the latter Nolina microcarpa.

B. Looking southeast in Soldier Canon. On right the treeless slopes of the Upper Desert, on left open Encinal

of Quercus oblong i folia and Quercus arizonica.

SHREVE

Plate 10

A. Flood-plain of Soldier Canon at 4,200 feet. The predominant plant is Baccharis sarothroides.

••

.
.

B. Lower Slopes and Flood-plain of Soldier Canon at 4,900 feet. The predominant trees are Quercus emoryi
and Juniperus pachyphloea, the shrubs Arctostaphylos punyens and Garrya wrightii.

SHREVE

Plate 11

A. Encinal in Soldier Canon at o.UOO feet, with Quercus emoryi, Juniperus pachyphliea, Garrya wrijhtii,
Yucca macrocarpa, and Nolina microcarpa. At lower left is Bouteloua rothrockii.

B. Quercus oblongifolia in Flood-plain of Soldier Canon. At left Quercus' emoryi, at right Nolina microcarpa.

SHREVE

Plate 12

SHREVE

Plate 13

A. Heavy Carpet of Summer Herbaceous Vegetation on a Flood-plain at 5T000 feet. Solidago sparsiflora
var. subcinerea, Monarda pectinata, Crotalaria lupnlina, and Sphceralcea sp.

B. Agave schottii in the Lower Encinal. Extensive areas between 4,200 and 5,400 feet are

covered by this plant.

SHREVE

Plate

A. Gymnopteris hispida occupying a ledge of rock in the Lower Encinal at 5,200 feet.

B. Mats of Selaginella sp. among rocks at 5,300 feet.

SHREVE

Plate 15

SHREVE

Plate 16

c3
O

O
02

o

IN

<: o
*^ ~*^

$0

"S o

Ss
o

o

"r:
°o

a

O

SHREVE

Plate 17

X

S

3
O

;*

ao

O
-w

o>
u

c
W

•o
o

SHREVE

Plate 18

SHREVE

Plate 19

-
c
O

ro
o

c
-^

o

o

CS

a

s-.

e

o

CO

in
0)

SHREVE

Plate 20

atTfctoL y. • -.-«? .-•,..-•• -ifx,

Snfflwcifc, '- • ~^-* *•* ';'. ••"$'••,. vi

*V*|yt?.'V., : . 'rl^-jt •„ r :'i . j. t •-
_; 7**y?f. . , , y^.i. £#.| /V \*<.J V -

• ^' '

A. Flood-plain in Bear Canon at (J.OUU feet, with Pinus arizonica, Populus jamesii, an open carpet of
summer-active herbaceous perennials, and Sporobolus confuse*.

B. Stream way in Bear Canon at 6,000 feet, with Pinus arizonica, Juglans rupestris, and Vitis arizonica.

SHREVE

Plate 21

Open stand of Pinny arimnita, I hum chihuahuana, and Juniperus pachyphlaea near floor of Bear Canon at 0,100
feet. In background are rocky slopes of north wall of Bear Canon.

SHREVE

Plate 22

SHREVE

Plate 23

SHREVE

Plate 24

SHREVE

Plate 25

T«* : ••-.;

A. Looking toward south face of Mount Lemmon from crest of Marshall Gulch, near site of S, 000-foot
climatological station. Pinus arizonica and scrub of Quercus reticulata.

B. Looking southwest into Marshall Gulch. The open area in the forest is a thicket of Populus tremuloides.

SHREVE

Plate 26

SHREVS

Plate 27

A. Typical heavy stand of Pinus arizonica on a bench in Marshall Gulch at 7,SOO feet.

B. Looking east along south-facing slope of Marshall Gulch at 7,700 feet. The bunch-grass is

Muhlenbergia virescens.

SHREVE

Plate 28

SHREVE

Plate 29

A. Open Forest on Steep South Slopes of Main Ridge at 8,500 feet. The shrubs are Quercus reticulata.

B. Stream and Narrow Flood-plain in Marshall Gulch near Montane Garden. Alnus acuminata, Acer

brachypterum, and Abies concolor.

SHREVE

Plate 30

SHREVE

Plate 31

SHREVE

Plate 32

A. Looking northwest from main ridge toward Samaniego Ridge.

B. Looking east along main ridge of Santa Catalans from a point east of Mount Lemmon at about S, <>(><) feet.

SHREVE

Plate 33

A. Open Forest on summit of Mount Lemmon, at 9,000 feet, with good reproduction of Pinus arizonica and

a close stand of Dugaldia hoopcsii.

B. Mature thicket of Populus tremuloides on summit of Mount Lemmon at 9,100 feet. In foreground are
Pteris aquilina var. pubescens and flowering plants of Frasera speciosa.

SHREVE

Plate 34

SHREVE

Plate 35

An alluvial flat in Fir Forest on north slopes of Mount Lemmon at 8,600 feet. Pinus arizonica, Abies concolor,

and Populus tremuloides.

SHREVE

Plate 36

A. Santa Catalinas viewed from north, showing grassy plains at elevation of 4,200 feet in the vicinity of

Oracle. At right Prosopis velutina, at left Yucca alata.

B. At north base of Santa Catalinas looking toward San Pedro River, at 4,500 feet. At right Quercus emoryi

and Nolina microcarpa, at left Yucca alata.

i i II !l""J|"iNH III I/HI l/ll