Yale Peruvian expeditions, 1911-1915.
Publications, v. 50.

YALE
UNIVERSITY LIBRARY

Agriculture and Native Vegeta
tion in Peru

OK X

Reprinted from the Jojjrnau of i>he. Washington
Academy of Sgi-ences, Vol: VI, No. 10,
Ktiy 19,: 1916 .'; '
•¦:" j

Reprinted from the Journal of the Washington Academy of Sciences
Vol. VI, No. 10, May 19, 1916
BOTANY. — Agriculture and native vegetation in Peru. 0. F.
Cook, Bureau of Plant Industry.
Alternation of forests with open grass lands or sparse desert
vegetation is one of the most striking of the biological phenomena
of tropical countries. Since the time of Humboldt many travelers
in tropical America have sought to explain the presence or absence
of the different types of vegetation by reference to differences
of geological formations, altitudes, prevailing winds, or other
natural features. More recent observations in Central America
have led to the opinion that the chief factors governing the
distribution of the forest vegetation are the agricultural occupa
tion of the land and the continued action of fire on lands aban
doned from cultivation.
There are reasons for believing that most of the forests of Central
America do not represent original or virgin growth, but different
stages of reforestation. Likewise most of the open grass lands
and deserts appear to be consequences of the native system
of farming — to be interpreted as artificial conditions rather than
as natural features. The climatic, geologic, or topographic
factors, though not without influence in determining the rate
of reforestration, seem in general to have very little importance
in comparison with human activities and exposure to fire. The
complete reforestation of fireswept grass lands is a long and
gradual process, but the successive stages can be recognized
by taking account of the habits of the different kinds of trees.1
Opportunities of studying the relations of agriculture to forest
vegetation under a different combination of natural conditions
have been afforded during four months (April to July, 1915)
spent in southern Peru and Bolivia as a member of the Expedi
tion conducted by Professor Hiram Bingham, under the aus
pices of Yale University and the National Geographic Society,
with the cooperation of the United States Department of Agri
culture. Most of the time was spent in the region traversed
by the Urubamba River and its tributaries, from the Pass of
1 Cook, O. F. Vegetation Affected by Agriculture in Central America. Bull.
145, Bureau of Plant Industry, U. S. Dept. of Agriculture. 1909.

cook: vegetation in southern pertj 285
La Raya, at an elevation of 14,000 feet, down to Santa Ana, at
an elevation of 3000 feet, including a visit to the Panticalla
Pass and the Lucumayo Valley. The region includes Cuzco,
Pisac, OUantaytambo, and Machu Picchu, the chief centers of
the Inca and pre-Inca or Megalithic civilizations, and is of great
agricultural and ethnological interest as the original home or
place of domestication of numerous species of cultivated plants.
In this part of Peru, as in Central America, it apears that
the present distribution of the principal types of vegetation is
not a natural effect of altitudes, climates, or soils, but an artificial
result of an intensive agricultural occupation of the land, extend
ing through a long period of time. If we wish to think of an
original condition, a biological background, so to speak, of the
primitive agricultural civilization that occupied this region,
we must imagine a country well covered with forests. The
destruction of forests appears to have been carried much further
than in Central America, in many localities to the complete
extermination of all forms of arboreal vegetation. The chief
considerations that seem to support these conclusions are stated
in the following paragraphs.
biological conditions favorable to forest growth
Though many districts are now entirely treeless and true
forests are found in only a few localities, there appear to be no
natural conditions that are definitely unfavorable to arboreal
vegetation. Light, heat, and moisture are sufficient to sup
port the growth of trees and there is ample fertility of soil, at
the higher elevations as well as in the lower and more tropical
valleys. In other words, there- seems to be no climatic or biologi
cal factor to preclude the growth of trees on any part of the
land surface except the bare rocks and snow fields at the summits
of the high Cordilleras.
From the positions of the moraine deposits and the lack of
soil accumulations above them it may be inferred that the
glaciers have receded in comparatively recent times, perhaps
following the destruction of the forests. Some, of the moraines

|r O «#*|p

¦2S,t*sr

SO

Fig. 1. Native agriculture in a branch of the Ushcopata Valley above Sicu-
ani, Peru, at an altitude of nearly 13,000 feet, with fields of potatoes, barley,
and broad beans, and remnants of the native forest flora — the white trees qui-
shuar, the others mostly quenuar and capldi.

cook: vegetation in southern peru 287
are as low as 9000 feet, with the present glaciers ending from
2000 to 4000 feet above. Under the Peruvian conditions it does
not seem unreasonable to believe that the removal of a forest
covering might tend to bring about a recession of the glaciers.
Greater exposure of the rocky slopes would bring increased heat
and dryness of atmosphere. Less snow would fall and the ac
cumulations on the high summits would be exposed to longer
periods of melting under direct sunlight.
As trees are often found above the moraines, there is no reason
to doubt that the ancient forest covering extended up to the
glaciers, as forests are known to do in other glacial regions.
Several isolated tracts of forests have been found by Professor
Bingham at very high elevations, even up to 15,000 feet. These
high-altitude forests are of interest as affording the most definite
demonstration of the fact that tree growth is not limited by
elevation alone. ANDINE FOREST FLORA
The possibility of a forest covering for all of the inhabited
areas of this region is shown not only by the fact that trees
grow when planted, but also by the presence of an indigenous
forest flora whose different components are well adapted to the
various natural conditions afforded by different exposures and
elevations, up to the line of glaciers.
Two of the high-altitude trees, quenuar (Polylepis) and quisuar
(Buddleia), have been noted frequently by travelers because
they are often planted in villages or allowed to grow among the
fields. (See fig. 1.) Other members of the Andine forest flora
of southern Peru are lambran (Alnus), chachacoma (Escallonia),
unca (Eugenia), lengli (Hesperomeles), quisca (Berberis), mulli
(Schinus), chicjlluromay (Vallea), and numerous other trees,
including various arboreal Compositae. Several of the genera
are represented by two or more species. Some of these, such
as the species of Escallonia, are reported by botanical writers
only as shrubs, but under favorable conditions they attain true
arboreal proportions, especially at altitudes of 10,000 to 12,000
feet.

288

COOK: VEGETATION IN SOUTHERN PERU

REFORESTATION OF TERRACED VALLEYS
Though no original or virgin forests are now known to exist
in this part of Peru on any lands that could be cultivated, re
forestation with native trees is in progress in many places,
notably in the valley aboVe OUantaytambo, and in the next
valley to the west, leading up to the Panticalla Pass. In both
valleys the growth of the native trees has progressed so far that

Fig. 2. Valley above OUantaytambo, Peru, at an altitude of about 10,000
feet, showing terraced slopes partly overgrown with forests of native trees,
including quishuar, quenuar, lambran, unca, and lengli.
genuine forest conditions have been restored, in the OUantay
tambo Valley covering several hundred acres, in the Panticalla
Valley thousands of acres. Reforestation is demonstrated by
the fact that the trees stand on ancient agricultural terraces
supported by skilfully constructed stone walls. (See fig. 2.)
The survival of native trees in these valleys may be ascribed
to the presence of lateral ravines too deep, narrow, and precipitous
for cultivation. Such places may well have escaped the other-

cook: vegetation in southern peru 289
wise complete clearing of the land for agricultural purposes. '
Hardy types of woody vegetation, growing in the ravines and
on the rocky slopes above, might survive even long periods of
agricultural occupation of the terraced lands below. The neigh
boring areas might be seeded easily from the old trees. Such a
ravine, leading up from the other side of the spur shown in the
figure, is now heavily wdbded and contains many well matured
trees of the same species that now cover the terraces, although
very few of the trees appear to be. very old or to have reached
the stage of natural decay. Some of the terraces in the Panti
calla Valley are covered with much older trees than any found
in the valley above OUantaytambo.
HABITS OF SURVIVING NATIVE TREES
Although the genera of native trees, as mentioned above,
belong to as many different families, there is a general similarity
in habits of growth, in that all of them sprout readily from the
stumps and endure repeated cutting. Seed is produced in a
few years, if the sprouts are allowed to grow. The limitation
of the present forests to such trees may be taken to indicate that
persistent vitality was necessary to pass through the periods
when these valleys were occupied by large agricultural popu
lations, as shown by the extensive terracing of the slopes. A
period of complete denudation of a valley would mean the
extermination of all kinds of trees that were unable to sprout
from the stumps, but trees like Escallonia, Eugenia, and Schinus
might survive centuries of pollarding. The last is familiar as
the ''pepper-tree," now grown by thousands in southern Cali
fornia for shade and ornamental purposes. Other members of
the Andjne tree flora are even more attractive in appearance
and promising for introduction into the United States.
DENUDATION OF UNCULTIVATED LANDS
The former presence of large agricultural populations accounts\l
not only for the clearing of all the lands that could be cultivated, h
but also for the denudation of lands that were not capable of I ¦

290 cook: vegetation in southern peru
cultivation. The growth of each native community means
that supplies of fire-wood have to be sought farther and farther
away. A large Indian town is usually surrounded by a broad
belt of denuded lands, no forest being allowed to remain within
two or three leagues. Judging the past by the present, a period
of denudation of all the neighboring slopes must have followed
the building of the extensive systems' of terraces in the valleys
about OUantaytambo, Torontoy, and Machu Picchu. The
country around these centers must have reached the same
treeless state as the districts that now have large agricultural
populations, such as the Vilcanota Valley and the slopes around
Lake Titicaca. reforestation prevented by fire
In many localities cultivation is confined to the bottoms of
the valleys or to the lower slopes, while the higher slopes have only
a sparse covering of grass or low bushes. This gives the im
pression that the interior of the country is naturally treeless,
like the desert regions along the coast. But the coast deserts
are explained by the rainless climate, whereas in the interior the
rainfall is sufficient to support forest growth.
The former cultivation of many of the higher slopes is indi
cated by the ridges and terraces that still remain. These show
in turn the previous existence of forests, since forests must have
preceded cultivation in order to accumulate soil and make it
possible to clear the land by the primitive method of burning.
This method is ineffective on grass lands, which have to be re
forested before they can be re-occupied by a primitive agricul
tural people. At the higher altitudes grassy slopes are culti
vated by spading, but this method is used only where turf is
formed. When treeless slopes are seen in tropical valleys meeting the
tropical forest vegetation, it is plain that some active enemy
of forest growth must be at hand, and this is fire. The fires that
are set to clear land for cultivation commonly escape and over
run the slopes above. As the grass-covered slopes are used
only for grazing, no effort is made to protect them from fire.

cook: vegetation in southern peru 291
forests in inaccessible places
Trees are often found growing under very unfavorable natural
conditions, in places that are too steep, rocky, or isolated to
be cleared for cultivation or for providing fuel. In the lower
Urubamba Valley it was observed that the driest and rockiest
hillsides in the vicinity of Santa Ana are covered with forests
of huillca (Piptadenia) and other tropical trees, while the smooth
er and more fertile lands on either side have no trees, but a
heavier growth of grass.
Tree seedlings often appear in grass lands, but are killed
when fires sweep over them. Hence, the forests are confined
to the rocky slopes as long as the adjacent grass lands are visited
by fire. Grazing reduces the danger of fire, and this assists
in reforestation; but the forests themselves may burn after
sufficiently long periods of drought. In the lower Urubamba
Valley, at altitudes of from 4000 to 8000 feet, the forests have
been burned on many slopes altogether too steep for cultivation.
This not only kills the trees but often has the effect of loosening
the soil and rocks, causing destructive landslides.
PAUCITY OF THE HUMUS FAUNA
Another indication of the more complete denudation of this
region in former times is the paucity of the humus fauna, com
prising the insects, millipeds, centipeds, and other small ani
mals that live normally in the upper layers of the soil. These
creatures become very abundant under conditions that afford
permanent moisture in the soil, but are killed when the land is
burned over or parched by severe drought. In southern Peru
the humus-inhabiting animals are everywhere extremely scarce,
and often lacking altogether. The number of species is very
small, as well as the number of individuals. Of millipeds only
three orders are represented, Merocheta, Anocheta, and Diplo-
cheta; in many localities only Merocheta, and most of these of
Antarctic types rather than tropical. The three orders of very
primitive arthropods, Symphyla, Rhabdura, and Dicellura, are
present, but were nowhere found in abundance, even in places

292 cook: vegetation in southern peru
where reforestation has advanced so far that ample deposits of
humus have accumulated. Humus-inhabiting insects of other
orders, including the Thysanura and Collembola, are also few.
ABSENCE OF PALMS IN TROPICAL FORESTS
The flora of the valley between San Miguel at 6000 feet and
Santa Ana at 3000 feet is distinctly tropical, and large areas are
forested, but not with original or virgin growth. That reforesta
tion is still far from complete is shown by the general scarcity
and often complete absence of palms. Instead of a normal
palm flora, no locality in the Urubamba Valley was found to
have more than two species, a large Geonoma and a small
Chamaedorea. This deficiency seems the more significant because the natural
conditions are extremely favorable and the palm flora of the
adjacent regions of South America is one of the richest in the
world. The original palm flora of this district can not be esti
mated at less than a dozen species, and may have included two
or three times that number. But denudation would involve
a complete extermination of the palms, and these plants are
very slow to return, even after forest conditions have been re
established. RECONSTRUCTION OF THE FLORA AND FAUNA
In view of these indications of prolonged interference with
the original conditions of plant and animal life it does not seem
reasonable to ascribe the present distribution of the native
vegetation entirely to differences of soil, temperature, altitude,
or rainfall. An ecological account of the Peruvian flora that
ignores the factor of human activity, as in the treatise by Weber-
bauer, does not convince. To present an adequate conception
of the native flora and its relations with the external conditions
would require a process of reconstruction, a careful collecting
and piecing together of the parts of the flora and fauna that
are left. Then there would need to be a careful comparison
with the floras and faunas of neighboring regions that were not
occupied and denuded by the ancient civilizations, if such re
gions can be found. This would give a better appreciation of

COOK: VEGETATION IN SOUTHERN PERU 293
the extent to which the presence of man has modified the original
conditions of the environment along the eastern slopes of the
Andes. AGRICULTURE AT HIGH ALTITUDES
Corresponding with the wide range of altitude there is great
diversity in the forms and habits of the wild vegetation and in
the agricultural arts of the native inhabitants. In the lower
valleys where corn is the principal crop the method of clearing
the land and the relation of agriculture to the native vegetation
are much as in Central America.
The high plateaus of, Peru, where the native agricultural •
population is now chiefly centered, are unlike any part of Central
America, the nearest approach being found in the tablelands
of Guatemala. In Central America cultivation is hardly carried
above 8000 feet; whereas in Peru potatoes and other Andine
crops are commonly grown at 13,000 feet and in some places at
14,000 feet. Moreover, it is in these elevated districts that the
native system of agriculture attained its highest development
and was least disturbed by the Spanish conquest.
CONCLUSIONS
The native agriculture of southern Peru is self-limiting.
Cultivation may be maintained for longer periods on the table
lands and higher slopes, but when the soil is once exhausted and
removed by erosion there is less prospect of renewal through
reforestation than at the lower elevations. Although under
the high altitude conditions the accumulation of soil goes on
to a certain extent in open grass lands, without the aid of forests,
such gains evidently do not make good the losses incidental to
cultivation. Large areas of the higher slopes that appear to
have been cultivated intensively in former times are now com
pletely sterile and abandoned.
Thus Peru may be said to afford even more striking evidence
than Central America of the fact that the primitive agricultural
civilizations were not permanent, but of limited duration.
Eventually the soil became unsuited to cultivation by the native
methods.

, I 08300 3006 E