University of Virginia Library

IX. ALLUVIATION AND AGRICULTURE[1]

That portion of Chaco Canyon in which we are primarily interested
is a 15-mile-long section walled by massive beds of Cliff House
sandstone. Shales lie above and below the Cliff House, and the
disintegration of sandstone and shale has filled the canyon with clay
and sand to an estimated depth of 100 feet or more. Of this
valley fill approximately one-third lies exposed in the banks of the
present arroyo—a flood-carved gully that had its beginnings in the
middle 19th century (Bryan, 1954, p. 15). This exposed third provides
an index to past geological conditions, but nowhere is there
evidence, at any time in the past, of a living stream in Chaco Canyon.
The entire valley fill, 100 feet of it, has been piled up, inch by inch,
by wind and rainwash.

In consequence of his investigations Bryan concluded that the
present arroyo is part of a fourth cycle of alternating erosion and
sedimentation. We are not especially concerned here with the origin
of the canyon itself nor with the processes by which it was initially
filled with alluvium, then eroded again, and subsequently refilled.
But the arroyo identified with the third period of erosion may not be
ignored since it provides a very plausible reason for the abandonment
of Pueblo Bonito and other communal dwellings along its
course.

The inhabitants of these multiple-storied settlements were farmers.
They cultivated fields of corn, beans, and pumpkins wherever soils
were suitable and moisture adequate. Agriculture was their livelihood.
The builders of the great Chaco Canyon pueblos were floodwater
farmers because they had no choice. Lacking a permanent
stream for irrigation purposes they necessarily depended upon the
runoff following midsummer showers.

But summertime showers in Chaco Canyon are neither reliable nor
predictable. They may advance promisingly to within half a mile,

as we witnessed repeatedly, and then turn abruptly aside and leave
the valley floor dust dry. Although the local climate has long bordered
on the arid, we have ample evidence that precipitation was
once more plentiful. Perennial grasses and greasewood formerly
flourished here; pines and cedars grew in side canyons and on the
mesas above. Cottonwoods and willows once traced a green ribbon
down midvalley, and some few were still there as late as 1920, the
year of the Society's Chaco Canyon reconnaissance. Vegetation
multiplies when the water table is high and a ground cover both
hinders runoff and quickens the deposition of silt and sand. Navaho
elders recall a time, or say they do, when water could be had anywhere
in the valley with a little digging.

Surface vegetation, however, has not always been lush enough to
hold Chaco floodwaters in check. There were times of drought or near
drought when shrubs and grasses withered and the runoff ran faster.
Below-average rainfall is recorded repeatedly in the annual growth
rings of timbers salvaged from Pueblo Bonito. Douglass (1935,
p. 49) points to a period of subnormal rains extending from A.D.
1005 to 1036 or even later. The 10-year drought between 1090 and
1101 was a severe one and may have had a more devastating effect
upon local plant life than is possible now to measure. Conceivably
it could have initiated the 12- to 15-foot-deep "buried channel" that
so intrigued Bryan—the channel that destroyed an annually increasing
amount of Bonitian farmland and, in so doing, probably hastened
migration of the Bonitians.

DROUGHT AND THE BURIED CHANNEL

Bryan first chanced upon this buried watercourse near the southeast
corner of Pueblo del Arroyo without knowing at the time that a
nearby exposure had previously been discovered by W. H. Jackson
(1878, p. 444). At this site, the section 4 of his map, Bryan (1954,
p. 54) described the old channel as 13 feet 5 inches deep and its
gravelly bottom 15 feet below the existing surface. The difference
in the two figures illustrates the fact that, after the 13-foot gully
had been completely refilled, an additional 19 inches of sand and
silt were spread upon the refill. Elsewhere the overburden is thicker
and the gully, deeper. The bottom of it lies 18 feet below the surface
in Test Pit No. 3, on the plain fronting Pueblo Bonito (pl. 7, upper).

A short distance upstream from section 4 and on the opposite
bank, Jackson (ibid., p. 443) in 1877 observed a small ruin whose
foundations were "5 or 6 feet below the general level of the valley."

Fig.24 -The "Far East Trench" (upper); the "Middle Trench"; and
the "East Mound Trench" from the post-Bonito channel to the
foundations of Room 171.

Plate 78

Left: To increase
its height,
stones were piled
loosely upon the
east retaining wall,
East Refuse
Mound.

(Photograph by
O. C. Havens,
1924.)

Right: A makeshift
wall enclosed
the East Refuse
Mound on the
south.

(Photograph by
Neil M. Judd,
1922.)

Plate 79

Left: A partly
razed second-typemasonry
structure
was buried under
village waste, East
Refuse Mound.
North retaining
wall at workman's
elbow.

Right: Four feet
beneath the south
foundation of
Room 171, our extended
East Mound
exploratory trench
bared a broad open
hearth.

(Photographs by
O. C. Havens,
1925.)

Plate 80

Until January 22, 1941, the Braced-up Cliff towered above the
fourth-story wall of Room 14b.

(Photograph by Neil M. Judd, 1926.)

A fourth-type wall between Rooms 55 and 57 was supported upon 2 timbers
their ends inserted into the stonework of Old Bonitian Room 58.

(Photograph by Neil M. Judd, 1922.)

Plate 81

Upper: A supposed multiple shrine, built upon blown sand 3 feet above the east foundation of
Room 176.

(Photograph by Neil M. Judd, 1923.)

Lower: Beams seated in pecked holes roofed this small talus room, west end of the
Braced-up Cliff.

(Photograph by Neil M. Judd, 1927.)

This little P. III house has since been destroyed but beneath it, as
seen in 1920 and later, layered alluvium extended to right and left,
the patient deposits of floodwaters flowing over a grassy plain
(Bryan, 1954, pl. 6, upper). In Chaco Canyon where the gradient
has always been a modest one, summertime floods formerly advanced
slowly and dropped their silty burden on the way.

Bryan traced his buried channel up and down canyon a distance
of approximately 5 miles and recorded his observations at 23 numbered
sites. While so doing he repeatedly found proof of human
occupancy coincident with deposition of the main valley fill. He collected
Pueblo I potsherds as much as 20 feet below the surface, while
those attributable to P. III, the period of Pueblo Bonito and its
contemporaries, rarely occurred below 4 feet. Of known P. I pithouses
in the canyon, two stood with roof levels at a depth of 6 feet
or more (Judd, 1924, p. 403; Roberts, 1929, p. 71; Bryan, 1954,
p. 32).

The fact that water-borne sediments were sometimes laid down
so gently that even charcoal in ancient hearths was not appreciably
disturbed evidences both a low gradient and a long-continuing grasscovered
surface. We did no testing south of the present arroyo, but
on the north side we met floodwater silt layers repeatedly. Although
not always deposited as evenly as those underlying the little P. III
ruin near Bryan's section 4, they are undeniable.

When exploring the Northeast Foundation Complex (fig. 11),
we exposed 22 inches of excellent Late Bonitian masonry at Station 1
upon a sturdy foundation 4 feet 7 inches high (pl. 48, right). At
6 inches deeper, or 7 feet from the present sandy surface, is a pavementlike
sheet of floodwater silt, smooth as a kitchen floor. And
wherever we dug deep enough throughout that whole foundation complex
we came upon the same or a similar silt layer.

Close under "Hillside Ruin" are five quadrangular fireplaces
(pl. 47, lower). A test pit 4 feet 10 inches deep between the second
and third revealed another adobe pavement, floor-smooth but covered
by debris of reconstruction and wind-blown sand. Presumably
that pavement is a continuation of the one we had previously
encountered outside Room 184—the pavement that prompted our
5-foot-wide trench to the lower terrace below the Braced-up Cliff
(pl. 42, left); the same pavement, presumably, as that at a depth of
4 feet outside Room 187 where it is overlain by 20-24 inches of
blown sand.

SILT DEPOSITS

In every deliberate test pit, within the ruin or outside, we invariably
came upon one or more of these silt layers. My field notes frequently
refer to them as "pavements" because they were so smooth
and hard as to seem man-made. When more than one was encountered
only the amount of the overlying materials set each apart.
For example, 35 inches under the floor of Room 225 we exposed
the pavementlike surface of a 5-inch-thick adobe stratum with a
slight downward slope to the east and south; beneath it were three
lesser surfaces separated from one another by from 9 to 20 inches
of sand, clay, or occupational debris. A similar, or perhaps the same,
southward-sloping surface was noted 37 inches under the floor of
adjoining Room 241.

Another silt layer 4 inches beneath the northeast steps, East
Refuse Mound (pl. 76, lower), may be a record of the same flood
that deposited a like layer 19 inches below the present surface at
the northeast corner of the West Mound enclosure despite the fact
nothing comparable appeared in the extended East Mound trench
(fig. 24). Although the main valley fill offers no evidence that the
surface gradient ever differed appreciably from what it is today, I
neglected to test this possibility fully. I neglected, for example, to
run a level between the two silt layers last noted or between that under
Station 1, Northeast Foundation Complex, and the silt surface surrounding
Pueblo del Arroyo and underlying at a depth of 5 feet
the small P. III ruin Jackson saw on the opposite bank.

That buried silt layers should appear more clearly defined in some
trenches than in others is understandable, considering the characteristics
of floodwaters and flood-borne silt. That the overburden should
be sand in its infinite variety is obvious, knowing that Chaco Canyon
is bordered by sandstone whose disintegration provided the canyon
floor. Thin clay streaks, gravel lenses, pellets, and chunks of waterwashed
adobe identify in this overburden places where floodwaters
ponded.

As Bryan (1954, p. 25) observed, these pavementlike silt beds
are generally of limited area although an occasional one may extend
a quarter-mile or more. They are the leavings of runoff wandering
laterally from the main flow—silty sand retarded by grassy ground
cover. In lesser measure we noted this same retarding effect at
present-day Navaho fields where a chico bush, a handful of bunchgrass,
or a chance rock sufficed to check or turn the advance of a
muddy rivulet. Unretarded, floodwaters in greater volume might

carve a deeper path for a short distance; they might destroy all or
part of the paths carved by earlier floods. Turning first one way
and then another, such divergent courses were the "discontinuous
channels" of Bryan's Chaco Canyon observations.

FLOODWATERS

He repeatedly calls attention to features that distinguish these
lesser waterways. They were a characteristic of the second cycle of
alluviation, spreading out disjointedly while the main valley fill was
building up. They tend to be round-bottomed wherever confined
laterally, with a crescentic fill of sand, clay, and clayey gravel; they
occur most frequently in midvalley but often reach out on either
side and one actually underlies the south rooms of Pueblo Bonito.
This latter may have been directed in its course by some undisclosed
upcanyon obstacle for it was remarkably persistent. We came upon
it quite unexpectedly during the digging of our extended East and
West Mound trenches (figs. 7, 24, c).

In both exposures the one-time floodway is floored with coarse
sand in which clay streaks, pellets, and gravel lenses occur. But
these water-borne deposits are overlain by quantities of household
sweeping and the discards of constructional activities—sandstone
spalls and chunks of dried adobe mortar bearing the imprints of
building materials. The source of all this wastage was Pueblo Bonito
itself. Some of it had been dumped there early, as witness the lenticular
bed of household sweepings 10 feet below the surface at Station
180, West Mound trench (fig. 7), and the metate fragment at
comparable depth above Station 164 in the second trench, 250 feet
distant (fig. 24, c). Witness, also, the two deep-lying hearths built
on the dry bed of the channel to meet the momentary need of local
picnickers or traders from far countries.

Obviously here was a former channel that had held transient water
from time to time and that had become a common dump-ground for
the village people living nearby. As village rubbish mounted, the
floodwaters were diverted and new courses eroded. I do not pretend
to experience in such matters, but later stream channels certainly
are indicated by the higher, more or less crescentic layers of
jointed clay and silt-streaked sand seen in the East trench between
Stations 55 and 80.

Our Middle Trench (fig. 24, c), an exploratory venture between the
East and West mounds, was neither long enough nor deep enough
to bare the full history of this intermediate section. The laminated

sand and clay to the right of Station 50 clearly mark the bed of a
former watercourse, subsequently filled with debris of demolition
and it is possible the two walls beyond were erected to curb the
outward drift of that debris. The outermost wall is a little higher
and later than the other. But between and below the two is a narrow
channel in which water once flowed, 5 feet below the present
valley floor.

Running water had threatened the earlier wall from the outside
and, presumably to prevent undercutting, a willow mat, its component
rods horizontal, had been suspended against the lower stonework.
But the water continued on its way, leaving clay-streaked
sand and gravel, until one of the local engineers brought it under
control by erecting a wattled wall opposite the mat. Rotted sticks,
apparently juniper, stood here at the time of our excavation, still
thickly coated on both sides with sun-dried mud. Three sandstone
blocks above the laminated silt may have been placed there to provide
footing for the mason who built the second wall. Outside this
latter, filling what appears to be an old arroyo bed, is a succession
of sandy-clay, gravel, rocks, and water-washed village debris.

Our two major trenches likewise revealed evidence that Pueblo
Bonito farmers undertook at times to harness midsummer floodwaters
and turn them to advantage. A channel on which man clearly exerted
a directing influence appears above Station 60 in the West Mound
trench (fig. 7). The bottom of it, gouged through occupational debris
filling an earlier and broader watercourse, lies 6 or 7 feet below the
present level of the plain and is marked by a quantity of broken
rock purposely thrown there. Above the rock is a 2-foot-thick deposit
of cross-bedded sand and, above the latter, a wide crescent of
laminated sandy clay. While this second layer was collecting, a
masonry wall was built along the south side of the channel.

Shortly thereafter, as I read the record, this masonry wall toppled
outward and was lost under the continuing clay layer before a less
vigorous stream carved a shallower course outside the first and a
bit higher. To restrain this successor, or perhaps merely to check the
spread of household waste, a second and less stable wall was hurriedly
erected (above Sta. 35). Beyond this second barrier, to the
right of Station 10, are the ill-defined limits of Bryan's post-Bonito
arroyo which, in Test Pit No. 3, measured 18 feet 3 inches deep—
well below the channels described above.

What may be a continuation of this same walled waterway was
disclosed in the East Mound trench between Stations 20 and 35

(fig. 24, c), its silt-streaked bottom 4 or 5 feet below the present level
of the plain and its margin again eroded by the post-Bonito, or thirdcycle,
channel.

Bryan (1954, p. 45) believed these successive channels were entirely
artificial because they paralleled the south side of the two
refuse mounds, were confined on the downhill slope by masonry, and
were repaired or replaced whenever a section filled up or washed
out. He believed they were designed to convey upcanyon floodwaters
to fields beyond the village—a practice followed by the inhabitants
of Peñasco Blanco and Kinbiniyol (Judd, 1954, p. 59). That floodwaters
continued to flow this way despite Bonitian efforts to turn
them aside is obvious from the layers of laminated sand and silt they
laid down in passing. That they persisted in their accustomed course
long after Pueblo Bonito had been vacated is further suggested by
the water-borne gravels Dodge (in Pepper, 1920, p. 24) observed in
some of the south-side rooms.

Thus a new and pertinent chapter in the history of Pueblo Bonito
has been disclosed by these prehistoric floodways, rubbish-filled and
forgotten until bared by our exploratory trenches. It is noteworthy
that the south wall of the pueblo, its foundation 2-2½ feet above the
present level of the plain, rests upon a 10-foot-deep rubbish fill the
beginnings of which lie 8-8½ feet below that same plain level (fig. 7).
As the village continued to increase in population and ground area
its accumulating refuse piled up 18 and 20 feet, forcing westwardflowing
floodwaters farther and farther toward midvalley until they
converged to form Bryan's 12th century arroyo.

How long floodwaters had followed this east-west course where
Pueblo Bonito now stands is still open to conjecture. Our geologists
would venture no opinion. We did not attempt to dig deeper than the
floor of our two trenches and this appears to be as deep as Dodge
(in Pepper, 1920, p. 24) ventured. However, John Wetherill during
a visit to Pueblo Bonito on November 2, 1929, told me that his
brother, Richard, had dug a hole 43 feet deep between the ruin and the
south refuse mounds before finding clean sand.

East of Pueblo Bonito, 400 feet beyond Room 176, a shallow
trench exposed still another round-bottomed channel (figs. 11, 24, c).
Its presence adds weight to Bryan's belief that upcanyon runoff, including
that from the rincon back of Chettro Kettle, was sometimes
purposely directed past Pueblo Bonito to down-valley plantings. I
am loath to suggest that the long, lone foundation wall from
Room 176 to Station 2, or that the curved pair south of it, had anything

to do with directing or diverting those floodwaters but it is
possible.

SOIL AND WATER ANALYSES

Since the subsistence problem of prehistoric Pueblo Bonito was
one of our paramount subjects for inquiry, we dug a number of pits
in order to test the character of soils formerly available for cultivation.
Our first pit was located northwest of the ruin, about halfway
to the north-cliff stairway. Others were dug on the east side, north
of the upcanyon road. Because none of these seemed to offer more
than blown sand and silt they were promptly refilled against possible
injury to wide-ranging Navaho horses and only one, No. 3, left open.

Test Pit 3, 9 feet 3 inches deep and situated on the plain midway
between Pueblo Bonito and the Expedition camp, looked more promising
than the others, and so it was fenced and held for further
study (pl. 7, upper). At the suggestion of C. S. Scofield, soil
chemist and a member of the National Geographic Society's Committee
on Research, earth samples were taken at 10-inch intervals,
bottom to top, and sent for analysis to J. F. Breazeale of the Office
of Western Irrigation Agriculture. Mr. Scofield warned, however,
that the soils represented by those samples would have less agricultural
value than anticipated if, like Chaco well and flood waters previously
analyzed, they were found to contain a high percentage of
sodium in proportion to calcium. (Scofield, 1922, discusses irrigation
and alkali salts.)

Although I have already quoted freely from it, Mr. Breazeale's
report of September 27, 1924 (Judd, 1954, pp. 10-12), on the 11 soil
samples from Pit Number 3 is so pertinent to our present subject
partial repetition seems justified:

A year after receiving the foregoing report Pit No. 3 was deepened
at the suggestion of Dr. Bryan and Late Bonitian potsherds
were collected at depths between 10 and 15 feet (Bryan, ibid., p. 58).
These finds established the fact that, unwittingly, I had located this
particular 1922 test within the banks of Bryan's post-Bonito channel.
Our soil samples, therefore, represented a varied alluvium transported
from a distance and not fields cultivated by the Bonitians.

Nevertheless, that transported alluvium originated in the same
place, upcanyon, as had the annual accretions comprising the main
valley fill. And some of those annual accretions, upcanyon, assuredly
were cultivated. But, if all local soils were as impervious to water
as were those from Test Pit 3, and we have no reason to believe
otherwise, farming for a livelihood in Chaco Canyon would have
proven increasingly discouraging and, as Breazeale points out, the
Bonitians would have had ample reason for moving elsewhere.

Each of our 11 samples from No. 3 pit, originally 9 feet 3 inches
deep, was not only deficient in soluble calcium but contained sodium
bicarbonate, or black alkali, in approximately the same amount, 0.144
percent. If as little as 0.144 percent of black alkali can so tighten a
soil that days rather than minutes are required for water to percolate

FLOODWATER FARMING

The fact remains, however, that the Bonitians lived a long while
in Chaco Canyon and that farming was their livelihood. Lacking a
living stream, they necessarily depended upon floodwaters for irrigation.
Bryan believed that the best agricultural lands available to them
were the alluvial fans at the mouths of the rincons. But scattered
areas in midvalley were also favorable—the sandy borders of discontinuous
channels and places where water temporarily ponded.
Experience dictates the selection of such places.

Inherited knowledge and understanding of soils and drainage are
prerequisites in the choice of areas suitable for floodwater farming.
Those overflowed by slow-moving floodwaters are everywhere preferred
areas. The flow must wet the ground without uprooting plants
or smothering them under washed-in sand. It is the freshly deposited
film of sandy silt that renews these floodwater fields. Hence the
operator must know how to gage the amount of runoff and check
erosion; he must anticipate the muddy onrush and stay on the job
until it has passed, according to the most successful Navaho farmer
we visited in the Chaco area (Judd, 1954, p. 55).

As a result of his study of local topography and following observations
among the Pima, Papago, and other desert tribes, Bryan (1929,
1941, 1954) believed that Chaco Canyon in Bonitian times was eminently
suited for floodwater farming and that it would be equally
suitable today if annual precipitation were what it was formerly and
if there were no modern arroyo.

This present-day arroyo, now 30 feet deep and 100 to 300 feet
wide, presumably began about 1850 and has proved more destructive
than either of its predecessors. If it follows a like cycle it, too,
eventually will fill with transplanted alluvium; additional soil will
blanket the refill and a new ground cover will take root. Indeed, as I
write these lines, those very processes are being expedited by the
National Park Service in a studied program of erosion control that
includes introduction of check dams and the replanting of willows
and cottonwoods. Since stockmen have moved their herds elsewhere
and Navaho sheep and goats are restricted, it should be possible once
again to reestablish a green carpet lengthwise of Chaco Canyon, but
the pine forests that furnished timbers for Pueblo Bonito and its
neighbors cannot be reclaimed.

CLIMATIC CHANGE

Bryan was not alone in believing that the climate of Chaco Canyon
had changed toward the less humid sometime after the Bonitians
settled there. Presumably these agricultural folk were drawn to the
valley by its broad, flat acres, by an abundance of water just beneath
the surface, and by pine forests on bordering mesas. But they soon
exterminated the forests and they eventually learned that the flat acres
of the valley were altogether unproductive. They may have noted
also that floodwaters bearing upcanyon silt did not always arrive
when expected and that crops failed for lack of moisture and replenished
soil.

It was possibly a period of subnormal precipitation at the end of
the 11th century rather than destruction of the Chaco forests that
initiated the erosion culminating in the third-cycle arroyo, Bryan's
"post-Bonito channel." Yet annual growth rings in many constructional
timbers from Pueblo Bonito are so uniform in thickness as to
indicate the pines had grown where moisture was fairly constant
year after year. This fact, plus the quantities of tule or bulrushes
(Scirpus actutus Muhl.) and coarse grasses utilized with those timbers,
led Bryan to the conviction that rainfall in Chaco Canyon formerly
exceeded his 1925 10-inch estimate. He offered no geological
support for this conviction, but the botanical evidence alone led
N. H. Darton, another geologist and a guest of the Expedition
at its 1922 symposium, to share Bryan's belief in a climatic shift
from the dry toward the less dry. Bulrushes grow in wet places.

Previously and for different reasons the theory of climatic change
throughout the Southwest had been advanced by Hewett, Henderson,
and Robbins (1913), Huntington (1914), Gregory (1916), and
others. But, as emphasized by Kidder (1924, p. 54), these advocates
did not always consider the fact that an Indian requires far
less potable water than a white man and is content to irrigate his
fields with the runoff following summertime showers. Together,
Darton and Bryan believed that as little as one additional inch of
precipitation annually would go far toward restoring the habitableness
of Chaco Canyon. Together, black alkali and the 12th-century
arroyo brought an end to local agriculture; floodwaters subsequently
refilled that arroyo and buried the old Bonitian farmlands under an
additional 2-5 feet of upcanyon alluvium.