Arid Lands Colloquia (1959-1960, 1960-1961)

Publisher University of Arizona (Tucson, AZ)

Download date 02/10/2021 22:09:48

Link to Item http://hdl.handle.net/10150/303144 THE UNIVERSITY OF ARIZONA ARID LANDS COLLOOt1IA

1959 -1960 1960 -1961 THE UNIVERSITY OF ARIZONA

ARID LANDS COLLOQUI

1959 -1960 1960 -1961 CONTENTS

Page

Foreword

William G. McGinnies ...... 1

1959-1960 COLLOQUIA PAPERS

Some Considerations of the Problem of Recording Data

A. Richard Kassander, Jr. . . . . 2

Origin of the Drainage and Geomorphic History of Southeastern Arizona

Mark A. Melton . . . . 8

Prehistoric Agriculture in East - central Arizona

Richard B. Woodbury ...... 17

Natural History of the Saguaro

Stanley M. Alcorn . . . . . 23

Precipitation and Saguaro Growth

James Rodney Hastings . . . . . 30

Chemical Constituents of the Saguaro

James W. Berry and Cornelius Steelink . . 39

1960 -1961 COLLOQUIA PAPERS

Economic Alternatives for Native Peoples in Arid Lands

Robert A. Hackenberg . . . . . 46

Changes in the Properties of Water Due to the Interaction of Soil and Water

Duwayne M. Anderson . . . . . 58

Economic Evaluation of Water Conservation Practices

Sol Resnick . . . . . 65

Status of Cenozoic Geochronology in the Southwest

Terah L. Smiley ...... 69 FOREWORD

William G. McGinnies Coordinator, Arid Lands Program

This second volume of colloquia proceedings includes copies of six papers presented in the eight meetings held during the school year 1959 -1960, and four papers presented in 1960 -1961.Written papers were not prepared for two of the talks presented in 1959 -1960 and for four in 1960 -1961.

The purpose of processing these papers is to make them available to the members of the University faculty and their associates. As the results presented are often preliminary and the treatment of the material is informal, it is not intended that this report be considered as publication. Citation of any of this material should only be made after personal communication with the author.

The Arid Lands Colloquia were oriented toward promoting the "interdisciplinary approach ". The programs were designed to keep the participants in the Arid Lands Program up to date in each field. Speakers were requested to keep this in mind in preparing material and to bring out points of special interest to workers in other fields.Ample time was also provided for questions and discussion and this gave an opportunity to bring out interrelationships.

The papers presented during the second year were mostly progress reports of research under way in the Arid Lands Program. The 1959 -1960 series was introduced by a general paper on the recording of data, followed by reports on some geologic and prehistoric aspects of the Arid Lands Program.The last three meetings were devoted to various aspects of research on the saguaro including its life history, the relation of growth to precipitation and some of its chemical characteristics.

During the 1960 -1961 school year the program included general information papers as well as reports on specific phases of arid land research. As in the previous Colloquia as many different phases of the arid land program as possible were included.The aim in the third year was to round out the three year colloquia program, by covering various aspects that had not been included in the first two years. 2

SCME CONSIDERATIONS OF THE PROBLEM OF RECORDING DATA

A. Richard Kassander, Jr. Director, Institute of Atmospheric Physics The University of Arizona

To understand the fundamental biological and physical mechanisms of the various processes at work in an arid land, one rather quickly proceeds from the purely descriptive phase of investigation to a phase involving the measurement and recording of the variables thought to be influential. This rapid procession has taken place in several of the arid lands projects, emphasizing further that one of the desirable features of an interdisciplinary program is the ability to call on the techniques of one investigation to be applied to a new one in a different field.

Although most modern laboratories and many field investigations are well equipped with recording apparatus of many different types, the problem of in- strumenting a new measurement is sometimes very subtle and perhaps more involved than is suggested by a first look at an instrumentation catalogue. Even when the recorder appears to be working satisfactorily, the results may cover the whole spectrum from inadequacy to deception.Thus,it is worth while to look at some of the basic concepts involved in recording.

A complete recording system may be thought to contain five basic elements. These are:

1. The Sensor 2. The Interpreter 3. The Recorder 4+. The Analyzer 5. The Computer

The sensor is that element which, when acted on by the variable of interest, converts the energy of the variable into some form which is more efficiently recorded. Typically, the sensor converts temperature, pressure, moisture, displacement, light intensity, or any of the myriad of other imaginable variables into an electric variable such as voltage, current, resistance, or frequency change. This is not necessarily the case, but for the purpose of this discussion, it will be well to restrict interest to systems whose sensors produce an electrical "output" as a function of the variable of interest, i.e., the "input."

The interpreter (a term not in general usage) is here thought of as the element of the system which renders the sensor output compatible with the required recorder input for reasonable scale deflection. It may be simply an amplifier or a battery. In more complicated systems it may be an impedance- match- ing device, a rectifier or demodulator, or an electronic device to linearize a non - linear output from the sensor.

The recorder is the "memory" of the system. It must render the data into some form that can be analyzed at a convenient time after the data have been recorded. Sometimes the recorder may be, or may take the place of, a pencil and paper. More often the recorder is used to make a record of data that are changing too rapidly for manual notation, or to eliminate the requirement of constant human attendance on an experiment. 3

The analyzer reduces the data to such form that subsequent mathematical, statistical, or merely speculative use may be made of it. If the recorder involves magnetic tape, punched cards or paper tape, or some form of ink recorder using electrically conducting ink, the analyzer may be more electronic equipment for optically sensing holes in cards, detecting magnetic variations on magnetic tape, or following a conducting -ink line on the recorder. In the same sense as the interpreter was the intermediate step from sensor to recorder, the analyzer mustperform the intermediate step from recorder to the element here thought of as final in the system.

The computer is the final element in the basic recording system. In practice, it may be anything from a simple electric- counter to the most elaborate electronic computer. Its function is to produce the final number or numbers describing the experiment for which the system has been designed.

Recording systems have an infinite number of possible variations. Every experiment has involved a system analogous to the one described, even though a graduate student may have performed every one of the five basic functions of the system. Since almost every conceivable experiment is limited by personnel, finances, or time, careful design of recorder systems to make optimum use of available resources or minimize the limitations, can pay large dividends in productive research. Just as important experiments have failed because one variable was omitted, so also have expensive and elaborately instrumented experiments failed because it proved impossible to analyze the data.

Ideally, a recording system should be designed by starting from the back, or the computer element. The basic equation or relationship to be tested should be clearly established. Then the number of variables needed and the minimum acceptable accuracy required should be established. The manner in which these variables will be combined should be planned, and it should be known whether the final step will be via a graduate student with a desk calculator or the IBM 650 electronic computer. The assumption that data once recorded or on punched cards can surely be analyzed scmehow, can be a devastatingly costly mistake.When the choice of experimental objective is made and the availability of final analysis and computing facilities determined, the rest of the system falls into place fairly well, provided that sensors are available for the variables of interest. A survey of recording systems will immediately suggest that the requirements for "automation" and " datomation" have been so extensive that well -designed and, in many cases quite reasonably -priced components are available for almost any imaginable recording operation.However, a sensor for the particular variable to be measured is sometimes a difficult thing to find, since original research, by definition, suggests that the particular measurement under the particular conditions of interest has never been made.

A complete discussion of all types of recording systems is obviously far beyond the scope of this paper. Competent people are available, particularly in the University's Electrical Engineering Department and the Numerical Analysis Laboratory, to advise on any aspect of such systems, no matter how complex they might be.

Discussions of recording problems with a number of workers in the Arid Lands Program indicate that they are not yet ready to consider complete systems involving electronic digital computers as the final step.Moreover, they have found ways to measure the variables of interest, but are restricted to the use of single variables recording on meters. The next important step is to multiple - variable recordings, particularly in field work where power is not available. This lack of power is a difficult obstacle to highly refined systems. Modern technology from satellite and remote weather station development offers important possibilities, but the expense of these systems is still very great. The most practical recording system will still probably involve either photography of meters or the use of clock - driven recording milliammeters. These are not the best components for economical data reduction but they probably still represent the best financial compromise.

When one considers the problem of recording several variables, the first decision must be as to whether the variables are to be recorded "in series" or "in parallel." Serial recording involves the sequential sampling of each variable on a single record. Parallel recording involves the simultaneous recording of all variables on a multiple channel recorder or on a number of recorders. In parallel recording we must have as many channels of recording capacity as we have variables and a corresponding number of amplifiers or other interpreter elements. In the case of serial recording the multiplicity of recorders and amplifiers is replaced by one amplifier (hopefully), one recorder, and a sampling switch. Thus in the serial system, °expense is essentially l /Nth that of parallel recording, where N is the number of variables. However, the cost saving is at the expense of time resolution, since, assuming that the speed of response of the recorders in the two systems is comparable, each variable in serial recording is examined l /Nth of the total time. Fortunately, in typical biological problems, the time variations are slow enough that serial recording is adequate. In many cases, such as those involving atmospheric turbulence, this is not true, but very often only the mean is required so that introduction of lag into the sensor to damp the higher frequency oscillations will render the variable amenable to serial sampling without loss of essential information.

Let us then assume that the recording is likely to be done either by photographing a meter or using a recording milliammeters such as the E_sterline- Angus, probably the best -known although by no means the only example.Although recorders of the self -balancing potentiometer type (Brown, Bristol, Leeds, and Northrup) are more accurate, they do require considerable power and are not :. nearly as well- adapted for field use. Much of the analysis which follows, how - ever, can be applied to these recorders. Even though the theory of the servo system is much more complex than that which is to be presented, the basic relations between sensitivity, speed of response, and frequency response characteristics are analogous even though mathematically not identical.

Basically, the recording device we shall consider here will use the d'Arsonval movement. This movement consists of a permanent magnet surrounding a suspended coil of wire to which a needle or pen arm is attached. The coil suspension is a torsion spring of some type so that the coil is always returned to its equilibrium position. The magnetic flux is parallel to the plane of the coil so that when an electric current is passed through the coil the resulting electromagnetic field opposes the permanent magnetic field and the coil rotates against its spring an amount proportional to the current. The basic d'Arsonval movement is shown in Figure 1.

1The flat rectangular coil BCDE of n turns surrounding a fixed cylinder of soft iron, is suspended by the torsion fiber F between the poles NS of a strong magnet so as to lie in a plane parallel to the lines of force when no current is flowing. The current i enters the coil through the upper suspension, and leaves below.

1 Page and Adams, Principles of Electricity, D. Van Nostrand and Co., 1931. 5

If 1 is the length DE = BC, the force on DE due to the field H of the magnet is iH1 in the backward direction.An equal forward force is experi- enced by BC. If h is the width EB = CD, the torque due to this couple is iHlh. This is the total torque on one turn, as the currents in the horizontal arms are subject to no turning moment.Multiplying by the number of turns, the torque on the entire coil is seen to be niHA, where A is the area lh of the rectangle.

The soft iron core, which the coil surrounds, concentrates the lines of in- duction in its interior. Therefore the field is nearly radial for small deflections of the coil, and the torque which it exerts on the current remains unchanged.

The torque of restitution exerted by the torsion fiber is proportional to the deflection angle C, and may be written k( where k is the springcon- stant of the fiber. Equating this to the torque on the coil, we have for small deflections k i= 01\. (1) nAH

The deflection sensitivity in terms of angular displacementper unit current may be written

= nAH . i k (2)

This coil represents a mass suspended by a torsion spring and thus in rotation is a torsion pendulum.The period of a torsion pendulum is given by

(3) where I is the moment of inertia of the mass about the point of support andk is the same spring constant as referred to above. The thin, flat coil of wire will have a moment of inertia closely related to the product of itsmass and its area, and since its mass, for a given size, wire will be related to the number of turns in the coil, we can crudely combine (2) and (3) to write

CT2 which states that the sensitivity of our recorder movementmay be expected to be approximately proportional to the square of its natural period. This is a surprisingly general observation. Those familiar with an analytical balance have observed that the sensitivity is always increased by moving themass on the pointer arm closer to the fulcrum, thus substantially increasing the time per swing of the balance.The above conclusion also indicates that we can avoid intermediate amplification between sensor and recorder by choosing the slowest possible recorder. However, we do this by sacrificing response time in the recorder. Consideration must be given to the relation between the most rapid fluctuations that must be recorded and the natural frequency of there- corder.

Much has been written on the subject of free and forced vibrations of oscillatory systems. For the purposes of our discussion this can be summed up 07PEN, STYLUS, OR -MIRROR AXIS

PEN, STYLUS, OR MIRROR ATTACHMENT

FIGURE I. Generalized d'Arsonval Galvanometer Movement. 6

in figures 2 and 3. These curves are drawn for systems that are considered "well- damped." This means that either electrical, magnetic, or viscous damping has been introduced into the meter movement to keep the meter from reson- ating when near its natural frequency or from 'over- shoot" when responding to a sudden transient. Damping is essential and is either included in every recorder or instructions are given for the use of shunt resistance across the recorder or meter to accomplish this.

Figure 2 shows the response of a galvanometer -type movement to a sinusoidal oscillation. By "relative amplitude" is meant the ratio of indicated amplitude to true amplitude. By "frequency ratio" is meant the ratio of driving frequency to natural frequency of the recorder. This figure, interpreted in terms of the period of the variables indicates that for less than 5 per cent error in there- cording, the shortest period fluctuation to be recorded should be at least twice that of the natural period of the recorder. If the period of the variable to be measured is half that of the recorder, it can be anticipated that greater than 75 per cent error will result:

Figure 3 shows the response of a damped galvanometer -type movement to a "step function" or step -wise change of the variable to be measured. In this case the time scale has been expressed in the number of natural periods after the step change has taken place. The curve shows that approximately ten natural periods of the recorder must elapse before less than 5 per cent error can be expected: This curve is important when considering serial recording of data since the switching from one variable to the next represents a step function to the recorder.Thus the switching interval should be at least ten times the free period of the recorder, yet should be such that the highest frequency in the variable is well sampled. In designing a recording system that will give a faithful representation of the variable, these factors are particulary important.

The following table gives some figures on representative commercially available recorders and galvanometers, showing some relations between period and sensitivity:

Manufacturer Type Natural Period Deflection Sensitivity (seconds) (millimeters /microampre)

Pimex Optical 25 18,000 (at 1 meter Galvanometer distance)

Esterline -Angus Pen Recorder 0.05 (approx.) 0.116

Sanborn Heated Stylus 0.022 0.001 Recorder

Century Optical 0.0025 0.065 (at 30 cm. Galvanometer distance)

The derived relation between deflection sensitivity and period should not be expected to hold in the above table since that relation was derived to show the angular deflection versus the natural period. Moreover, the proportionality "constant" which involves the magnet strength, certain geometrical factors, and the coil wire diameter, obviously will be different for the different movements. Moreover, the recorders that have to drive pens or styli will obviously have heavier movements than the optical recorders that typically need only a small 7 mirror to direct a beam of light to scme photosensitive surface or optical system. However, the curves of Figures 2 and 3 are perfectly general and will apply for the natural frequency figures given above.

Figure 4 shows the record of a serial recording system used by J. R. Hastings of the Institute of Atmospheric Physics. It was desired to record the circumferential variations with time of nine saguaro cacti. Bands were placed around each cactus with spiral slide -wire type resistors at the end of each band. The bands were arranged so that increase or decrease of the circumference caused a sliding contact to move over the resistance. This was the sensor. The interpreter in this case was a battery and some fixed resistors chosen so that full -scale traverse of the recorder was accomplished for full scale motion of the resistor.A rotating switch was attached directly to the chart drive of the recorder. The switch was on thirty seconds, off thirty seconds and had ten positions. Thus the complete cycle was repeated every ten minutes. The long bar on the record corresponded to a reference resistor, and the intermediate bars represent the sequential sampling of the circumferences of the nine cacti. GALVANOMETER FREQUENCY RESPONSE

I I I i 1 1 1 o 0.5 1.0 1.5 20 FREQUENCY RATIO (f)

FIGURE 2. Frequency response of well- damped galvanometer. ó STEP FUNCTION I- RESPONSE J 2 1.0 - 4

i i i 1 1 o TIME (naturalperiods)

FIGURE 3. Step- function response of well- damped galvanometer. viILYI ""o"*" wduu\l \ \ \1 \tlI[ \\\\\\N .0 \\1 \a L \ \ \ \UV \ \ \\ \tllilL\ \n\ V1\ \lt:í` \`1sáili\ \IVll \ \M + \1 \1 \ \ \\ ' I! ' 1111111 .ü11M111ldill \ WA I iXV! Ili ll1'II'itIflhII'll JILl IIi1141 f1 1 Ili IllhIft'M Í i Ili il AI/A 111k i L Ili ty,r1 i A 1 ' Iii ri ' i %I r A t ,1,1 7 J l1 / -- I : ,l' 8

ORIGIN OF THE DRAINAGE AND GEOMORPHIC HISTORY OF SOUTHEASTERN ARIZONA

Mark A. Melton Department of Geology The University of Arizona

Introduction and Acknowledgments

The general and historical geology of the Safford Valley, Graham County, Arizona, is under intensive study by faculty and students of the Department of Geology, University of Arizona. In connection with this, the writer has undertaken investigation of the geomorphic history of the area, particularly with the object of finding out details of the drainage development: changes of course and discharge of ancestral rivers, source areas of river sediment, and probable causes of changes of river regimen.*

The logical starting point of the work was to develop a theory which would adequately explain the origin of the transverse segments of the Gila River across the Mescal and Dripping Springs mountain ranges, at the lower end of the Safford Valley, and between the south end of the Gila Mountains and the north end of the Peloncillo Mountains, at the upper end of the valley. The theory must of course be consistent with the known and inferred regional geological and geomorphic history, and should help explain physiographic and sedimentary features within the Safford Valley and related areas.

The writer wishes to acknowledge the valuable help received from Drs. J. W. Harshbarger and J. F. Lance, University of Arizona, and Dr. L. A. Heindl, Groundwater Branch, U. S. Geological Survey.

The Problem of the Modern Drainage

Characteristics of the Modern Drainage -- The major drainage courses in the southeastern quarter of Arizona exhibit the following characteristics:

(1) The rivers alternately follow structural valleys in a northwesterly direction and cross the intervening mountain ranges through erosional gorges in a southwesterly direction. These gorges have general trends at right angles to the mountain ranges they cross. They show the effects of two or three major episodes of downcutting: the higher (older) parts of the gorges evidently were maturely developed into broad valleys before renewed downcutting produced the inner, deep and narrow gorges.

(2) The transverse segments of the Gila, Black, Salt, and San Francisco Rivers show a remarkable alignment one with another or with major gaps through the ranges that are not now occupied by rivers. This alignment could be fortuitous, perhaps, but there is a reasonable explanation for it.

*A shorter, preliminaryversion of this paper has been published in The Arizona Geological Digest, V. III, 1960. 9

(3) Nowhere does a river pass through a major granitic or gneissic mountain block, such as the Santa Catalina, Finale-1i°, or Chiricahua Ranges. However, "dry passes" through the granite ranges do exist, and they are probably of erosional origin, although localized by structures; e.g., Redington Pass between the Santa Catalina and Rincon Mountains, Eagle Pass between the Santa Teresa and Pinaleño Mountains.

(4) Underlying Recent and Late Pleistocene river and fan sediments in the longitudinal valleys, fine- grained clastic sediments, diatomite, limestone, and tuff form a series that has been only slightly affected by diastrophic movements, although locally, near volcanic centers, considerable deforma- tion has occurred. In many cases the fine -grained deposits lie in deposi- tional contact with the adjacent mountain fronts with a zone only a few feet thick of essentially unrounded cobbles and boulders intervening. Only where evidence of a definite channel can be seen do the coarser sediments extend any distance away from the mountain front into the finer lake beds. Vertebrate fossils found near the top of this sequence are of Blancan and Post -Blancan (Kansan) age.

In the Safford Valley, the base of the fine -grained sediments lies at a lower elevation than the present bedrock channel of the Gila River at the exit from the valley, at the Coolidge Dam (2286 feet elevation), and they have not been found higher than about 3650 feet elevation. "Lake beds" similar to those in the Safford Valley are found in the San Pedro, Dripping Springs, Tonto, Sulphur Springs, and Salt drainages, but not in the Santa Cruz Valley near Tucson.

(5) The transverse segments of the Gila River do not pass through the lowest part of the mountain ranges traversed. The general elevation of the Mescal Mountains near the Gila Gorge is about 5,000 feet above sea level. Other low points on the perimeter of the Safford -San Simon Valley are found at elevations of 4730, 4390, 4200, and about 3600 feet.

(6) The linearity and structural relations of the longitudinal basins strongly suggests that they are fault controlled. However, gravity surveys fail to show evidence for mountain -block -bounding faults, and in only a few cases are these known to exist. The basins are therefore not grabens, and a probably large fraction of the basins' width and depth is the result of long- continued erosion and pedimentation. The pediments, since their formation, have undergone dissection, burial, and exhumation.

These are the salient points from a geomorphic standpoint, however, more points could be listed. The theory that will be presented was constructed especially to explain these particular points.

Earlier theories -- Numerous attempts to explain the landform features of the region containing the Safford Valley can be found in the geological and geographical literature. These are mostly well known and need not be reviewed in detail here (e.g., Waibel, 1928; Sauer, 1930; Davis, 1931; Howard, 1942; Gilluly, 1949; Tuan, 1959). As one example, Waibel believed that following a long period of structural stability and erosion, the present mountain ranges were uplifted in Pliocene time; faulting and volcanism produced a large number of basins of interior drainage.Erosion of the mountain masses resulted in pediment development and partial filling of the basins with locally- derived, coarse, clastic debris.. A change to a wetter climate allowed the accumulation of excess water in lakes, and fine -grained, lacustrine deposits were laid down. The eventual overflow of some of these lakes at the lowest point of the basin 10 perimeter produced a successively -integrated drainage system.Later, the lower basins "captured" the higher ones by headward erosion, somewhat as in the standard description of the arid cycle of erosion. The Willcox Playa remained inter- nally drained because of supposed greater aridity in that area. I shall call this the "integration by lake overflow" theory.A number of difficulties make any simple lake - overflow theory unacceptable:

(1) Transverse gorges pass through parts of the mountain ranges where the general elevation near the gorges is greater than elsewhere on the basin perimeter - -the Gila River gorge through the Mescal and Dripping Springs Mountains, for example.Similarly, the Gila River enters the Safford Valley by a gap through mountains whose general elevation is considerably greater than the elevation of the same range, the Peloncillos, farther to the south. The deposition of the "lake" sediments was not contemporaneous with the uplifting of the mountains, and must have begun long after diastrophism had essentially ceased. The "lake beds" near the San Carlos Dam, at the entrance of the gorge, are undeformed, or at most show signs of slumping. Therefore, it is most unlikely that post- overflow uplift could have raised the regions near the hypothetical point of overflow.

(2) The lake- overflow theory cannot explain the alignment of the transverse river segments except as the result of chance.

(3) The lake- overflow theory does not explain why the fine - grained "lake" deposits in many cases lie directly against the mountain front. Hypothet- ically, the lake stage would precede an integrated -river stage; it is probable that the vast amount of fine sediment was imported by rivers from scme other, possibly more humid area, and was not the product of local weathering.

A more fruitful hypothesis was anticipated by Sauer (1930, p. 375), in his conjecture that the sequence of broad mountain passes followed by the main line of the Southern Pacific Railroad could have been the former course of a major, transverse river. Sauer, however, despaired of ever finding supportive evidence.

Possible modes of origin of a transverse drainage system -- There are four plausible ways in which the transverse river segments present today could have originated:

(1) Lake overflow; this theory has already been discussed and rejected.

(2) Headward erosion and capture by favored tributaries to an original, entirely longitudinal system of rivers.This requires that the transverse river segments lie along cross faults or zones of weakness which would have favored certain tributaries over the major rivers. Evidence for such faults or zones cannot be found. In addition, this theory does not explain the alignment of transverse segments, and so is rejected.

(3) Regional superposition of a southwesterly- flowing drainage system, followed by capture by tributaries eroding headward along major faults striking northwest. This theory requires a post -orogenic, more -or -less planar cover whose surface sloped to the southwest from about 9,000 feet above sea level near Alpine, Arizona, to about 4,000 feet near Tucson. The hypothetical cover might have been either a sheet of lava, tuff, or alluvial sediment. Consequent streams on this surface, upon eroding 11

downward, would be positioned across the axes of the mountain ranges. After further downcutting and a series of captures, most of the southwesterly- flowing drainage would be diverted to courses along the structurally -controlled valleys. Although this would explain the alignment mentioned, there is no evidence that the required cover ever existed. In fact, postulating such a vast cover would raise more problems than it solves.

Regional superposition on a regional scale seems very unlikely, although local superposition from a volcanic or sedimentary cover may have occurred in scme instances.

(4) Antecedence of pre -orogenic, southwesterly- flowing rivers.Accord- ing to this theory at least four rivers flowed from northeast to southwest across southeastern Arizona prior to the major uplift of the present mountain ranges.These rivers maintained their courses throughout the period of uplift. Following downcutting, perhaps induced by gentle regional tilting, northwesterly - flowing tributaries following the major faults successively captured the transverse rivers. This is the theory favored by the writer, and is explained in detail in the following section.

Theory of Antecedent Drainage

Pre -orogenic rivers -- If we assume that the straight, aligned, transverse segments of modern rivers are remnants of pre -orogenic rivers, and are therefore older than the longitudinal river segments as well as at least most of the mountain -building activity, then it is necessary to postulate that at least four major rivers flowed across southeastern Arizona before the present topography came into existence. The earliest direction of flow is not definitely known; the writer favors a southwesterly flow because that is the direction of the modern drainage, and also because that would fit well with the regional pattern of radial drainage away from the Rocky Mountains in Colorado and northern New Mexico. The San Juan Mountains in that case could have been the major source of water. This system could have been established in the early Tertiary. The courses and proposed names of the four hypothetical rivers are shown in fig. 1.

Evidence of three, probably smaller, rivers positioned between the major rivers has been found:

(a) From the Final Mountains across the Mescal Mountains and Troy Gap in the Dripping Springs Mountains. (b) From the northern part of the Santa Teresa Mountains across the Galiuro Mountains near the present gorge of Aravaipa Creek (L. A. Heindl, personal communication). (c) From the Pinaleio Mountains north of the Winchester Moun- tains and through the Redington Pass area, between the Rincon and Santa Catalina Mountains.

No names are proposed for these rivers.

The northeastward extent of the old rivers is not known, but it is reasonable to suppose that the Clifton River (fig. 1) may have extended via the Plains of San Augustin from the vicinity of Albuquerque, N. Mexico, and perhaps drained ó-

o + n ?ó O y a tñ ó E ó

I I I I I I I I I I I I I I I I I

IIIIIIIIIIIIIIIIIill

M M 12 the upper Rio Grande watershed.Neither is the southwestward extend of the rivers known, and evidence on this is lacking.

Independent evidence -- Support for the theory of antecedence that is independent of the features listed above falls into three categories:

(1) Evidence that a transverse drainage system was formerly more extensive than now is found in a relatively high -level surface that extends from the New Mexico -Arizona line near San Simon westward at least as far as Gila Bend.Existing remnants of this surface are various accordant (i.e., of approximately equal elevation) summits of many of the smaller mountain ranges, and shoulders or spurs on the sides of the larger ranges. The surface itself must have been a compound erosional and constructional surface, perhaps consisting of a large number of coalescing pediments and bajadas whose general level was controlled by (probably) southwesterly - flowing rivers. The name, Tortolita Surface is proposed, because the Tortolita Mountains north of Tucson are about in the center of the known extent of the surface, and because the accordant summits of the Tortolitas provide striking evidence for its existence.

The Tortolita Surface declines slowly from an elevation of about 4,950 feet above sea level near the Peloncillo Mountains in southeastern Arizona, to about 4,250 feet in the Tucson area, to 3,950 near Phoenix, to perhaps 3,000 feet near Ajo. "Form lines" on the reconstructed surface would have roughly a NNW -SSE trend. Other remnants of the Tortolita Surface are believed to be part of the summit of the Tucson Mountains (4,200), Redington Pass (ca. 4,350), accordant summits of the Dripping Springs Mountains (4,400), the sub -volcanic granite surface north of the Pinal Ranch, Pinal Ranch Quadrangle, Arizona (4,400 to 4,500), the saddle of the pass through the Peloncillo Mountains southeast of the Whitlock Mountains (4,950), accordant summits of many hills north and northeast of Willcox (4,800 to 4,900); west of the Tucson Area there is a suggestion of summit accordances near Arivaca (4,100), the Sierra Estrella southwest of Phoenix (3,950), and possibly near Ajo (3,000). Much more field work will be required before existence of the Tortolita Surface in the western and west - central parts of southern Arizona can be satisfactorily demonstrated.

It is particularly interesting that the Tortolita Surface appears to be restricted to the area formerly traversed by the ancestral drainage referred to above. That is, it appears to be limited to the north by plateau structure, and no certain traces of it have been found in the higher mountains to the south. The mean gradient of the surface from the Peloncillo Mountains to Ajo is 9 feet per mile, which may be compared with the present, average gradient of the Gila River from Safford to Gila Bend, about 10 feet per mile. This is reason for believing that the Tortolita Surface represents a compound surface controlled by a number of major rivers flowing southwestward across southern Arizona during a fairly long period of relative stability following uplift of the mountains. The river system then must have contained more or longer transverse segments than now exist. This constitutes independent geomorphic evidence against the theory of integration by lake- overflow and the theory of capture of an original, entirely longitudinal system of rivers and lakes by headward erosion of favored tributaries, and is evidence favorable to the theory of antecedence. 13

An older, higher surface may have existed also. Possible remnants can be found at about 6,100 feet elevation in the Huachuca Mountains - Canelo Hills area, the north end of the Santa Rita Mountains, the Sierrita Mountains, the Mazatzal Mountains north of Four Peaks, and elsewhere. Correlation on the basis of elevation alone is hazardous, and as the localities mentioned are widely separated other explanations are possible.

(2) Former river channels, transverse to the ranges and buried by late Tertiary volcanic material, have been found by Heindl near the confluence of Aravaipa Creek with the San Pedro River (see above), by Harshman (1939) in Queen Creek Canyon above Superior, and by Melton in the Haunted Canyon Quadrangle, 0.7 miles south of J K Mountain. The channels found by Harshman and Melton have been affected by post -volcanic tilting; both line up well with the trend of the Rockinstraw River (fig. 1).

(3) Klaus Voelger (1953, pp. 49 -51, 76) notes the occurrence of very well rounded pebbles derived from the Barnes Conglome.rate(Apache Group, younger Precambrian age) in the lower unit of the Rillito beds on the south flank of the Santa Catalina Mountains. The Barnes Conglomerate is not known southwest of a line striking northwest through the Santa Catalina Mountains (Stoyanow, 1936). The drainage at the time of deposition of the lower unit of the Rillito beds must have come from the north or northeast, in order to bring in the Barnes -derived pebbles, and imbrication in the gravels of the Rillito beds supports this. It is believed that the Rillito beds were deposited prior to and during the uplift of the Santa Catalina Mountains, so this much of Voelger's evidence is consistent with the theory of antecedence. However, very recent field work has shown that large, sub -angular boulders of probable Barnes Conglomerate also occur in the lower Rillito beds, and these boulders must have come down from the Santa Cata- lina Mountains, perhaps before the relief was as great as now (Mr. F. Pashley, U. S. Geological Survey, Tucson, Arizona, personal communication). This throws some doubt on the value of the support afforded by Voelger's findings.

Post -orogenic events -- The sequence of events following the uplift of the mountain blocks is not yet known in great detail. However, if one accepts the theory of antecedence, certain events are necessary in order to produce the present topography and drainage pattern. The chief aim of the following section is to show that these events are consistent with what is known of the history from stratigraphic studies.

(1) Establishment of dominant southwesterly -trending drainage in early Cenozoic times, following the Laramide Orogeny. In southern Arizona, the positions and directions of the rivers were probably influenced by the northeast trend of lineation in Precambrian rocks, and locally by other structural features.

(2) Deposition of "Rillito" and "Pantano" beds and correlatives in southeastern Arizona, as basin fill.

(3) Uplift of the present mountain ranges across the courses of the rivers; downcutting of the rivers. The river courses were probably changed somewhat as uplift progressed, shifting away from areas of greatest uplift toward structurally lower areas. Basins were partially filled by coarse, clastic debris. Hypotheticaltransition stagein the drainage history of Southeastern Arizona; capture of transverse riversnot ye t complete d. Figure 3 Probable drainage of Southeastern Arizona at time of deposition of fine -grained :Lake :beds. (4) Development of the Tortolita Surface by both erosion of the mountain masses and deposition between the mountains.

(5) Further downcutting by the rivers of 800 to 1,200 feet, which may have been induced by gentle regional tilting down to the northwest.

(6) Headward growth of northwesterly - flowing tributaries and capture of the Winkelman, Clifton, ancestral Aravaipa, and other rivers. The Winkelman River was possibly captured first (fig. 2).

(7) During and after stage 6, large pediments were eroded across both deformed alluvial basin fill and mountain bedrock, thereby greatly increasing the width of the valleys.

(8) Completion of the capture of the old drainage; establishment of longitudinal drainage from Mexico in the San Simon and Aravaipa valleys (fig. 3). At this stage the rivers were probably flowing at elevations of 500 to 800 feet. The entire system, with the exception of the Santa Cruz and San Pedro Rivers, drained through or near the present Salt River gorge below the Roosevelt Dam.

(9) Gradual damming of the ancestral Salt River system by the accumulation and /or uplift of the volcanic pile extending from the south end of the Mazatzal Mountains to the Tortilla Mountains, west of Hayden. By this action, the base levels of the Salt, San Pedro, and probably the lower Verde Rivers were raised. The Santa Cruz River, joining the Gila River farther to the west, was not affected.This damming caused the deposition of fine - grained material in the valleys upstream, forming the so- called lake beds. Locally, basalt flows and intrusions and alluvial fans or deltas in the lake basins formed small ponds, such as near Mt. Triplet in the Safford Valley, in which lime and gypsum accumulated. The Upper Verde Valley, which at that time probably flowed through Rye Creek and the Tonto Valley, was dammed independently and at a higher elevation by volcanism and uplift south of Camp Verde (Jenkins, 1923,pp. 76 -78; confirmed by recent work of the Geological Survey, Ground Water Branch). The Gila River was probably also affected by diastrophism in the Globe area.

(10) In the last stages of the lake filling episode there was a large out- pouring of basalt in the Safford Valley in the Mt. Triplet area, with flows intercalated with water -laid tuffs.

(11) Probably following the volcanic eruption, and almost certainly in response to climatic refrigeration, large amounts of boulders and cobbles were shed from the higher parts of the mountains to form alluvial fans. Frye Mesa in the Safford Valley, Davis Mesa in the San Pedro Valley, and a similar fan below the northeast side of Four Peaks in the Mazatzal Moun- tains are remnants of these fans. A long period of stability followed during which a distinctive red soil was developed on the fans and also other alluvial deposits in southern Arizona. In the Safford Valley the alluvial fans were deposited on the undissected, approximately level surface of the "lake beds ". In the San Pedro there had apparently been an episode of erosion and pedimentation of the "lake beds" before deposition of the alluvial fans.

(12) A second, and generally coarser, boulder deposit was emplaced on Frye Mesa and elsewhere as a second fan, and as a blanket of ,cobbles on the floodplains of the Gila River in the Safford Valley, and the Sulphur Springs Valley near the Mule and Swisshelm Mountains. Davis Mesa in the 15

San Pedro Valley, on the other hand, does not have a second boulder deposit emplaced upon it. From this we infer that further dissection of the fill in the lower San Pedro Valley had already begun when the second gravel was coming out of the mountains; physiographic evidence indicates clearly that dissection of the Safford Valley fill followed deposition of the second boulder deposit.

(13) When the level of the river in the Safford Valley was raised to an altitude of about 3650, the approximate elevation of the gravel blanket near Safford, the Gila River reoccupied its former course through the Mescal Mountains, which had been left as a long series of wind gaps when the Gila had been captured much earlier. The gorge through the Mescal Mountains shows evidence of a broad valley stage at an elevation of around 3550 feet. As the San Pedro had already begun cutting down, and may have been as much as 800 feet lower than the Gila River, the Gila must have begun cutting down at a tremendous rate immediately upon overflowing.

Perhaps just prior to the diversion of the Gila River, the rivers that flowed into the San Simon - Safford and Sulphur Springs Valleys from Sonora were captured and diverted to the south by the Rio de Bavispe, which follows a more direct course to the sea.

(l4) The erosional history of the Safford Valley is complicated by six or more stages of lessened rates of downcutting, recorded as terraces and pediment levels. The valleys to the west have fewer of these levels, but their history is somewhat more complex.

Placement of these events in the geological time scale is uncertain, but a tentative dating is as follows: Origin of the transverse drainage system, Eocene or Oligocene; deposition of the Rillito and Pantano beds, Upper Oligocene to Lower Miocene; Basin and Range Orogeny, Lower Miocene; development of the Tortolita surface and extensive lower pediments, Upper Miocene; deposition of the "lake beds," Lower Pliocene to Aftonian; deposition of Frye Mesa, Davis Mesa, and similar alluvial fans, Illinoian; development of the red soil, Sangamon; deposition of the younger boulder deposits on the alluvial fans and on the river floodplains, early to middle Wisconsin; dissection of the Safford Valley fill and development of pediments on the "lake beds," late Wisconsin to Recent.

Conclusions

A number of generalizations can be emphasized here without excessive re- petition. Of primary importance, and contrary to the "text- book" concept of the arid cycle of erosion, drainage history of the desert basins in southern Arizona does not necessarily begin with uplift of mountains and formation of undrained basins. In the region studied, a pre- existing drainage system was able to maintain itself throughout the mountain -building epoch, probably with only minor changes of course influenced by local structures, and was disrupted not by diastrophism but much later by capture by structurally - favored tributaries. Secondly, desert basin fills are not necessarily entirely the product of erosion of the bounding mountain ranges. At times, erosion in the mountains is sufficiently slow to allow rivers to import fine -grained material and to deposit it next to the mountain front. Silt and clay -sized material can also be derived locally in small quantities. The cause of deposition of the "lake sediments" in the region studied apparently was not related to the primary 16 mountain uplift, but to the accidental coincidence of the mouth of the major drainage with a center of volcanism and local uplift that raised the base level of the entire system. Thirdly, the large pediment surfaces seen today were cut relatively lately, almost entirely on "lake beds" or alluvial fan material. In some cases, the mountainward extremity of the pediments roughly coincides with a bedrock pediment, now being exhumed, that was cut over a long period of erosion that predates the last filling of the valleys. In several instances it is known that these buried pediments extend to the center of the valleys where the major fault must lie.

REFERENCES CITED

Davis, W. M., 1931, The Santa Catalina Mountains, Arizona: American Jour. Sci., v. 22, pp. 289 -317.

Gilluly, J., 1956, General geology of central Co.chise County, Arizona: U. S. G. S. Prof. Paper 281, 169 pp.

Harshman, E. N., 1939, Geology of the Belmont -Queen Creek area, Superior, Arizona: Unpubl. Ph.D. thesis, Dept. of Geology, Univ. of Arizona.

Howard, A. D., 1942, Pediments and the pediment pass problem: Jour. Geomor- phology, v. 5, pp. 3 -31, 95 -136.

Jenkins, O. P., 1923, Verde River lake beds near Clarkdale, Arizona: Amer. Jour. Sci., 5th Series, v. 5, pp. 65 -81.

Sauer, C. O., 1930, Basin and range forms in the Chiricahua area: Univ. Calif. Publ. Geogr., v. 3, pp. 339 -414

Stoyanow, A. A., 1936, Correlation of Arizona Paleozoic formations: Geol. Soc. America Bull., v. 4+7, pp. 459 -540

Tuan, Yi -Fu, 1959, Pediments in southeastern Arizona: Univ. of Calif. Publ. Geogr., v. 13, pp. 1 -163.

Voelger, Klaus, 1953, Cenozoic deposits in the southern foothills of the Santa Catalina Mountains near Tucson, Arizona: Unpubl. M.S. Thesis, Dept. of Geology, University of Arizona, 101 pp.

Waibel, L., 1928, Die Inselberglandschaft von Arizona und Sonora: Gesell. Erdkunde Berlin, Zeitschr., Sonderband zur Hundertjahrfeier, pp. 68 -91. 17

PREHISTORIC AGRICULTURE IN EAST- CENTRALARIZONA1

Richard B. Woodbury Department of Anthropology The University of Arizona

Agriculture began in the Southwestern United States two- thousand or more year ago, and long before the arrival of the first white explorers it had become the principal means of subsistence of most of the aboriginal population. It was always supplemented to some extent by hunting and by the collecting of wild plant foods, but as groups increased in size they depended increasingly on farming. In spite of the impressive amount of research on the prehistoric peoples of the Southwest, we still know relatively little about their farming, which was the very basis of their existence. Our ignorance is partly due to the focusing of research on other problems, but equally to the difficulty of studying an activity that left so few conspicuous and datable traces.

Several kinds of information have been used to good advantage in the study of prehistoric agriculture. Foremost is the identification of plant remains by botanists.2The exigencies of preservation usually result in only a few fragmentary parts of plants being available for study, and even the most expert analysis cannot often provide a complete inventory of the domestic species grown by a prehistoric group. The cumulative data from many excavations have indicated that the Indians of the Southwest depended chiefly on maize, with several species of beans and squash grown as important subsidiary foods. Non -food plants were also sometimes grown, particularly tobacco and cotton. However, botanical analysis cannot provide any satisfactory evidence of the actual amounts of each species that were grown, nor evidence of the particular farming techniques practiced.

Another means of learning about prehistoric agriculture is by inferences based on the present farming practices of the Southwestern Indians, giving due recognition to the numerous changes that have occurred during the last three centuries as a result of conquest and acculturation.A few excellent studies have been made, particularly those of Castetter and Be113, Forde , Hi115, Brand°, and Hack7, butthese describe the farming of only a handful of groups and for dozens of others the information available from ethnographic reports and the writings of early travellers is scanty in the extreme. Therefore, only an incomplete and sketchy basis is provided for inferring past practices from those of the present, and any such reconstruction is necessarily uncertain and incomplete.

A third approach to prehistoric agriculture is by a study of surviving field patterns and systems of water control.There has been much interest in the large -scale irrigation that the Hohokam farmers of the Gila and Salt River valleys carried on from about the ninth to the fifteenth century, although much remains to be learned of the age and extent of the systems. But elsewhere in the Southwest, irrigation ditches, agricultural terraces, and other traces of prehistoric farming have been given almost no systematic attention, although they have been repeatedly observed and occasionally mentioned in print.0The present paper describes and discusses briefly the surviving field systems of one Southwestern locality, the Point of Pines area; a more comprehensive report is in preparation.9 18

The Point of Pines area is well suited to the investigation of prehistoric agriculture, because a considerable number of field systems survive, little dis- turbed except by natural forces since their abandonment some five centuries ago. Equally important, the area has received intensive archaeological investigation for the past fifteen summers, by the University of Arizona's Archaeological Field School, so that the general cultural development and the chronological sequence of events are quite well understood in comparison with many other areas of the Southwest. Briefly, the area has been occupied by man for three or four millen- nia, and by the eighth or ninth century A.D., if not before, small villages of farmers were permanently established there. Increasingly large clusters of masonry rooms were built as time went on, and the architectural climax was reached in the fourteenth and fifteenth centuries, with the population peak probably occurring then also. Shortly thereafter the villages were abandoned and the area was uninhabitated until Apache bands filtered in probably after the sixteenth century.

Research at Point of Pines was initiated for the Arizona State Museum by E. W. Haury and E. B. Sayles; both they and their numerous students, field assistants, and colleagues have collected the information used in this study, and I am indebted to them for the opportunity to use it. The excellent maps prepared by Murray C. Gardner in 1958 have been especially helpful. In 1959, I visited as many of the Point of Pines farming sites as possible and made additional observations and notes.

Point of Pines is located in the Mogollon Rim country of central Arizona, just north of the Nantack Ridge (also called the Natanes Mountains), about sixty -five miles east of Globe and forty miles north of Safford. A large grassy prairie with an elevation of about six thousand feet is surrounded by higher land covered with juniper, pinyon, oak, and ponderosa pine. The prehistoric settlements were mostly located near the edge of the prairie, usually on slight elevations or low ridges. The area is drained by numerous small, intermittent streams, few of them deeply entrenched, and many of them flowing through flat and even marshy valleys.

Two types of evidence were found for prehistoric fields. The first consists of lines of small boulders, generally laid more or less horizontally along a gently sloping hillside in groups of half a dozen to a score, the lines spaced from two to ten meters apart. I have called these "linear borders." These borders sometimes occur on level land, particularly at the Rocky Point site (Arizona W :10 :108, in the Arizona State Museum's system of site designa- tion). Here a variation also occurs, in the form of occasional stone lines perpendicular to each other, forming a partial grid,an arrangement I have named "grid border ". The second type of evidence consists of terraces in the beds of small intermittent streams with gradients of about 10 per cent to 20 per cent. These terraces were formed by building rough walls of unshaped boulders, probably less than a meter high, and from two or three to about twenty meters long. Behind each wall, the soil that accumulated formed a small plot, unusually well- watered, and suitable for intensive farming.

Associated with most of the prehistoric fields are the ruins of small one - room structures. Since the characteristic settlement pattern of these people was the village, a cluster of houses which often coalesced into a single ir- regular block of rooms, these isolated structures can be assumed to have served a different purpose. Their location leaves little doubt that they were tempor- ary shelters for the use of farmers whose homes were too distant to permit going and coming each day during the busiest part of the summer. They would also have 19 sheltered the guards, possibly children or youths, who protected the ripening crops from birds, rodents, deer, and other marauders. Such field shelters have continued to be a feature of Puebloan farming to the present,10 an indication of the conservatism of at least some aspects of Puebloan life.

The linear borders probably served several purposes simultaneously. Most of the summer rain at Point of Pines comes in brief, hard showers, and even on a slight slope the water runs off quickly without penetrating more than the uppermost few centimeters of soil. The stone lines would have helped to slow the runoff and increase its penetration of the soil, a significant benefit to agriculture. Another probable function of the stone lines was to hold the soil from washing down the slopes, since clearing for planting would leave much of it bare. At best the soil was relatively thin, and some of the hillside farms would have been rapidly washed away without this protection. Even now, after five centuries of neglect, the stone lines result in a greater thickness of soil on the upslupe side, demonstrating that movement of soil down the slope is effectively prevented.A third function of the stone lines, at least on some slopes, would have been to concentrate at the edge of each small field the stones that littered the surface and interfered with farming (this, of course, is also one of the functions of New England's miles of boulder walls on glaci- ated hillsides). Finally, these stone lines probably served as field bounda- ries, marking permanently the limits of each man's plot.

Many of the linear borders cover only limited portions of hillsides, and adjacent areas of similar soil and slope are without these visible traces of farming. No entirely satisfactory explanation can be given for this discon- tinuous arrangement, and it is probable that the intervening areas were also farmed, although with small earth ridges taking the place of the stone borders to slow runoff and impede erosion. As will be pointed out later, the farming needs of the region probably far exceeded the few acres marked with stone borders and terraces, so other locations must have been farmed in addition.

Three sites with terraces were studied in detail. At one there was a single series of ten walls across a very small stream channel, each wall only a few meters long. At each of the other two sites, three adjacent stream beds had been terraced, with twenty to fifty -seven walls across each channel. The more extensive group comprised about four thousand meters of wall, a considerable accomplishment even though the stones were unshaped, dry laid, and the walls probably never over a meter high and many of them only half this. At the best, each of these many walls could have held soil to form only a small plot of farm land, less than ten acres in all for the two groups of terraces. It has been suggested that these wall were not built to form terraces - there is no soil behind them today, and every wall has been breached by the stream in flood - but only to slow the runoff so that fields in the flats below the stream would not be washed out.This explanation seems unacceptable since only a rather small area could have been farmed at the foot of each stream and since approximately the same degree of flood control could have been achieved with less regularly spaced and uniform width walls placed partly on the main channel and partly on its small tributaries.Agricultural terraces similar to these have been Eeported at several other sites in the Southwest, particularly at Mesa Verde , and are still used by the Hopi7.With the hillside fields, marked by their stone borders, they provided additional and probably particularly desirable areas for farming.

The question can legitimately be asked, why did the prehistoric farmers trouble to prepare these hillside and stream channel plots, when considerable level land was available. Fróm studies of modern Indian farming in the 20

Southwest we can derive a reasonable answer. The primitive farmer cannot control the various natural phenomena on which he depends - temperature and rainfall, especially. But by planting his crops in a variety of locations he avoids the risk of total failure.One year a southerly exposure may be advantageously warm, but another year it may prove disastrously dry in midsummer. Some years it may be possible to plant large areas of valley bottom, but during an unusually wet year such fields may be drowned out and only the hillsides dry enough for a good harvest. During a drought year hillside fields may wither and only a few low -lying fields produce a crop. An unpredictable storm may send a flood down one stream bed and not another, so that one group of fields is destroyed and others left undamaged.Thus, the distribution of the fields of an individual farmer or of a village among a variety of locations insures against total loss and takes advantage of the special conditions of each location.

The probability of the hillside fields and terraces being designed to provide a greater variety of planting locations is increased by the fact that they could not have supplied more than a small part of the food needed by the population of the final centuries of occupation; during the fourteenth and fifteenth centuries, numerous large villages were occupied simultaneously, and the population must have been many hundred. Unfortunately, there is no simple and accurate way to estimate the prehistoric population of a region; we cannot depend on the number of graves encountered during excavations, or the number of rooms in a series of more or less contemporary villages - the errors of interpretation are too great.At best we can make only a rough guess as to the number of people the region might have supported, and from that, the amount of land the people might have had under cultivation. For example, the Hopi of northern Arizona were cultivating about 2 per cent of their reservation in the 1930's7; this amounted to about three acres per capita and provided about one half their livelihood, wages providing most of the remainder. Assuming a comparable prehistoric standard of living and crop yield (since the Hopi even as recently as the 1930's were living for the most part in traditional rather than new ways) we could consider four or five acres per capita as a reasonable amount for the prehistoric Point of Pines people to have cultivated. If no more than 2 per cent or 3 per cent of their land was arable, almost certainly an underestimation, this would be some three thousand or four thousand acres; that would have supported perhaps one thousand people. Even though this figure cannot be regarded as better than a very crude guess, it is not un- reasonable in view of the many large ruins in the area. Also, it serves to suggest that the total acreage cultivated must have been substantial, with the terraces and stone -bordered hillside fields only a part of the total.

The dating of the observable field systems in the Point of Pines area is difficult, since there are few clues in the fields themselves. A few of the farm houses had some pottery with them, which provided an approximate date for the houses, and by inference for the fields associated with them. However, a field could have been used for many centuries both before and after a particular farm house, so its date provides only one point in the unknown span of time. Only a few fields are so closely associated with a village that con- temporaneity is suggested; most of them could have been farmed from any or all of the villages within a radius of several miles. From all of the evidence so far obtained, it appears that at least some of the surviving field systems may have been in use in the eighth or ninth century, and by the fourteenth or fifteenth century, a peak was reached in population and cultivated land so that nearly all the field systems were in use at that time.

A final question to be considered is how widely the techniques used at Point of Pines for control of water supplies were distributed in the prehistoric 21

Southwest, and whether there were other important techniques not known and used at Point of Pines. Terracing of small intermittent streams was quite widely practiced, although few of the known occurrences have been reported in even brief manner. Stone borders have been less frequently noted, but are by no means restricted to the Point of Pines area. On the other hand, large -scale canal irrigation did not spread north of the desert area of Southern Arizona, perhaps due to lack of suitable river valleys or to the greater suitability of other irrigation techniques in the mountain and plateau country.At Point of Pines, there were no gardens or small plots irrigated from springs, with the water conducted through a series of small ditches, as is done by the Hopi today. Whether the techniques of spreading fairly large streams over a wide series of fields during spring and summer floods were practiced at Point of Pines is open to some question, but these techniques may not have been major because of the small size of most of the local streams and the lack of suitable broad but well- drained valleys. Wherever the stream bottoms tend to be swampy, as is often the case at Point of Pines, heavy growth of grass or other vegetation would have made the land excessively difficult to clear with primitive tools.Although the rainfall was unpredictable and many years of drought or near - drought must have occurred, no water storage systems were developed to supply water to the fields. Walk -in wells and small reservoirsl- were built in the Point of Pines area, but only for domestic water, and never became part of the agricultural system.

In sum, then, the prehistoric farmers of the Point of Pines area appear to have selected a few of the techniques known to them or practiced in neigh- boring areas for the control and the effective use of the water supply. They were able to farm several types of terrain by relatively simple but judicious- ly applied methods. They could not or did not employ a number of other techniques for the providing of an agricultural water supply, techniques used by some prehistoric groups in the Southwest. The Point of Pines area was farmed for at least six centuries, probably for many centuries longer, and even before the increasing population brought any need for maximum use of the arable land, it was found advantageous to distribute crops among varied locations, so that in no year would more than a part be lost by drought, frost, floods, or other un- controllable and unpredictable natural forces.Working with a few simple tools and techniques that were found through the centuries to be appropriate, these prehistoric farmers managed as best they could the most critical ingredient in their lives - the water supplyl2.

References

1) Contribution to Point of Pines Arachaeology, No. 20.

2) See, for example, Hugh C. Cutler, "A Preliminary Survey of Plant Remains of Tularosa Cave," in Paul S. Martin and others,Mogollon Cultural Continu- ity and Change, pp. 461 -79, Fieldiana: Anthropology, Vol. 40, 1952; Volney H. Jones and Robert L. Fonner, "Plant Materials," in Earl H. Morris and Robert F. Burgh, Basket Maker II Sites near Durango, Colorado, pp. 93 -115, Publication 604, Carnegie Institution of Washington, 1954; and Richard S. MacNeish, Preliminary Archaeological Investigations in the Sierra de Tamaulipas, Mexico, Transactions of the American Philosophical Society, n.s., Vol. 48, Pt. 6, 1958.

3) Edward F. Castetter and Willis H. Bell, Pima and Papago Indian Agricul- ture and Yuman Indian Agriculture, University of New Mexico Press, 1942 and 1951. 22

4) C. D. Forde, "Hopi Agriculture and Land Ownership," Journal of the Royal Anthrop. Inst. of Great Britain and Ireland, Vol. 61.

5) W. W. Hill, The Agricultural and Hunting Methods of the Navaho Indians, Yale Univ. Pubis. in Anthrop., No. 18, 1938.

6) D. D. Brand (ed.), Symposium on Prehistoric Agriculture, Univ. New Mexico Bull., Anth. Series, Vol., 1, No. 5, 1936.

7) John T. Hack, The Changing Physical Environment of the Hopi Indians of Arizona, Papers of the Peabody Museum of Am. Archaeol. and Ethnol., Harvard Univ., Vol. 35, No. 1, 1942.

8) See particularly Guy R. Stewart, Conservation in Pueblo Agriculture, Scientific Monthly, Vol. 51, Nos. 3 and 4, pp. 201 -220 and 329 -40, 1940; Guy R. Stewart and Maurice Donnelly, Soil and Water Economy in the Pueblo Southwest, Scientific Monthly, Vol. 56 Nos. 1 and 2, pp. 31 -44 and 134 -44, 1943; Kirk Bryan, Flood -water Farming, Geographical Review, Vol. 19, No. 3, pp. 444 -56; and J.O. Brew, Archaeology of Alkali Ridge, Southeastern Utah, Papers of the Peabody Mus. of Am. Archaeol. and Ethnol., Harvard Univ., Vol. 21.

9) Richard B. Woodbury, Prehistoric Agriculture at Point of Pines, Arizona (in preparation).

10) Frank H. Cushing, Zuni Breadstuff, Indian Notes and Monographs, Vol. 8, 1920, (p1.5).

11) J. B. Wheat, Prehistoric Water Sources of the Point of Pines Area, American Antiquity, Vol. 17, No. 3, pp. 185 -96, 1952.

12) The research which this paper describes briefly is part of a larger program of investigation of the prehistoric methods of water control in the arid portions of North America, particularly southern Arizona; eventually the North American data should be compared with those from South America and the Old World, in order to evaluate the role that various water -control techniques have played in different cultures of the past and present. 23

NATURAL HISTORY OF THE SAGUARO

Stanley M. Alcorn Crops Research Division Agricultural Research Service United States Department of Agriculture

Knowledge of the natural history of the saguaro is, unfortunately, far from complete, though it started with the chronicles of the early pioneers (Bancroft, 1889; Bourke, 1891; Pfefferkorn, 1795), was expanded through the efforts of those associated with the Carnegie laboratory in Tucson (Britton and Rose, 1920; MacDougal, 1905, 1908; Shreve, 1910, 1911, 1929, 1951), and now is being further interpreted by the research of a number of scientists in several departments of The University of Arizona and in the United States Department of Agriculture. Some of the following discussion is based on the results of current investigations, in considerable part unpublished, by J. A. Booth, G. D. Butler, Jr., J. R. Hastings, E. B. Kurtz, Jr., C. H. Lowe, Jr., and R. M. Turner, of The University of Arizona, and S. E. McGregor of the Department of Agriculture

Flowering and Pollination

Flower buds are commonly first seen on saguaros in the vicinity of Tucson during the middle or latter part of April.Their number per branch (arm) varies greatly within a given year, and between years. Collections suggest that forth to fifty buds per branch would be a rough estimate. Studies have shown that damage by the larvae of the Cactobrosis moth, Cactobrosis fernaldi- alis, may effect losses of 20 to 30 per cent of these buds.A number of others die for unknown reasons before they flower, leaving approximately 35 per cent of the buds surviving to maturity.

Pollination requirements definitely affect flower -bud survival.The saguaro flower usually opens at night and closes the following afternoon (Benson, 19+0; MacDougal, 1908; Peebles and Parker, 19+6). Thus, in less than twenty -four hours, pollination must be effected or the flower will not set as a fruit. Morphologically, the saguaro blossom is epigynous, perfect, and regular, with filaments adnate to the tubular corolla. The stamens are in such profusion that the interior of the corolla, commencing near the junction with the receptacle to a point just below the corolla -tube lip, appears as a solid golden mass of anthers. The hollow style is several centimeters in length and tipped with several stigmas.When the flower first opens, the stigmas are usually clumped. The style often elevates them above the anthers. The stigmas gradually open and radiate from the style, particularly after dawn. As the day progresses, the style shrinks, withdrawing the slowly wilting stigmas into the corolla -tube.

Changes in the positions of the stigmas relative to the anthers in several other cacti suggested (Kerner, 1897) that such blooms are capable of being either cross- or self -pollinated. However, tests have shown repeatedly that

1 Appreciation is expressed to John Cook, Superintendent; George Olin, Naturalist; and Barton Hirschler and John Lewis, past Superintendents Saguaro National Monument, U. S. Department of Interior, for cooperation in conducting many of these current investigations. 24

self -pollinated buds of saguaro will not set.Maximum set is achieved when the pollen source is another plant (Alcorn et al, 1959). The observations of Kurtz (1958) suggest that saguaro pollen is too large to be effectively disseminated by wind. It would therefore seem axiomatic that to produce fruits, the saguaro needs pollinating agents other than wind. Honeybees, Apis mellifera, are seen in profusion around blooms (McGregor et al, 1959) and are well known to be generally effective as pollinators (Vansell and Griggs, 1952). However, the honeybee has been in the West only since the late 1800's (Todd, 1950), too late to account for the seed production prerequisite for the establishment of most of the mature plants in the present large saguaro forests. Although the honeybee may now facilitate the fruit set necessary for the establishment of future saguaro forests, investigations of other possible native pollinators are in progress. The importance of determining such agents and their flight habits is particularly apparent when one considers the probability of seedlingprogeny becoming established around a mature saguaro, located miles from the nearest stand.

Seed Distribution and Germination

As the fruit matures, the receptacle walls redden, often beginning at the distal end. Eventually these split, revealing a reddish pulp in which perhaps several thousand black seed are enmeshed.This seed mass may be devoured on the plant by rodents or birds, or in some areas, collected by Indians for culi- nary purposes (Greene, 1936; Pfefferkorn, 1795). The pulp is often extruded from the receptacles as a unit and falls to the ground, where the seedmay be disseminated by water, insects, rodents, and other animals. In preliminary studies, viable seed have been recovered from the fecal material of several species of animals. Thus, the seed, measuring about 1.5 to 2.0 mm. in length by 1 mm. in width, and weighing approximately 1 mgm., can be distributedover a wide area.The final depository of each seed has a great bearing on the probability of its germination and the survival of thenew seedling.

Extensive experimental work has shown that properly treatedsaguaro seed will not germinate if exposed to darkness or far -red (7350 A) light, but will germinate if exposed to daylight or red (6550 A) light (Alcorn and Kurtz, 1959). Thus to germinate, seed require not only light, but the proper quality of light. This suggests that buried seed cannot germinate. Using quartz sand as a translucent soil simulator, repeated greenhouse tests have shown that germination can occur, but in decreasing amounts, when seed are buried at depths of one inch or less. However, where germination did occur, most seed had to be at a depth of not more than one -eighth of an inch to emerge. With one reservation, then, it can probably be assumed that to germinate, saguaro seed must beon the soil surface, or in crevices or holes exposed to light, or shallowly distributed in very translucent soils. The reservation relates to our lack of knowledge of light requirements and the viability of seed buried for extended periods of time and subjected to the stresses of microbial action, wetting and drying, and temperature fluctuations.

Germination, moreover, cannot be effected unless the seed are properly moistened. Work in progress indicates that seed must be in contact with free water to germinate. Data (Alcorn and Kurtz, 1959) suggest further that the length of time the seed are in contact with water (the imbibition period) in the dark, before being given a single light exposure, may subsequently affect the per cent germination. Such observations agree with those recorded for lettuce seed (Borthwick et al, 1954). 25

Favorable temperatures are a third requirement for seed germination. Under conditions of constant temperature in the laboratory, germination is restricted to the relatively narrow range of 20° to 35° C., with maximum germination near 25° (Alcorn and Kurtz, 1959).No seed have germinated at 15° and few at 35 °. Seed exposed to fluctuating temperatures, one of which is in the 20° to 30° range, may germinate. However, in such tests germination is less than at 25 °.

So far as I know, natural germination in the field has yet to be observed. However, the present studies indicate the times of year when germination might be expected to occur naturally. Additional laboratory investigations have shown that though viability is reduced, the saguaro seed may be stored dry for several years. In addition, dry seed may also survive several days of continuous exposure to such températures as -10° C. and +83 °. Such temperatures are extreme, but they can occur, if only momentarily, at the soil surface. In nature, the seed of the giant cactus probably can remain viable for several years until conditions favorable for germination occur.

Seedling Establishment

A newly germinated seedling has a host of environmental factors to con- tend with, among which are rodents. In one test only, fourteen of eight hundred one -inch, uncaged plants were alive six months after planting.After two years, the survivors and seven hundred seventy of eight hundred additional caged plants were gone.Rodents destroyed all but about one-hundred of the -. 1570 small plants that were lost.The rodent species involved in such depre- dations are not yet known. However, preliminary observations indicate that the packrat (Neotoma albigula) can survive very well on saguaro seedlings and water, but that the same diet will kill kangaroo rats (Dipodomys spectabilis) and D. merriami within several days.Kangaroo mice (Perignathus baileyi and P. penicillatus) are even less inclined than the kangaroo rats to eat the seedlings.

A more insidious attack on very small saguaros comes from the larvae of a species of Gerstaeckeria weevil.This insect feeds within the seedling, leaving the epidermis as an empty shell. In some plots the weevil destroyed 13 per cent of the seedlings.

Though rodents and insects threaten survival of the saguaro seedling, perhaps the physical location of the plant is an equally important factor. Lowe (1959,1960) studied the effects of low temperatures on acclimatized seedlings to determine their cold tolerance as a factor affecting the ecological and geographical distribution of saguaro.He has concluded that the distribution is controlled on its northern and upper elevational limits by extreme low temperatures, as recorded for twenty -four hour periods.Lowe (1959) has also considered moisture stress as a factor limiting the westward distribution of the saguaro.

Results from repeated tests show that shade is required in the desert for survival of half -inch transplanted seedlings. Further, field observations together with greenhouse tests, suggest that seedling survival and growth may be greatly influenced by the type of litter (soil) under the shade -providing "nurse- plant ". Survival and growth data, correlated with chemical analyses of various collections of these litters, suggest that at least pH, salinity, and levels of nitrogen are involved. The relationships of these chemical factors to survival of saguaro seedlings are being studied in nutrient solution cultures. 26

It is a long road, in terms of years, between the seedling and the mature plant (perhaps one hundred twenty -five to one hundred thirty yearsfor a thirty- foot saguaro). So far as is now known, death of a mature saguaro is determined primarily by two factors: bacterial necrosis and the ability of the plant to withstand heavy winds, particularly after soaking rains.

Bacterial Necrosis

Bacterial necrosis is not a new disease. To my knowledge, it was first noted shortly before the turn of the century (Hubbard, 1899). But, the disease is serious, particularly in view of the lack of reproduction of saguaro in scme areas. In long- established field plots in one region, mortality has been almost consistently 2 to 3 per cent per year. The causal organism is a bacillus, Erwinia carnegieana Standring, about two microns long (Lightle et al., 1942). Belonging to the genus Erwinia, it is a close relative of the bacteria that commonly rot carrots, and of those that cause "fire- blight" of pear and apple.

Intensive studies of Boyle (1949) indicate that larvae of the Cactobrosis moth are primarily responsible for carrying the bacilli from saguaros to saguaros. Boyle (1949) isolated the bacilli from the adult moth, the surface of eggs, and both.the surface and the intestinal tracts of larvae. The life cycle of the moth is not perfectly known, but observations indicate that the larvae may spend considerable periods, particularly during the non -bud seasons, tunneling within the saguaro. The circular, corked -over holes seen in profusion on any adult plant are evidence of larval penetration.We do not know whether all larvae carry the bacilli. However, if even a small percentage are actual vectors, the large number of larval invasions per mature plant, would indicate that most, if not all, such plants are infected.

At times the infection sites (larval tunnels) are walled -off by corky tissue produced by the plant, thus localizing the disease. But, for reasons yet unknown, there are times when the bacilli are not contained and the plant is eventually destroyed.

The disease symptoms are actually produced by cellular disintegration (Lightle et al., 1942). If unchecked, the rot can progress throughout the plant, even into the roots. Such disease -weakened roots contribute to the mortality effected by wind. More often, however, break -down and loss of inner parenchymatous tissues leave the familiar gaunt skeleton.

In the process of decay, the epidermis often ruptures, allowing the plant to "bleed ". Frequently copious amounts of liquid, teeming with the infectious bacilli, are deposited on the soil around diseased plants.Boyle (1949) isolated the rot bacilli from the infested soil as long as six weeks after contan- ihatión... Thus, such soil could be a factor in reducing saguaro seedling repopulation and in initiating infections in established plants through injured roots or fresh wounds.

No certain method is known for controlling the disease on a field -wide scale. To control a disease of this nature in the field, one must be able to eliminate or significantly reduce all sources of infection and /or to eliminate or reduce the various agents for disease dispersal - the vectors that could carry the pathogen to "healthy" plants. Work in the Department of Entomology here on the life cycle of the Cactobrosis moth may reveal some portion that will be vulnerable to control procedures. 27

To complicate matters further, recent research indicates that the bacterial necrosis bacillus may also effectively attack a number of other plants, some native to the saguaro habitat. Thus, these other potential infection sites must also be considered in a field program for disease control.

Chemicals, including antibiotics, are known which will control Erwinia carnegieana in vitro (Brown and Boyle, 1944).Some of these chemicals are known to move - to be translocated - within other plants (Alcorn and Ark, 1956; Brian, 1952). However, we do not yet know how to get any of the chemicals into saguaros efficiently on a field scale, whether they will move, or how long they will remain effective within the plant. Further, most of the chemicals are relatively expensive.

On an individual basis, however, the disease can sometimes be eliminated from a valuable ornamental saguaro by aseptically cutting out the occasional, superficial, small rot pockets.

In some regions, saguaro forests are maintaining themselves; some are even increasing in density. However, some stands are being reduced by minimal repopulation on the one hand and by mortality of mature plants from bacterial necrosis on the other. The studies required to understand such population dynamics are still considerable. Work yet to be done can only be accomplished through cooperative efforts of individuals of various interests in various disciplines - in this case, microbiologists, entomologists, botanists, and chemists - each pursuing his own interest, but very much aware of how his work can mesh with and abet that of the others.

Although current research is centered on the saguaro, it is hoped that these intensive studies may contribute to increased understanding of the nature of and interaction between other native biotic complexes present in our arid and semi -arid lands.

Literature Cited

Alcorn, Stanley M. and Peter A. Ark. 1956. Movement of certain antibiotics in cuttings of Pyracantha and Carnation. Appl. Microbiol. 4: 126 -130.

Alcorn, Stanley M. and Edwin B. Kurtz, Jr. 1959. Some factors affecting the germination of seed of the saguaro cactus (Carnegiea gigantea). Am. J. Botany 16: 526 -529.

Alcorn, Stanley M., S. E. McGregor, George D. Butler, Jr., and Edwin B. Kurtz, Jr. 1959. Pollination requirements of the saguaro (Carnegiea gigantea). Cactus and Succ. jour. Amer. 31: 39 -4+1.

Bancroft, Hubert Howe. 1889. The works of Hubert Howe Bancroft. Vol. 17. History of Arizona and New Mexico, 1530 -1888.The History Co., San Francisco. 829 pp.

Benson, Lyman. 1940. The cacti of Arizona. Univ. of Ariz. Biol. Science Bull. #4, Vol. 10 (1). 134 pp.

Borthwick, H. A., S. B. Hendricks, E. H. Toole, and V. K. Toole. 195+. Action of light on lettuce seed germination.Botan. Gaz. 115: 205 -225. 28

Bourke, John Gregory. 1891. On the border with Crook. C. Scribner's Sons, New York. 491 pp.

Boyle, Alice M. 1949. Further studies on the bacterial necrosis of the giant cactus. Phytopathology 39: 1029 -1052.

Brian, P. W. 1952. Antibiotics as systemic fungicides and bactericides. Ann. Appt. Biol. 39: 434 -438.

Britton, N. L. and J. N. Rose. 1920. The cactaceae. Vol. 2, pp. 164 -167. Carnegie Inst. of Washington.

Brown, J. G. and Alice M. Boyle. 1944. Effect of penicillin on a plant pathogen. Phytopathology 34: 760 -761.

Greene, R. E. 1936. The composition and uses of the giant cactus,Carnegiea gigantea. J. Chem. Ed. 13: 309 -312.

Hubbard, H. G. 1899. Insect fauna of the giant cactus of Arizona: Letters from the Southwest. Psyche 8: Suppl. 1, pp. 1 -14.

Kerner von Marilaun, A. 1897. The natural history of plants. F. W. Oliver, trans. 5: 347. Blackie and Son, Ltd., London.

Kurtz, E. B., Jr. 1958. Pollen grain characters of certain cactaceae. Bull. Torrey Bot. Club 75: 516 -522.

Lightle, Paul C., Elizabeth T. Standring and J. G. Brown. 1942. A bacterial necrosis of the giant cactus. Phytopathology 32: 303 -313.

Lowe, Charles H., Jr. 1959. Contemporary biota of the Sonoran Desert: Problems. Univ. of Arizona Arid Lands Colloquia Series.Vol. 1, in press.

. 1960 Second annual report of the University of Arizona Arid Lands Biology Program. In preparation.

MacDougal, D. T. 1905. The saguaro or tree cactus. J. New York Bot. Garden 6: 129 -133.

. 1908. Problems of the desert. Plant World 11: 28 -39.

McGregor, S. E., Stanley M. Alcorn, Edwin B. Kurtz, Jr., and George D. Butler, Jr. 1959. Bee visitors to saguaro flowers. J. Econ. Ent. 52: 1002 -1004.

Peebles, R. H. and Harvey Parker. 1946. Watching the saguaro bloom. Desert Plant Life 18: 55 -60.

Pfefferkorn, Ignanz. 1795. Sonora, a description of the province. T. E. Treutlein, trans. Coronado Historical Series. Vol. 12. Univ. of New Mexico Press. 1949. 329 Pp.

Shreve, F. 1910. The rate of establishment of the giant cactus. Plant World 13: 235 -240.

. 1911. The influence of low temperature on the distribution of the giant cactus. Plant World 14: 136 -146. 29

. 1929. Sahuaro --Its flowers and the way it grows. Desert 1: 10.

. 1951. Vegetation of the Sonoran Desert. Carnegie Inst. of Washington Publ. 591. pp. 139 -141.

Todd, F. E. 1950. Farmers Friend.Ariz. Highways, 26: 4 -9.

Vansell, G. H. and W. H. Griggs. 1952. Honeybees as agents of pollination. The Yearbook of Agr., U. S. D. A., Wash. D. C. pp. 88 -107. 30

PRECIPITATION AND SAGUARO GROWTH

James Rodney Hastings Institute of Atmospheric Physics The University of Arizona

A section in the Arid Lands Colloquia 1959 dealtwith vegetation change in southeastern Arizona, and reviewed some historical evidence suggesting that several major changes accompanied by a cycle of arroyo- cutting began in the years between 1880 and 1890 (Hastings, 1959).

The interest of the Institute of Atmospheric Physics in such occurrences stems from the possibility that climatological trends initiated them. There are, however, many equally plausible explanations. The one advanced perhaps most vigorously, attributes the changes to Anglo- American immigration and to the imbalance that man and his domestic animals wrought on the precarious ecological stability of an arid land.

The Colloquium 1959 by no means provided any sort of answer to the question of primary cause. It pointed rather to the necessity of first investigating subsidiary problems and answering other less complex questions.

Precipitation and saguaro growth, not remote,represent one such line of investigation. The link between these and the parent topic, "vegetation change," lies in work done by Forrest Shreve about 1910, who observed that in two areas he had studied around Tucson, the giant cactus was not repopulating itself.

Although not explicitly stated by Shreve, three assumptions regarding the structure of a living population underlie his work:

1. If a population is to remain stable, i.e., of about the same size and composition year after year, it must be constituted so that there are more individuals of one age than of the age next older. If "number of individuals" is plotted as an ordinate against "age" as an abscissa, their scales respectively increasing and decreasing from the origin, a stable population will be represented by some sort of line from lower left to upper right. Its over -all slope will depend upon mortality characteristics of the species in the particular habitat, and its constancy of slope, i.e., whether the line is straight, concave up, concave down, or has several points of inflection, on the compara- tive mortality among component age groups.

2. If a population is increasing, the slope of the line will be greater than that for a stable population.

3. If a population is decreasing, the slope will be less than that for a stable population. If the slope is zero, or is negative, the population must be declining.

Shreve's raw data consisted of height measurements made of all unbranched saguaros under five meters tall, at each of two locations, one on Tumamoc Hill, the other on "a steep slope at the base of the Santa Catalina Mountains." 31

By the use of repeat photographs, from measurements made by Mrs. E. S. Spalding of the Desert Laboratory on sixteen individuals, and through his own observations on partially irrigated seedlings, Shreve was able to plot the approximate relationship between height and annual apical growth. By summing annual increments of growth, he defined age as a function of height, a relationship given empirically in the following table:

Table 1

Height in meters Age in years

.10 8.o .20 12.5 .4o 19.1 .8o 27.3 1.00 30.3 2.00 40.5 3.00 47.5 4.o0 54.o 5.00 60.0

Applying these figures to his height measurements, Shreve plotted "number of individuals" against "age."Both of his curves showed a negative slope. In the case of the Tumamoc Hill population,the greatest number of individuals was found in the fifty -five to sixty year age group (the oldest measured), the next most in the fifty to fifty -five group, and a steady decline thereafter in each next younger interval. In the Santa Catalina population, the greatest number of individuals was found in the forty to forty -five and forty -five to fifty age groups, with rather abrupt declines proceeding in both directions from that peak.

His study left little doubt that the saguaro was failing to repopulate at either location. Its decline at Tumamoc Hill seemed to date from at least 1850 and possibly earlier; on the Catalina slope from about 1865 -70.

He made the further remarkable statement that "young plants less than 1 dm. in height are so rare, or inconspicuous, that nine botanists who have had excellent opportunities to find them report that they have never done so. . ." (1910, p. 236).

The importance of Shreve's work to anyone concerned with vegetation change is obvious. Of particular interest to the project in the Arid Lands Program were the dates that he assigned to the beginning of saguaro decline, 1850 -1870. A vegetation change during that period would antedate any other documented ones in this area. More important, it would also antedate the main stream of Anglo- American immigration into Arizona and the time when cattle- raising began as a major industry.

In conjunction with Dr. Raymond M. Turner, the first Arid Lands Program investigator to become interested in Shreve's population studies, a series of surveys have been carried out during the past year, aimed at testing the earlier conclusions. 25

20 Ventana Cave 15

r1 n Gl rP7F`ri 40 30 20 10

10 Saguaro N. M, .

40 30 20 10 0

40 30 20 10

15 1 ! Macdougal Crater 10

40 30 20 10 HEIGHT IN FEET 6 ¿? s delimit 95%MEAN confidence GROWTH levels BY HEIGHT CLASS, 1951 - 1959 5 o 0 A A of mean. 0 o A A 0 A o 0 ca A ° A A 0A A A A A o A A A ` \`. `. 0- 12- 99 4- 6- 8- 10- 12- 14- 16- 18- 20- 22- 24- 26- 28- 30- 32- 34- 36- 38-FIGURE 2, SampleMean Annual at Saguaro Apical National Growth ofMonument, Saguaros 1951-1959. of Various Sizes in U. S. D. A. HEIGHT CLASS (feet) 40.99 - MEAN GROWTH BY HEIGHT CLASS , 56 A A A A'S delimit 95% confidence levels 1951 - 1959 of mean. 4 A A A A A A A 0 0 A A -A A A p A 2 A A A p pA p A A A A A A A 1 4 I 2- 4- 10- 12- 14- 16- 1.990- FIGURE 3. 6- 8- DataSmoothed by Shreve. Curve of Mean Annual Apical Growth Based on Figure 2 and HEIGHT CLASS (feet) 18- 20- 22- 24- 26- 28- 30- 32- 34- 36- 38- 40- 40.99 o

No Q) O

o o '{;fti .....'t\':'::?

. '.....'.

o Z óJ

:._...... oo w co o MO

.:.: 8 co

o o JlDN3n03aA 32

Permanent plots have been staked out at four locations across the saguaro range; at two of these, MacDougal Crater and Ventana Cave, the plots have been mapped to facilitate future study. Either directly or by use of instruments, heights have been determined for all saguaros on each plot. The procedure has differed from Shreve's in two respects only. English mensuration was adopted after a short but unequal struggle with the surveying instruments and stadia tables available. Secondly, all individuals were measured, irrespective of their height or the number of arms they had. Distributions for the four locations are shown in Figure 1.

Even before translating height into age, the decadence noted by Shreve appears for the Saguaro National Monument and Tucson Mountain plots. At the former location, there is only one individual less than ten feet tall; the height classes from eighteen to twenty -five feet contain the greater numbers, and a rather abrupt decline occurs with the seventeen -feet class and those shorter. At the latter location a more gradual but still marked decline is evident after a peak occurring at about twenty -three feet.

When comparison is made with the distribution for Ventana Cave, where vigorous repopulation seems to be going on, the decadence at Saguaro National Monument and in the Tucson Mountains becomes even more marked. One is tempted to call the Ventana curve that of an increasing population; that of MacDougal Crater a stable or slightly decreasing one. Until something is known of mortality rates at the two locations, however, such categorizing must be largely speculative.

The problem of constructing a time abscissa for the distributions is, of course, contingent upon knowing mean height -age equivalents. Those formulated by Shreve are unfortunately based on a limited number of cases, and in any event apply only to unbranched specimens under five meters in height.

A much more comprehensive body of data was available in work done by Dr. Stanley Alcorn and his colleagues in the United States Department of Agriculture, incidental to their disease investigations at Saguaro National Monument. These data comprise some 1100 height measurements made over a period from 1951 to 1959. Except for 1955, direct measurements were made each spring during that period on a group of saguaros of diverse heights, 130 -140. Except for deaths, in which cases plants of approximately the same size were substituted, the group has remained constant.

From these basic materials,977 growth increments result, enough so that curves for this particular habitat can be constructed with a reasonable degree of confidence. Figure 2 is the first of such curves and shows mean annual apical growth by height for each interval of two feet.

As can be seen, growth varies widely from one height class to another. In general, it reaches a peak at four to eight feet, declines fairly rapidly at about sixteen to eighteen feet, then remains essentially constant until the interval between twenty -eight and thirty feet is reached. That these differences for various heights are not chance, can be shown by an analysis of variance. The F- ration -variance among the means divided by variance within the means --is three times higher than necessary to be significant at the 99 per cent level.

To say this does not imply, of course, that the mean for each height class is significantly different from that for every other height class. T -tests applied to the standard error of the difference of means, as computed from over -all variance, show that the definitive means on the curve (those 33

for 0 -1.99, 6 -7.99, 16- 17.99, 32- 33.99) do differ significantly from each other at the .99 confidence level.

Using Figure 2 as a basis, and combining with it Shreve's data on growth rates in individuals under two feet in height, a group inadequately represented at Saguaro National Monument, the smoothed curve of Figure 3 results. It is not entirely satisfactory, since it ignores the increase in growth rate that Figure 2 shows to occur at about thirty -two feet. The probability is less than .01 that this increase is mere chance, and can therefore safely be ignored. With this qualification, however, Figure 3 probably represents a good approxi- mation to the actual curve.Table 2, which has been derived by summing the a annual growth increments shown in Figure 3, summarizes what seems to be the most probable height -age relationship among the saguaros on the USDA plots at Saguaro National Monument.

Table 2

Estimated Mean Ages for Saguaros of Various Heights at Saguaro National Monument

Height in feet Height in Meters Age in years

.5 .15 9 1.0 .3o 13 2.0 .61 18 3.0 .91 21 4.0 1.22 24 5.0 1.52 27 6.o 1.83 3o 7.o 2.13 32 8.o 2.44 35 9.o 2 .74 38 10.0 3.05 41 12.0 3.66 48 14.0 4.27 56 16.0 4.88 64 18.o 5.49 74 20.0 6.10 83 25.0 7.62 107 30.0 9.14 131 35o 10.67 157

Using these equivalents, one can now translate the abscissa of Figure 1 from height to age. Figure 4 does so for the Saguaro National Monument population. If the break after eighteen feet is taken to be the time when decadence set in, then the most probable date for it seems to be not 1850 -1870 as thought by Shreve, but the 1880's. This is, of course, the critical decade discussed so much already when mass immigration into southeastern Arizona occurred, when cattle- raising began on a large scale, when arroyo- cutting was inaugurated, and when other changes in vegetation are tentatively supposed to have commenced. If this analysis can be believed, the decline of the saguaro may well be a result of the same factors - climate, man, over- grazing, fire - whatever they are, responsible for the other occurrences. 34

In any event one distribution is hardly conclusive.The configuration of Figure 4 is clearly not reflected in the curves for Ventana Cave and MacDougal Crater Tsee Figure 1). It may or may not be at the Tucscn Mountain plot, depending on by how much mean growth rates there may vary from those calculated for plants at Saguaro National Monument.

Figure 4 may reflect local environmental happenings that have nothing to do with events happening regionally. Or it may reflect events happening regionally, but only in those localities with a common denominator x.And what this common denominator might be, one can at the moment only speculate.

One aid to speculation exists.Because of the remoteness of the Pinacate Mountains, and because of the rugged terrain of the crater itself, it is unlikely that cattle have ever grazed on the MacDougal Crater plot. Neither is there an obvious imbalance in the height distribution for the location. These two statements look interesting when juxtaposed. A cause- effect relation, however, must be established by evidence, not by mere proximity. That the link in this case may be more apparent than real can perhaps be brought home by reflecting on probable past conditions at Ventana Cave. No reason exists for supposing that area to be in any sense "undisturbed."Yet it exhibits the healthiest repopulation characteristics of any surveyed.

Of the four plots studied so far, then, decadence is clearly present in two. Some evidence - far from conclusive - links this decline tentatively with other major environmental changes taking place in the 1680's. The failure of the saguaro to re- establish itself at two localities is by no means true of two others.

The primary question remains unresolved:whether climatic or biotic stress played the major role in effecting saguaro decadence.Yet, as with so much of research, an investigation into one problem has uncovered information bearing on others.

If Alcorn's yearly growth increments at Saguaro National Monument are distributed not by height class, but by year, and if mean annual growth rates are then calculated, one arrives at the figures presented in Table 3.

Table 3 Mean Annual Apical Growth, Saguaro National Monument, 1951 -59

Year Mean Annual Growth Sample Standard Std. Error in Inches Size Deviation of Mean.

1951 -2 4.04 142 2.84 .24 1958 -9 2.94 141 1.56 .13 1952 -3 2.87 138 1.36 .12 1954 -6* 2.72 135 1.17 .10 1953 -4 2.43 139 1.66 .14 1957 -8 2.29 145 1.22 .10 1956 -7 1.77 137 1.48 .13 All Years 2.73 977 1.82 *No measurement in 1955. One -half the growth between 1954 and 1956 has been assigned to each year, 1954 -5 and 1955 -6. 35

Mean annual growth for the period 1951 -9 is 2.73 inches.The variation from year to the next is surprisingly large - from 1.77 inches between the 1956 and 1957 measurements to 4.04 inches in 1951 -2.An analysis of variance shows that the F -ratio is eight times greater than necessary to be significant at the .99 level. Thus the probability is virtually nil that differences can be accounted for by chance.

Experience with trees indicates that height growth is considerably less sensitive to environmental conditionsthan diameter growth (Kramer and Kozlowski, 1960). The configuration in Table 3 nevertheless suggests the sort of variation that dendrochronology has come to associate with soil moisture conditions and rainfall.

A series of statistical correlations between mean annual growth data on the one hand, and the values for various seasonal and annual climatic variables for the same years at Tucson area stations on the other, yields some interesting results.

The best correlation found so far is between mean annual apical growth and summer precipitation - to be more precise, the total of July and August rains equal to or greater than .20 inches, as recorded at the Campbell Ave- nue Experimental Farm. The scattergram for this correlation is shown in Figure 5.Rho, the coefficient of curvilinear correlation, amounts to .93, a value which is significant at .99 with five degrees of freedom. The standard error of estimate is .41 inches; the regression equation of growth on precipitation is given by:

Yc = .360 + 5.014 log X

Tempting as they may look initially, these statistical indices should be approached with caution.

In the first place, the method of "shopping around" for relationships is, in itself, suspect. Over one hundred correlations were run; in at least one case out of a hundred one would expect to generate a probability level of .99 by chance, where no correlation does, in fact, exist at all.

In the second place, as McDonald has pointed out (1956), the spatial variability of summer precipitation in Arizona is large. There are no rain gages on the plots where saguaro measurements were made; no correlation can be undertaken between actual precipitation there and apical growth.To predicate such a relationship on a correlation of .93 between apical growth at one location and precipitation at another fifteen miles away is specious, because rainfall itself at the two locations might well not be that closely related.

Finally, the correlation is of the product- moment type in which values are squared. The point in the upper right of the scattergram is, by itself, responsible for much of the "correlation."This point happens to be of questionable validity since it represents combined precipitation and combined growth for the two -year period, 1954 -6. (In 1955, it will be remembered, no measurements were made.)No such addition of annual values is valid unless the two variables are linearly correlated; the correlation here is curvilinear.

While these three qualifications limit severely what such a correlation may "prove," they do not limit its usefulness in suggesting a hypothesis worth testing experimentally: that apical growth in the saguaro is a function of summer precipitation. 7

4 a

IC =.360+ 5.014 log X 3 8 P = .93

ca _ .41" 2

0 0 2 4 6 8 10 12 E JULY AND AUGUST PPN. > .20"

FIGURE 5. Relationship between Summer Precipitation and Annual Apical Growth in U. S. D.A. Sample at Saguaro National Monument, 1951 -1959. 36

To test this hypothesis the Institute of Atmospheric Physics has built and is presently maintaining an installation behind Tumamoc Hill on the grounds of the old Carnegie Desert Laboratory. The site lies immediately to the west of Quadrat Fifteen, used by Shreve as a permanent plot for long -range vegetation studies. Its selection was governed by three considerations: (1) it affords one of the few locations where a variety of saguaros of different sizes are conveniently grouped; (2) at least in theory it is a protected area, where trespass is forbidden, and where instruments are reasonably safe; (3) it provides two contrasting types of terrain. Part of the plot lies on a level mesa, part on a rocky, south -facing slope.

Soil throughout the plot appears reasonably homogeneous, and is a heavy, volcanic clay of high field capacity-41 to 4+8 per cent.

The installation consists essentially of two parts. One is photographic in nature and provides intermittent, non- metered data on apical growth.Each of eleven saguaros is photographed periodically from a fixed position, in a fixed film plane, against a fixed aluminum rod calibrated in millimeters.Al- though the installation was not completely standardized until mid -June, photog- raghybegan in April, 1959, and has continued to the present. Six photographs taken from the series on one individual are shown in Figure 6.Background variation in the palo verde due to wind and to foliation has since been elim- inatedby photographing the cactus against a white cardboard. The subsequent use of a telephoto lens has provided a greater degree of clarity.

Considering the saguaro's reputation as a slow grower, the extent to which its apical growth can be seen is perhaps surprising. Hardly less striking are the changes in diameter with whose measurement a second part of the installation is concerned.

Figure 7 shows the main components of the Institute's saguaro dendrograph, an instrument discussed by Dr. A. Richard Kassander at an earlier colloquium ( Kassander, 1960). The sensor consists essentially of a brass band held in place by a coil spring. To one end of the band is attached a wire -wound resistor, to the other a contact arm.As the saguaro expands or contracts, resistance in the circuit changes. The current, whose source is two small mercury cells, is relayed to an Esterline -Angus recording milliammeter. By means of a sampling switch attached to the meter's clock drive, circuits leading to ten cacti are recorded in sequence every ten minutes. Calibration charts permit reduction to units of circumference and diameter.

Changes of 0.1 mm. in circumference (.03 mm. in diameter) can be easily detected by the instrument. While this sensitivity is not particularly great as dendrometers go, it is more than sufficient where variations are as gross as those here.

We had hoped originally to determine with the dendrograph (1) the minimum amount of rainfall capable of stimulating expansion in Carnegiea, (2) the lag time between precipitation and expansion, (3) the relation between this lag time and temperature. The instrument has provided this information. In addition, because of its sensitivity relative to the magnitude of daily variations in the stem of the plant, it has given us the record of a clear diurnal cycle related evidently to transpiration.

A standard United States Weather Bureau recording rain gage and a hygro- thermograph maintained by Dr. Raymond M. Turner complete the installation. Part of the combined growth- expansion -precipitation record for one cactus is 37 shown in Figure 8. The direct dependence of diameter on rainfall is immediately apparent. Some other observations seem warranted as well:

1. The saguaro seems unable to pick up water from rains under about .20 inches. This agrees well with Shreve's observation (1914, 20 -21) that rains under .15 inches are probably wasted insofar as desert plant life is concerned because the water is evaporated before penetrating the soil. Shreve appended to his rule -of -thumb figure the statement that the exact value would no doubt vary with temperature, with evaporability, and with soil conditions. The same qualifications should be made here.

2. The beginning of apical growth in the saguaro coincides generally with the onset of the summer rainy season. How precise this dependence may be remains yet to be determined. In Figure 8, triangles mark the points of actual photographic measurement; lines connecting the triangles are the product of inference only, and do not delineate a continuous height record. Between June 25, 1959, and July 21, 1959, no measurements were made. On June 25, apical growth had not started; by July 21 it had. That the rain of July 16 triggered it, as indicated by the line, seems likely, but is not certain.

3. Apical growth appears to be confined to summer. A rapid leveling off occurs about the end of August, also the end of the summer rainy season. Between the last measurement shown in Figure 8 (September 28, 1959) and the time of this colloquium (March, 1960) no growth has occurred.At the spring end of the time -scale, no growth occurred in 1959 between May 24 and June 25. Furthermore, the growth between June 25 and September 28, 1959 (5.2 ") is already greater than the total annual growth to be expected frcm Figure 3 (4.5 ").

4. At those times after the onset of the rainy season when there are water deficits - they can be easily picked out in Figure 8 by noting when decreases in diameter occur - apical growth appears to taper off. Here again, however, the evidence is inconclusive. Too few photographs were taken; intervals between them were too long.

In general, experimental evidence from the Desert Laboratory installation seems to confirm the statistical suggestion that saguaro growth is dependent on sunnier rainfall. It seems likely that growth commences with the onset of the summer rainy season, that growth is confined to summer, and that its amount is contingent, first, on when the rains begin; secondly on their amount, and thirdly on their spacing.

If these inferences are correct - and until another summer's data are in hand, we cannot be sure that they are - so complete a dependence of apical growth on seasonal precipitation suggests other problems that need looking into. Cne problem concerns the part that the summer monsoon, which declines in importance going westward, may play in determining the western limit of the saguaro range.A second concerns the secular trends that may exist in summer rainfall alone: what part, if any, these secular trends, if any, may have played in the larger and elusive problem of our changing natural vegetation. JUNE 25, 1959 JULY 21, 1959 AUG 3, 1959 AUG 31, 1959 SEPT 28, 1959 AUG 24, 1959 FIGURE 6. Grounds.A Summer'sDecimeter. Apical Growth in One Saguaro on the Desert Laboratory Lines are Superimposed at Intervals Corresponding to On_i FIGURE 7. ComponentsLeft, the Sensor of the in Saguaro Place on Dendrograph a Plant. in Use at the Desert Laboratory Plot, Right, the Recorder and Container. (CM.) CACTUS 4 - 1959 DIA.HT.

36 176

34174 Diameter '

32 172

height '' 30 170

28168

PPN (IN.) -....I 26 1661.00 I

24 164. 80 Rainfall A 22162 .40

2n Ian nn - l' I11L r L. . 20 30 10 20 30 10 20 31 10 20 31 10 20 30 JUNE JULY AUGUST SEPT

FIGURE 8. A Four -Month Record of Rainfall at the Desert Laboratory Station, together with Height and Diameter Changes in One Saguaro. 38

LITERATURE CITED

Hastings, James Rodney. 1959. Vegetation Change and Arroyo - Cutting in Southeastern Arizona. Arid Lands Colloquia, 1958 -1959.

Kassander, A. Richard. 1960. Some Problems in the Recording of Data. Arid Lands Colloquia for 1959 -1960.

Kramer, Paul J. and Kozlowski, Theodore T. 1960. Physiology of Trees. McGraw -Hill Book Co.

McDonald, James E. 1956. Variability of Precipitation in an Arid Region: A Survey of Characteristics for Arizona. Technical Report No. 1, University of Arizona, Institute of Atmospheric Physics.

Shreve, Forrest. 1910. The Rate of Establishment of the Giant Cactus. The Plant World, 10:235 -24+0.

1914. "Rainfall as a Determinant of Soil Moisture."The Plant World, 1:9 -26. 39

CHEMICAL CONSTITUENTS OF THE SAGUARO

James W. Berry and Cornelius Steelink* Department of Chemistry The University of Arizona

The saguaro (Carnegiea gigantea) has attracted the active interest of many scientists in the past fifty years. Despite this interest, no systematic chemical studies of this plant are reported, although a few scattered references have been noted. 1,2,3Therefore, systematic chemical analysis of the saguaro should prove useful.

The saguaro, largest of the cacti growing in the United States, is endemic to the Sonoran Desert. It thrives under conditions of high temperature, low rainfall, and loose rocky soil. Individual plants may attain a height of fifty feet, a weight of six tons, and an age of two hundred years. The ability of the cactus to accumulate and retain water enables it to flower and bloom during periods of prolonged drought. The cortex of the plant is unusually large, permitting a variety of studies not possible with the smaller cortex of other dicotyledons.

A unique feature of the plant is the inner framework of ligniferous ribs, the secondary xylem, which is the main structural member. (Figure 1.)The chemical composition of this woody rib material as related to the composition of typical heartwoods is of interest to organic chemists and taxonomists alike. Another interesting feature of this plant is its response to injury or bacterial infection following injury. A hard callus tissue is formed in concentric layers around the injuredpart. This callus tissue may extend deep into the pulpy cortex, sometimes more than six inches and, like the ribs, it is highly ligniferous. The mechanism of callus formation, as well as the formation of related wound tissues and the mechanism of abscission, has been studied in other plants from an anatomical, physiological, and histochemicalviewpoint.6Y7 Bonner9,10and others have investigated the effect of plant growth hormones on the rate of formation of these tissues. Little is known, however, concerning the chemical composition of these pathological excrescences or their co- occurring chemical precursors.

The saguaro appeared to be ideally suited for a study of callus formation, lignification, and suberization because of its large cortex, its easy mechanical separation into distinct tissues, and its availability.

This preliminary investigation was concerned with a proximate chemical analysis of the ribs and callus tissue (the two ligniferous parts of the plant). In order to compare adequately these two tissues with each other and with other plants, the researchers have determined the concentrations of lignin, poly- saccharides, and extractives. Standard methods of heartwood analysis were used.5a, 5hThe data obtained from saguaro tissue are listed in Table I. To emphasize the relation of cactus wood to economic woods, data showing the composition ranges of some typical heartwoods are presented in Table II.

*Part of this work is reported Journal of Organic Chemistry, Volume 25, 1267 -8, (1960) Epidermis

Cortex (pulp)

Rib (secondary xylem)

Pitti

FIGUREI. Cross section of the saguaro 4o

Results of the above may be summarized as follows:

1. The chemical composition of the ribs is quite similar to that of a number of representative hardwoods.

2. The extractive content of the callus tissue (particularly that portion which is soluble in organic solvents) is significantly greater than that of the rib tissue.

3. The lignin content of the callus tissue is approximately 50 per cent greater than that of the rib tissue, whereas the holocellulose content _is-considerably lower.

The last two results are not entirely unexpected, since it isknown6' 7 that the formation of plant callus tissue is accompanied by a large increase in cell -wall lignin, suberin, resinous substances, and compounds soluble in organic solvents.

Although most callus tissue is found deeply inbedded in the cortex of the saguaro, no detectable concentration of lignin is present in the uninjured portion of the cortex, Since a clearer understanding of the mechanism of callus formation was dependent on the determination of the nature of the callus lignin and the identities of some of the co- occurriiig extractives in the adjacent tissues, studies were carried out in;this area.

A chemical degradation of the lignin fraction of both callus tissue and rib tissue revealed the presence of three aldehydes: vanillin (I), syringaldehyde (II), and p- hydroxy -benzaldehyde (III). The analytical results are tabulated in Table III.These compounds occur in the same relative ratio,

CHO CHO

i . -oCH3 CH30 bH OH I II I /II /III (1.0/0.4/0.3) in the lignin of both tissues.

To assess the significance of these results, one must first briefly

examinç previous investigations of lignin. Freudenberg, 11Erdtman,l2 - Pearl,13 Nord, and others have carried out extensive studies to determine the chemical structure and the mode of formation of lignin.These studies show lignin to be an amorphous polymeric substance, which is always associated with cellulose or polysaccharide in the plant and whose molecular structure varies with plant origin. One generalized structure of this polymer is shown below (IV). The monomeric unit is a substituted phenylpropane (C6- C3)fragment. Nitrobenzene C CH ---

-6-CH3 I ) CH

0 -/ CH -0 OCH \`í 3 \ Iv '--oCH3 +l

oxidation of lignin yields substituted aldehydes and ketones such as I, II, III, and acetophenones derived from these compounds, and thus provides information about the structure of the parent polymer. Conifer lignin usually yields vanillin (I) as the major degradation product, while hardwood lignins are characterized by the presence of vanillin (I) and syringaldehyde (11).11 Gramineae are characterized by lignin containing the p- hydroxybenzaldehyde (III) moiety. Thus, the chemical structure of plant components provides useful taxonomic information ( Erdiman, reference 12, extensive discussion).On the basis of the present results *, the saguaro rib tissue appears to be related chemically to other hardwoods.

The formation of conifer lignin arises from the oxidative polymerization of coniferyl alcohol (7).11The latter is found in significant concentrations in the cambial zone and can be converted enzymatically in vitro to a lignin. The structural relation of coniferyl alcohol to vanillin (Iis obvious; lignin synthesized from coniferyl alcohol yields vanillin on nitrobenzene oxidation. Thus, soluble C6 -C3 compounds in plant tissues are the immediate precursors of

CH=CHCH2OH CH=CHCH2OH CHO s

CH30 OCH3 OCH3 % , OR (R=glucose) OH VI V

lignin and presumably are available for rapid lignification in many parts of the plant.

The structures of such soluble precursors may be deduced from a chemical examination of lignin.The existence of both a vanillyl and a syringyl (II) moiety in the polymer would suggest the presence of soluble substances such as syringin in the tissue (VI). Since all three aldehydes (I, II, III) appear to be copolymerized in saguaro lignin, we analyzed the alcoholic extracts of saguaro rib and callus tissue for the presence of these substances.Thé results are tabulated in Table IV and show striking identity between extractive composition and lignin composition.Percentage -wise, the callus contains a higher concentration of aldehydes than does the rib. It is significant that this is also reflected in the higher lignin content of the callus.

Work in the Department of Chemistry here is continuing with the identification of the major phenolic components of callus and rib tissue, as wellas of some of the waxy constituents. Researchers hope to determine the identities of some of the pigments, alkaloids, and other plant phenolics for the purpose of evaluating biosynthetic patterns in the plant. (Geissman,i6 Hünsel,17

* Actually, most hardwoods have a higher syringaldehyde content relative to vanillin than is reported here. It is quite possible, however, that the true concentration of syringaldehyde may be twice the reported values, since it is known (see reference 15) that the degradation procedure is responsible for more than 50 per cent loss of this aldehyde. 42

Birch,i8 related studies.)Once these components are identified, we plan to study the chemical course of disease resistance in saguaro. For example, in one investigation, saguaro plants will be inoculated with Erwinea carnegiana, a soft -rot bacterium. Sections of the sound plant adjacent to the infected area will be analyzed periodically during the course of callus growth. Paper chromatograms of these extracts will be used to determine the changes in composition and concentration of the significant phenolic compounds. It is anticipated that the results of such studies will help elucidate the mechanism of wound formation and abscission, not only in saguaro but in other plants as well.

In summary, the present work has demonstrated the following:

1. Saguaro rib appears to be related chemically to hardwoods, as con- trasted with softwoods. This conclusion is supported by a proximate analysis of the tissue as well as by an analysis of the lignin fraction.

2. Callus tissue is a result of active lignification in response to injury. Analytical data show a high concentration of lignin as well as a proportionally high concentration of soluble lignin precursors in the callus tissue. The number and identity of all callus tissue precursors have not been determined to date.When more of the phenolic components are known, however, they may serve as valuable tracers in later biosynthetic studies of callus formation.

Bibliography

1. Georg Heyl, Arch. Pharm., 266, 668 -73 (1928)

2. Robert A. Green, J. Chem. Ed., 13, 308 (1936)

3. Robert R. Cruse, Economic Botany, 3 111 -31 (19141 -)

)4. Alice M. Boyle, Phytopathology, 39(12), 1029 -52 (19-9)

5a. R. L. Browning and I. H. Isenberg, "Wood Chemistry ", L. E. Wise and E. C. John, Editors, Reinhold Publishing Company, New York, 1952, Chapter 34+

5b. B. L. Browning, ibid, Chapter 32

6. E. J. Butler and S. G. Jones, "Plant Pathology ", Macmillan Company, London, 19+9

7. E. Gaumann, "Principles of Plant Infection ", Hafner Publishing Company, New York, 1950

8. L. E. Wise, M. Murphy and A. A. D'addicco, Paper Trade J., 122, 2, 35 (19+6)

9. J. Bonner, J. M. Henderson and Mary E. Durrell, Am. J. of Botany, 39, 467 -73 (1952)

10. J. Bonner, "Plant Biochemistry ", Academic Press, New York, 1950 43

11. K. Freudenberg, "Modern Methods of Plant Analysis ", Paech and Tracey, Editors, Volume III, 1955, Springer- Verlag, Berlin

12. H. Erdtman, "Perspective in Organic Chemistry ", A. Todd, Editor, 1956, Interscience Publishers, N.Y.

13. I. A. Pearl, J. Am. Chem. Soc., 76, 6106 (195.)

11+. F. F. Nord, W. J. Schubert and S. N. Acerbo, J. Am. Chem. Soc., 79, 252 (1957)

15. R. I. Morrison, J. Soil Science, 9, 130 -40 (1958)

16. T. A. Geissman and E. C. Jorgensen, Arch. Biochem. Biophysics, 5h-, 72 (1955)

17. R. Hünsel, Archiv. Pharmazie, 61, 619 (1956)

18. A. J. Birch and F. W. Donovan, Australian J. Chem., 6, 360 (1953) TABLE la TISSUE EtherPet. EtherEthyl Benzene EXTRACTIVESb Ethanol H2OCold H2OHot NaOH 1% Volatile Steam Lignine cellulose Hobo- Percent of HolocelluloseAlpha Beta Gamma SAGUAROSAGUARO CALLUS RIB 0.163.16 0.112.20 0.051.02 4.242.85 2.923.41 1.461.68 13.511.7 0.080.08 21.9030.40 68.5653.30 49.3552.95 15.0014.36 36.3132.00 TABLE II d HARDWOODS 0.5 -2.0 1.8 -4.0 1.5 -7.0 14 -21 9 -15 2519 -29-24 6071 -74-78 62 -73 CONIFERCONIFERS 0.4 -5.6 4 -34 1.4 -8.0 1 -33 0.4 -5 5 -41 20 -44 27 -55 66 -75 BARK a. All results in % of oven -dry (105 °C) unextracted samples. b.d.c. TheseTAPPIandSuccessive 1%Standardvalues Na0H. extractionsrepresent T -13m Method. a withrange petroleum from minimum ether, to ether,maximum benzene, percentages 95 per of centten speciesethanol, (see cold ref. H20, 5a). hot H2O Analysis of the Lignin Degradation Products of the Saguaro TABLE III Tissue Lignina Vanillin Syring- Aldehydes in Lignin%of Lignin p- Hydroxy- RatioI:II:III RibCallus 30.4021.90 1.200.66 (I) aldehyde 0.500.30(II) benzaldehyde0.400.20 (III) 1:0.43:0.331:0.44:0.30 Analysis of the Hydrolysate of Extractives from the Saguaro TABLE IV Tissue tivesaExtrac- Vanillin (I) Syring-aldehyde (II) Aldehydes in the Extractives % of Extractives benzaldehydep- Hydroxy- (III) RatioI:II:III CallusRib 10.50 2.80 0.070.13 0.05 0.03 11:0.38:0.23 :0.29 :0.14 a. Results in per cent of oven -dry (105 °C) unextracted wood samples. 0.02 0.01 46

ECONCMIC ALTERNATIVES FOR NATIVE PEOPLES IN ARID LANDS

Robert A. Eackenberg Assistant Director Bureau of Ethnic Research The University of Arizona

It has long been the tendency of social scientists to catalog the subsistence patterns of primitive peoples according to a simple categorical scheme as either hunters and gatherers, pastoralists, or agriculturists. Beyond agricultural beginnings, recent evolutionary schemes have indicated the processes through which simple farming societies have developed into states, and even empires.l

But, as Wittfogel observes, the post -agricultural phases of this developmental scheme are much better documented than the pre- agricultural ones.2 Little is known of the circumstances under which a group assumes the practice of agriculture, since this event is usually obscured by time, and unrepresented by definitive archaeological remains.

The circumstances of at least one primitive group's emergence as a farming society can be described from materials assembled by the author in the American Southwest. To understand this event, the assumption of farming practices will be described against the background of logical economic alternatives available to a primitive people occupying an arid environment with a high agricultural potential. The manner in which selections and combinations of these alternatives have figured in the group's total pattern of environmental adaptation should further illustrate the dangers inherent in the oversimplified approach to primitive subsistence patterns referred to in the first paragraph.

The environmental setting for this study is the central and southwest desert of Arizona, which receives between five and ten inches of rainfall per year, and is divided between river basins and poorly watered intermontane valleys. Soils are fertile, and gravity flow irrigation from three sources has been possible for centuries: the Gila, Santa Cruz, and Sonoita Rivers. The ground water table has been adequate to support heavy pump irrigation for the past twenty -five years.

Beyond its agricultural potential, the region produces an abundance of wild edible plants, and, during the aboriginal period, supported a variety of game, fish and bird life. The nature of plant cover, and the nature and distribution of animals has changed greatly since the beginning of the American occupation following the Civil War.

The aboriginal occupants of this area consisted of two groups, designated by different names but sharing many behavioral traits: the Pima and Papago Indians. The two groups had much in common, including language, physical type, and many items of material culture and technology. The reason for applying different terms of reference to the two groups (a custom persisting since Spanish times) is to some extent environmental.The Pimas have staid largely within the valley of the Gila River, while the Papagos have ranged widely throughout the other river valleys, and across the drier intermontane regions. 47

Together with this environmental difference, ethnographers further separate the tribes in terms of variant subsistence patterns, political organization, and the nature of their village settlements. It will be a sub -thesis of this presen- tation that these differences (1) consist of elements which are interdependent; and (2) proceed from a choice of different economic alternatives.

This proposition, like the major question concerning the circumstances surrounding incipient agriculture, leads to a consideration of the economic alternatives themselves.These may be arranged along a gradient of increasing technological intervention. They are intended to indicate the range of logical possibilities in subsistence typology, extending from pre -agriculture through agriculture.

1. Hunting and Gathering. The presence of ample plant and animal food in the area has already been mentioned.The described subsistence patterns of the primitive Yavapai and Maricopa Indians demonstrate adequately that small scattered groups could make a living from wild food resources entirely.3 This alternative involves a retreat from the demands of agricultural work, but also pro- hibits realization of the rich potential of the environment.

2. Marginal Agriculture. Elementary cultivation places maximum reliance upon those environmental features which, without human interference, are con- ducive to maturing a crop. Such features are seen in the soil fertilization and land preparation by natural floods.Human technological intervention to either control or improve the environment is minimal, consisting primarily of planting, occasional cultivation and harvesting. Even though marginal agriculture improves the environment as a human habitat, man does not secure himself against loss of his advantage through flood disaster, drought or crop failure. Hence, marginal agriculture is still a precarious type of subsistence, made doubly so by the tendency for population to increase during several decades of "good" years. Agricultural failure under these circumstances is doubly dis- astrous.

3. Pre- industrial Agriculture. At this level, primitive man attempts to intervene with nature to secure more reliable productivity, and to increase yields far beyond the potential of the undeveloped forces present in nature. These can only be secured through massive intervention with the "given" features of the environment.This intervention takes the form of reservoir construction, dams, canals, land leveling and terracing, bordering and related operations. Without machines, these activities require the coordinated efforts of massed manpower. The social structure, which is the instrument of coordination, is the key to successful agricultural production at this level.Hence, the emphasis placed by Wittfogel upon the need to understand the importance of the "hydraulic society" as a social instrument of enormous powers in those areas of the world which are still struggling toward industrialization. Since human intervention of this sort tends to remove some of the aspects of risk frcm crop production, many disasters are averted and, other things being equal, population continues to increase and social structure to become complex.4

4. Industrial Agriculture. This alternative involves primary reliance upon non -human energy sources as the instruments for transforming the productive capacity of the environment. While machines are vastly superior transforming agents, they maximize social structure demands above even those of the hydraulic society. The structure of finance, credit, transport, supply, 48

communications, and maintenance needed for a machine technology has no parallel in complexity. Yet, only through the application of such vast structures and massive technical resources is environmental mastery possible.

These categories appear to exhaust the logical possibilities, though the opportunity to combine and create interim stages is conceded. Faced with these alternatives then, how did the Pimas and Papagos choose? The answer may baffle anyone committed to notions of simple technological, cultural, or even economic determinism. For the flexibility of the Indian adjustment to the resources of the environment makes it difficult to assign them to one or another alternative position throughout much of their history. To validate these assertions, let us consider some aspects of Pima and Papago history and ethnography in the light of the economic alternatives.

Pimas and Papagos as Hunters and Gatherers. The important role of wild plants and game animals in the diet of both groups between the 17th century and the end of the 19th is conceded by all writers on either group. The rich resources of the desert region may be appreciated from a consideration of the following partial list:4a

"During the ethnographic present they used such foods as seeds, buds, fruits and joints of various cacti; seeds of the mesquite, ironwood, palo verde, amaranth, salt bush, lambsquarter, mustard, horse bean, and squash; acorns and other wild nuts; screw bean; ...the greens of lambsquarter, salt bush, canaigre, amaranth and pigweed; boxthorn and other berries; roots and bulbs of the sand- root (wild potato), covenas and others; and yucca fruit.

...dear, antelope, mountain sheep and mountain goats, peccary, muskrats, bears, rabbits, quail, dove, mockingbird, wild ducks, geese, bittern, heron, snipe, wild turkey, rats, terrapin, lizards, grasshoppers, moth larvae, locusts, iguanas, snakes, toads, and beaver were used."

Both groups resorted to hunting and gathering, but there is considerable difference in emphasis in the extent to which each depended upon it. The foods listed above were the primary source of subsistence for the Papagos, making up more than 3/4 of their total annual food intake.5 Since the cultivated crops used aboriginally by both groups were harvested in October, stored foods were relied upon primarily in the winter Spring and summer subsistence depended largely upon hunting and gathering.°

The Pima search for wild plants and animals was less intensive than that of the Papago, since they often obtained an early corn crop in June. But wild foods were never dispensed with entirely, making up about 2/5 of the annual Pima diet in the pre- Spanish period.? When cultivated crops were abundant, Pimas tranded the surplus with the Papagos for additional supplies of sahuaro syrup, cholla buds, And dried meat. The riverine people regarded these wild foods as delicacies.8

Only in exceptionally dry years did Pima farm produce fail completely. About every fifth year, however, there was a drought of sufficient intensity to reduce the entire tribe to a diet of mesquite beans and jackrabbits.9 A 49

serious famine was reckoned in terms of a failure of the mesquite beans to mature. Several instances of this were recorded in the latter half of the 19th century. Nothing appears to have interfered with the proliferation of the rabbits.

The basic difference between the two groups, as future sections will confirm, was the greater uncertainty of crop production among the Papagos. Hence, this group was more completely dependent upon foods obtained through hunting and gathering. The inadequacy of cultivated foods, together with the vagaries of their domestic water supply, forced many of them to maintain several residences. These residences have been called "field villages" when they were adjacent to the field locations employed for flood -water farming (see below). They were called 'well villages" when located near sources of year -round domestic water supply in the mountains.9a

Well villages among the Papagos, though smaller, were considered to be more permanent than field villages in aboriginal times. Since the "wells" were often merely seepage springs among the rocks, the mountain locations were often inhabited by only one or two extended families. There were, of course, more well locations than field locations.

Papagos spent part of their time in crop producing activities at field locations, but the major portion of the year was consumed by activities for which the well village served as the base of operations. In moving between fields and wells, the Papagos assumed a semi- migratory residence pattern, which had political implications as well as economic.

The organization of native Papago villages, until the period of intensi- fied Apache raiding, was very loose, and tended to follow extended patrilineal family lines. Only in the field locations were supra- family groups associated with each other under aboriginal conditions. The migratory Papago tendency reached its peak of development among the Sand Papagos. This group, which occupied the extreme southwest of the Papagueria, was entirely dependent upon hunting and gathering, and had no field locations.

Summary: Hunting and gathering was the primary subsistence pattern for most Papago villages, and the exclusive source of food for several of them. It was only a supplementary source of diet for Pima villages, and was never relied upon exclusively except in times of drought.

Pimas and Papagos as Marginal Agriculturists.Marginal agriculture, like hunting and gathering, was a pattern known to both groups. In the pre- Spanish period, crops followed the traditional Southwestern maize -bean - squash complex and were shared by both peoples.

The difference between Pima and Papago use of marginal agriculture was, again, largely one of emphasis. Pima fields were permanently laid out, adjacent to either natural or artificial water courses, and permanent houses making up the village settlement were located nearby.

In early Spanish times, the Pimas were grouped in less than a dozen such villages, extending a total distance of 55 miles along the Gila River below the present site of the Casa Grande monument.The regularity of crop production in the Gila bottom lands permitted these villages to be sedentary, though they were spaced at a considerable distance from each other. 50

In size, each Pima village probably consisted of several hundred individuals. Because they were sedentary, and of larger size than the Papago communities, the Pima villages were superior in internal political structure, though there is some question about the precise nature of this structure for the first half of the 18th century.

The Pima villages, like those of the Papago, were organized under a headman, but greater dependence upon agriculture gave the Pima chief more political authority. Direction of warfare and supervision of hunting were the village headman's responsibilities in both groups. In addition, the Pima leader initiated the construction and maintenance of irrigation ditches and diversion dams. He also figured in land clearing and apportior_ment.10These activities ccm- prised the bulk of Pima technical intervention with natural conditions.

Other forms of Pima cooperative labor did not depend upon the leadership of the headman necessarily.Planting and harvesting were organized coopera- tively, though the total manpower of the village was not frequently enlisted for this purpose. Other tasks which were less demanding were discharged by the individual families.

Papago flood farming was more of an individual family matter. Land suitable for cultivation appeared in small fan -like patches at the mouths of washes. The fertility and irrigation of these alluvial fans was secured from the flooding action of the washes themselves.Papago technical intervention was less.than that of the Pimas, consisting largely of placing earthen dikes for water spreading and collection, tank- excavation for water storage, and other minimal conservation measures.

The Papagos, who were more dispersed in their farming activity, were generally without settlements and political and economic organization comparable to that of the Pimas. But again, the matter is one of emphasis. Diversion and canal irrigation using river water was primarily a Pima pattern, and permitted a more intensive agriculture and more complex social structure than the Papago. Still, the Papago corrmunity at San Xavier on the Santa Cruz River had considerable land under cultivation using techniques analogous to those employed by the Pimas. In the valley of the Sonoita River, to the south of the present Sells Reservation, there were several other Papago communities making use of permanent streams also.

Conversely, Pima communities were reported at several places in the Kohatk country on the northern periphery of the Sells Reservation, practicing flood water farming of the Papago variety in the vicinity of Vekol Wash. Russell, the Pima ethnographer, considered these settlements to be Pima.11

Summary: Pimas were predominantly irrigators, settled in sedentary communities along the Gila River, but perhaps not exclusively so.The Papagos were predominantly flood farmers, dependent upon rainfall for crops, but certainly not exclusively so. Some Pimas may have been flood farmers, and scme Papagos surely were irrigators.

Pimas, Papagos and Pre -industrial Agriculture.A division occurs between the two groups at this level. There were tendencies toward the creation of a hydraulic society among the Pima, but not among the Papago.These tendencies 51

could have been a primary step in the sociocultural evolution of an irrigation society, had not other historical circumstances intervened.

The essential event producing the Pima change in the direction of pre- industrial agriculture was the introduction of wheat in the 18th century by the Spanish. The presence of this winter crop permitted them to augment their summer harvest of corn and procure a year- round subsistence from agriculture.12 This act freed the group completely from dependence upon hunting and gathering, but the social and political consequences were even more important.

The expansion in the scope of agriculture permitted the Pimas to adjust to the challenge presented by the Apache wars. The year -round farming cycle permitted a greater population density to accumulate upon a reduced acreage. This, in turn, festered the development of a more tightly integrated social structure.

The range of Pima village settlement contracted frcm 55 to 12 miles. Social integration was manifested in the growth of joint community dependence upon, and management of, a common water supply. Several villages came to rely upon a common canal system, and arranged jointly for maintenance and water allocation through a supra -village structure.All villages, acting under a common chief, established a tribal system of defense including lockouts, sent- ries, and patterns of mobilization to resist sudden surprise attacks.

Some evidence of potential diversification in the Pima economy may be seen. Surpluses of goods were produced and stored, and outlets were sought to trade them. Papagos were hired as laborers in the Pima wheat fields, and there was other evidence of the commencement of social differentiation.

The Papagos received the same stimuli as did the Pimas. They also received wheat frcm the Spanish, and cultivated it as widely as moisture permitted. However, they were unable to produce sufficient quantities to supplant hunting and gathering. Their farming areas could not be consolidated, since the multiple sources of water could not be replaced with a single adequate one.13

Consequently, though the Papagos formed defense villages of a sort to ward off Apache attack, these were used more as fortifications than as residences. There was no tendency toward group -wide consolidation, nor was a common chief appointed for all the Papago villages.

Summary: Wheat permitted the Pimas to assume pre - industrial agriculture, gain superior control over their environment, and actually rise above the level of mere subsistence productivity. By mid -19th century, the Pimas were on the way toward achievement of tribal -wide social integration.The Papagos were unable, because of their deficient environment, to capitalize upon the new opportunity.

Pimas, Papagos and Industrial Agriculture.The appearance of industrial agriculture inevitably accompanied the intrusion of non - Indians into Southern Arizona. Industrialized farm production was preceded by construction of a massive dam and reservoir system on the Salt River, completed in 1905. This was followed in a decade by the introduction of a new cash crop for which there was a waiting marked: cotton. 52

Within the new environmental framework of flood control and managed water supply, non - Indian agriculturists brought vast desert tracts under cultivation. These tracts were located upon three sides of the Pima reservation, but the Indians did not participate as producers in this new economic challenge.None of the Papago lands were within the zone of economic development, so they did not receive a similar opportunity.

The Pima -Papago agricultural institutions, at their peak of development, did not display technological complexity.Tools were the simplest, and machin- ery unknown. Nor did their social structure resemble in any sense the complex network of formal political and economic institutions essential to absorb and apply the procedures of industrial agriculture.

Thus, despite their margin of superiority under aboriginal conditions, the Pimas were no better equipped to participate as producers in the new controlled environment situation than were the Papagos.The gap between their rudimentary irrigation society and the national level of sociocultural integration required to operate and maintain hydraulic works, and to participate effectively in managed irrigation districts, was much too great. It has not yet been closed.

All participants in industrial agriculture need not be land -owning and operating producers, however.The Pima -Papago has found his place in the new system at the level of laborer, if not as owner -operator. At this level the Indian needed few technical or organizational skills he did not already possess.

The Papagos and Pimas of today retain their aboriginal lands, but are unable to put them to economic use. Several attempts have been made through the Bureau of Indian Affairs to introduce both groups to the producers' role within the newly defined farming situation of industrial agriculture. A pump irrigation project for the Papagos at Chuichiu, and a much larger irrigation district on the Pima reservation have yielded unsatisfactory results.

Neither group of Indians is, as yet, capable of coming to terms with a new set of agricultural variables: credit, land preparation, fertilization, equipment rental, seed selection, strict water measurement, systematic irrigation, etc. Only when reservationlands have been leased to non - Indians have they been consistently and profitably productive.A recent, if somewhat uncertain, exception to this statement is the Gila River Pima -Maricopa Community Farm. A tribal enterprise, the Community Farm has successfully grown and marketed several thousand acres of cotton annually since 1953. At this time, however, it is largely a token effort.

Summary: Neither Pimas nor Papagos have succeeded in making the transition to industrial agriculture in their traditional role of producers of farm products. Both groups have become incorporated into the unskilled labor force employed by industrial agricultural producers, however.

Interpretation. The chart opposite provides a summary of the economic alternatives, and the extent of Pima -Papago participation in each alternative. The Papagos are depicted as predominantly hunters and gatherers, but the Pimas are not excluded from this mode of subsistence. The Pimas were predominantly marginal agriculturists throughout most of the aboriginal period, but the Papagos shared this alternative to a limited degree. 53

The Pimas alone made a promising start at pre- industrial agriculture. It began late in their aboriginal history -- -after 1750, and the trend lasted little more than a century. During this period, however, the Pimas developed the following patterns unknown to the Papagos:

1. Concentration of settlement pattern. 2. Cooperative intervillage water management. 3. Tribal -wide political leadership in war and mobilization for defense. 4. Production of surplus farm commodities for sale and trade. 5. Escape frcm the need for wild foods except in poor water years. 6. Commencement of social differentiation seen in employment of laborers and accumulation of wealth.

The Pimas never developed pre- industrial agriculture to the point of secur- ing environmental control, however. The drought of 185+ reduced them to eating mesquite beans, and the flood of 1868 destroyed three of their villages. Proper water conservation measures and flood control construction were never undertaken. After 1870, the social fabric of Pima life began to disintegrate, and there was no opportunity to organize for such an output of group effort.

The relationship between the Pima group, the Papago group, and the economic alternatives has been represented in the opposite chart. Instead of the traditional subsistence typing within the unrealistic limits of a single type, the Papagos have been placed in an intermediate position between the two alternatives which formed the range of their subsistence pursuits: hunting and gathering, and marginal agriculture.

The Pimas have also been placed in an intermediate position between marginal agriculture and pre - industrial agriculture, the alternatives which form the range of their subsistence pursuits. The subsistence patterns of both peoples may thus be seen to form a continuum.The continuum is "pure Papago" at one end, "pure Pima" at the other, but there is an area of overlap in the middle:

HUNTING MARGINAL PRE - INDUST'R IAL GATHERING AGRICULTURE AGRICULTURE

Sand Pima- Pima Papago Papago after 1750

The institutions which have, in the ethnographic literature, served to differentiate Pima from Papago, may be seen as those which the former group developed in consequence of its acceptance of wheat, and development of a year -round '. agricultural cycle. In the absence of these institutions, the difference between the two peoples seems to be one of emphasis, rather than one of kind.

It must now be decided whether this discussion has shed any light on the question of the assumption of the sort of agricultural productivity which leads to the establishment of the hydraulic society.One thing seems clear. Marginal productivity of the aboriginal Papago type, which also characterized Pima cultivation before the introduction of wheat, leads to nothing in the way of social evolution. GatheringHunting and Agriculture Marginal Pre - industrialAgriculture AgricultureIndustrial PINTA producers Pimas as All villages All villages Tendency in ' Tendency in exceptnonesecondarily; exclusively in drought. (1700 -1900) 1870.Declineprimarily. after (1700 -1900) (1775 -1890.)all villages all(1890labor.toward villages -1950) farm PAPAGO producers Papagos as primarily;Most villages a All secondarily;villages inNo anytendency village allTendèncy villages in exclusively.few villages(1700 -1900) a 1875,Increasefew primarily. Down after after 1925(1700 -1900) at any time. labortoward(1915 -1950)farm 55

I. Economic Choice Among Alternatives

A. Hunting and Gathering

NATIVE PAPAGO

B. Marginal Agriculture.

NATIVE PIMA

II. Irrevocable Economic Commitment.

C. Hydraulic Society (pre- industrial).

D. Mcdern Industrial Irrigation Agriculture.

MODERN PIMA - PAPAGO 56

Management of small plots of farm land, by individual families or extended families, whether under desert or riverine conditions, can produce a stable food supply which is marginally adequate. Under these condition, sociocultural stability is possible. There is no impetus to population growth or concentration, no unbearable strain on existing resources.

Among the Pimas two external stimuli destroyed this stability and appear to have started the group on an ascendant evolutionary sequence of changes: the appearance of wheat and the challenge of the Apaches. In the absence of the motive power provided by the severity of the threat of annihila- tion, it is questionable whether the upward spiral would have been initiated. Some writers feel that the Pimas knew about wheat long before they began to cultivate it. The social and political changes occurring as a result of their decision have been listed (c.f., p. 54).

The failure of Pima society to survive the appearance of non - Indian industrial agriculture also deserves comment. There are two apparent reasons for this. First, the creation of a pre- industrial agricultural society appears to be an irrevocable commitment. The assumption of increased agricultural productivity, and growing population density, create an environmental adjustment. This adjustment generally reduces the capacity of the area to produce wild foods and to support people in the absence or failure of their own productive activity. For example, in the vicinity of the consolidated Pima villages, timber soon became scarce, fish disappeared from the stream, and hunting parties ventured increasingly far away from their homes.

Furthermore, the decision to attempt to control water through the creation of diversion dams and canal systems also implies a commitment to cooperative effort. The effort required is usually beyond that of the single family or extended family to mobilize. Thus, should the pre- industrial society fail in its expanded agricultural attempt, retreat to marginal agriculture or to hunting and gathering is extremely difficult. If the society has been long - established and extreme environmental alteration has occurred, it is impossible.

Faced with the non - Indian challenge, then, the Pumas could not retreat to simpler patterns of subsistence.This brings us to the second point.Other non - western societies of irrigation farmers, such as the Burmese and Thai, have made the transition to western agricultural productivity within an industrial framework. This has been achieved without breakdown in the social fabric.

These peoples, however, have had long experience in meeting the requirements of a fully developed hydraulic society infinitely more complex than that begun among the Pimas. In Southeast Asia, for example, the following cultural and structural concepts were present:

1. Devotion of a portion of each year to labor on public works, with or without compensation.

2. Widespread use of writing and record -keeping, with consequent calendri- cal knowledge and precision in performance of timed activities.

3. Extreme social differentiation into managerial and working classes.

4. Authoritarian direction of economic activity by officials representing a remote central government. 57

While not directly comparable, each of these features has a parallel in the demands placed upon the peasant within the framework of industrial agriculture. He already understands wage labor, timing of activities, supervision by remote and powerful social structures, and participation in parts of plans which he does not fully understand.

In the absence of these cultural and structural concepts, Pima Indian society had no chance to absorb and apply the technology and structure of industrial agriculture. The destruction of group life among the Pimas was then foredocmed by (1) their inability to retreat to previous patterns of subsistence activity; (2) the insufficient development of the pre -industrial pattern of agricultural institutions, which, had they matured into a full hydraulic society pattern, might have protected them.

REFERENCES

1. Steward, Julian H., "Cultural Causality and Law: A Trial Formulation of the Development of Early Civilizations," American Anthropologist, Vol. 51, no. 1, 1949.

2. Wittfogel, Karl A., "Developmental Aspects of Hydraulic Societies," in Julian H. Steward, ed., Irrigation Civilizations, Pan American Union, 1955.

3. Spier, Leslie, Yuman Tribes of the Gila River, Chicago, 1933; Gifford, Edward W., The Southeastern Yavapai, University of California Publications in American Archaeology and Ethnology, Vol. 29, no. 3, 1932; The North- eastern and Western Yavapai, UCPAAE, Vol. 34, no. 4, 1936.

4. Wittfogel, Karl A., Oriental Despotism, New Haven, 1957.

4a. Mark, Albyn K., Ecological Change in the History of the Papago Indian Popu- laticn, M.A. Thesis, University of Arizona, 1960, p. 4b.

5. Castetter, Edward F., and Willis H. Bell, Pima and Papago Indian Agricul- ture, Albuquerque, 1942.

6. Ibid.

7. Ibid.

8. Russell, Frank, The Pima Indians, 26th Annual Report, Sureau of American Ethnology, 1908.

9. Ibid.

9a. Underhill, Ruth M., Social Organization of the Papago Indians, New York, 1939.

10. Hill, W.W., "Notes on Pima Land Law and Tenure," American Anthropologist, Vol. 38, 1936, pp. 586 -589.

11. Russell, Frank, op. cit. 12. Castetter, Edward F., and Willis H. Bell, op. cit.

13. Ibid.

14. Ezell, Paul H., Hispanic Acculturation of the Gila River Pima Indians, Ph.D. Thesis, University of Arizona, 1955. 58

CHANGES IN THE PROPERTIES OF WATER DUE TO THE INTERACTION OF SOIL AND WATER1,2

Duwayne M. Anderson Associate Professor Department of Agricultural Chemistry & Soils The University of Arizona

The purpose of this paper is to present in a very brief, yet coherent, manner our present concept of the structure of water sorbed by soils. The topic, inasmuch as it deals with water, is of intrinsic interest. After all, water is the principal substance of which most living things are composed. It is, therefore, absolutely necessary to survival. Profound changes are brought about in water when it exists in thin films on soil particles as compared to its state in bulk. Not all aspects of the subject can be treated in a short paper, and certain simplifications will be made which will be apparent to those well versed in the several scientific disciplines involved in this discussion.

Water is known to chemist as H2O although they all realize that this is an oversimplified representation of its molecular structure. However, it does convey the correct idea that two hydrogen atoms associated with one oxygen atom make up the water molecule. In recent years, it has become known that there are three isotopes of hydrogen.They are named hydrogen, deuterium, and tritium. Chemically the three are hardly distinguishable and differ only in mass. The hydrogen atom, by far the most abundant of the three, is assigned an atomic weight of one, whereas deuterium and tritium have atomic weights of two and three, respectively. There are also three isotopes of oxygen. The most abundant of the isotopes of oxygen has an atomic weight of sixteen, the two others have atomic weights of seventeen and eighteen, respectively. Therefore, it follows from the laws of permutations and combinations that there are no less than eighteen separate and distinct ways of combining these isotopes to form the water molecule H2O. All these are presumed to exist in normal water. It is known that water dissociates into hydrogen and hydroxyl ions (an ion is an atom which possesses one or more electrical charges of either positive or negative sign.)When this is considered, one concludes that there are no fewer than 33 different and distinct entities in water customarily regarded as pure. As if matters were not already complicated enough, evidence has been collected which indicates that single molecules of H2O may be rare and that water is composed of groups of these molecules which may be represented as (H20), (H20)2, (H20), etc., the relative number of each depending upon temperature and pressure

1. Technical Paper, Arizona Agricultural Experiment Station, Tucson, Arizona. Contribution from the Department of Agricultural Chemistry and Soils.

2. Presented to the Arid Lands Colloquia,l2 April 1961. 59

It is helpful at this point to examine a diagram illustrating our present concept of the geometrical arrangement of the atoms making up the single molecule H20, such a diagram is shown in figure 1. The two hydrogen atoms may be thought of as being imbedded in the oxygen atoms like raisins in a bun so that the angle formed by a line connecting their centers to the center of the oxygen atoms is about 105 °. The diameter of the water molecule is about 2.6A° units (2.6 x 10-8cm.), disregarding the protuberances of the two hydrogens. When these are considered, the effective diameter may be expected to be larger, and experiments indicate it to be about 2.9A °, or roughly 3A °. Since the two hydrogen atoms tend to give their negatively charged electrons to the oxygen atoms and since they are located on one side of the molecule, the electrical properties of the molecule are not symetrical about its center. The hydrogen side of the molecule tends to be positive and the opposite side negative. This separation of charge is responsible for many of the electrical properties of water, notably its high dielectric constant.The extent of this polariza- tion is expressed in terms of the dipole moment of water, and it is a property of great interest and significance to chemists and physicists.

When two or more molecules of H2O approach each other in just the right way, there is a tendency for the molecules to "share" hydrogen atoms and in this way link themselves together to form larger units. This is called hydrogen bonding. Figure 2 shows three possible ways in which this may occur. In the first instance, two hydrogen atoms may be "shared" by a pair of molecules. This results in the polar molecules arranging themselves so that positive and negative regions are juxtaposed in harmony with the well -known law: unlike charges attract, whereas like charges repel. Clearly, adding on molecules in this way would result in a chain being built up. In the second instance, only one hydrogen is "shared ". This type of bonding between molecules permits the greatest variation in geometry as successive molecules are added on. In the third situation, three hydrogen atoms are "shared" between two molecules. This situation, if perpetuated, leads us, as in the first instance, to the formation of a chain.There is no good reason to expect that any one type of association between molecules exists to the exclusion of the rest. There is reason, however, to suppose that all three situations exist together resulting in the possibility of a variety of geometrical forms as associated groups of molecules form. Cne must always bear in mind that on a molecular level, everything is in a constant state of turmoil, atoms and molecules vibrating, rotating, bounding and rebounding about so that the associations we speak of here probably last for very short times before they are broken up and new ones formed.

At this point it is appropriate to discuss what happens to all this frantic motion when the temperature is lowered and the "thermal energy" is removed. But before doing so, it is desirable to make a short degression to introduce the concept of ion hydration in terms of the diagrams of water molecules just presented.

When many substances such as common table salt dissolve in water, they actually split up into charged particles called ions.Table salt splits up in water into one positively charged sodium ion and one negatively charged chloride ion. The former is called a cation since it is positively charged, and the latter is called an anion since it bears a negative charge.Recall that the water molecule is polarized; one part of it is negative and one part ii ii /00 , % ;4:1 * +` 105O

, ti

DIAGRAM OF THE WATER MOLECULE, H2O

FIGURE I I. (I) (2) (3) THREE POSSIBLE MODES OF POLYMERIZATION OF FIGURE 2. H2O 6o of it is positive. Since unlike charges attract each other, we can easily predict how the water molecules might arrange themselves about a positively charged cation. Figure 3 illustrates this. It shows how six water molecules might orient themselves about the cation with their negative ends pointing toward the positively charged cation.The relative sizes of the water molecules and the cation are correct in this illustration, and six is about the maximum number of water molecules that can surround the cation. It is interesting to note that extensive experimental work indicates that for many substances about six water molecules are associated with each cation in accordance with this prediction. For the anion, shown in figure 4, the situation is similar but reversed. Here the positive ends of the water molecules point toward the negative anion. In general, anions are larger than cations and the force of electrical interaction is diminished so that, although more than six water molecules can be accommodated in the hydration shell around the anion, in general, fewerwater molecules are attracted to the anion than is the case with the smaller cations.

Let us now refocus our attention on the rapidly vibrating and rebounding water molecules among which hydrogen bonds are continually being formed, broken, and then reformed.What happens as the temperature is decreased? This is another way of asking what happens when the energy of motion is removed. The answer is the motions become less violent and occur less frequently until a point is reached where the hydrogen bonds are strong enough to resist break- ing and the associated groups grow larger and larger until they form only one large group. When this happens, we say the water has frozen. It now possesses the properties of a solid. Because of the variation possible in molecular association, several crystalographic structures of ice may result.Actually, seven different forms of ice are known although only three have been studied by x -ray diffraction and their structures worked out. At normal temperature, ice I is formed; it is the ice we are accustomed to getting out of the refrig- erator.

In contrast to this, let us ask what happens when the temperature of water is increased.Energy is added to the water when its temperature is increased. This causes the molecular motion to be more violent and the associated groups to become smaller and to exist for shorter periods of time until the associated groups virtually disappear and the molecular motion becomes so large that intermolecular attraction becomes insignificant and the molecules behave nearly independently.When this happens, the water possesses the properties of a gas and we call it steam. Figure 5 illustrates this concept and shows the correspondence of viscosity with the degree of hydrogen bonding.The viscosity, of course, was measured directly. The degree of hydrogen bonding has been estimated from magnetic susceptibility, infra -red spectra, scattering of light, x -ray diffraction, and many other measurements.

Having laid this foundation, let us proceed to examine a few of the properties of water in somewhat greater detail. It is convenient to begin with its density which is a measure of the average distance between molecules.The specific gravity of all materials is referred to water, and water is assigned a density of 1 gram per cubic centimeter at its temperature of greatest density, 4 °C. The density of water depends on temperature and pressure.Raising the pressure tends to increase its density; whereas, above 4 °C, raising the temperature tends to decrease it. Another property of interest is the viscosity of water which 61 may be taken as a measure of its fluidity or its "internal friction ", referring to the difficulty of water molecules sliding past each other.Yet another is its vapor pressure. The vapor pressure of water is measured in essence by confining pure water and determining the tendency for some of the water molecules to leave the liquid phase and exist as a gas.Cnly the molecules with great thermal energy can do this, but as the temperature is increased more and more have the requisite energy and the pressure they exert increases.At any given temperature there is an "equilibrium" vapor pressure which is a constant for pure water. All three of these properties are different in water existing as thin films on soil particles.

There are many other properties of water which have been defined and found to be useful in characterizing water. A few of these are its specific heat capacity, its dipole mcment, its dielectric coefficient, its surface tension, and many others too numerous to mention. It will suffice in this limited space to refer to a 670 page monograph entitled Properties of ordinary water substance by N. E. Dorsey to illustrate the vast amount of work which has been expended in measuring the properties of water.

It now becomes necessary to briefly discuss the character of soil materials before actually examining the consequences of the interaction between water and soil.

Unlike water, which at first glance appeared to be a simple substance and then was found to be so complicated, one contemplates, initially, the tremendous variability in composition, appearance, and properties of soils and is somewhat surprised, but nevertheless pleased, to find a great simplifying similarity among them. Soils are predominantly composed of silicate minerals and the silicates are essentially polymorphs of the oxides of silicon and aluminum. (recall that water is an oxide of hydrogen). The silicon atoms is small in proportion to the oxygen atoms so that the silicates appear as clusters of oxygen atoms with the tiny silicon and aluminum atoms and other incidental metallicions tucked away in the interstices. It is not difficult to imagine that water molecules might form links with the oxygen atoms which make up the silicate minerals in much the same way that they form links among themselves. So that it is distinctly possible for water molecules in close proximity to the surface of soil particles to form hydrogen bonds with the oxygen atoms which make up and form the surface of these particles.

When water freezes, the water molecules link together by means of the hydrogen bonds and form the regular structure of ice. In normal ice the tetrahedrally- bonded molecules arrange themselves in hexagonal rings, and it is this regular pattern repeated throughout the solid that is responsible for the six -fold symmetry of the snowflakes. It is of considerable significance, then, to learn that in the silicate minerals a similar six -sided ring of oxygen atoms is repeated over and over as a part of the structure of most of the silicate minerals. This similarity between the structures of normal ice and the silicates was noticed about twenty years ago by Hendricks and Alexander (2) in this country and H. H. Macey (3) in England. They reasoned that the water next to the surface of a silicate particle might, by hydrogen bonding to this surface, build up and propagate an ice -like structure in the thin films of water which often coat the particles. This idea implies that the water next to soil particles is actually in a rigid state and that its density is less than that of normal water. DIAGRAM OFAHYDRATED CATION

FIGURE 3. q

DIAGRAM OF A HYDRATED ANION

FIGURE 4. (asioduu90)JllISOOSI n

o In in (.; _ O. o

o CIO MDeaIN o

o OM co

I /o /I o I.

MN No ,i. --- o. o a) U o S co 8 Ñ ( %) 9NION08 N3908dAH JO 3321030 62

Another view which seemed equally plausible was that the hydroscopic nature of soil clays, i.e. their avidity for water, inferred strong forces of sorption and that these resulted in very high pressures which tended to compress the sorbed water and increase its density.Moreover, it was known that all soil clays have more or less free ions "stuck" on their surfaces and that these "exchangeable ions" tend to hydrate as shown in the previous diagrams resulting in a closer packing of the water molecules in the adsorbed water layer.

It was generally believed, then, that the adsorptive forces at the soil- water interface alter the structure of the water. But opinions differed as to the nature of the alteration.The most prominant point of controversy concerned whether or not the adsorbed water had the properties of a solid or a liquid. And, on the basis of the theories proposed, a measurement of the density of the adsorbed water seemed to offer a means of helping to settle the controversy. So during the past twenty years at least a dozen full -blown experiments, and many lesser ones, were undertaken to settle this point. Unfortunately, no experiment free of some crucial assumption or difficulty was hit upon. The results of all these experiments were conflicting and contradictory.

The author was at Purdue University beginning his Ph.D. work during a time when this issue was of some interest there and he undertook to devise an experiment which would not suffer from the critical defects of preceding experiments. A method was found, but it required that the study be done on a clay that would shrink and swell as its water content was changed.A suitable clay was found and prepared for the experiment and the measurements were made and reconfirmed. The results were definite and, for the clay, which is typical of many soil clays, quite conclusive.Figure 6 shows the results obtained with a sodium -bentonite. By sodium -bentonite is meant a bentonite clay which was prepared in such a way that all its exchangeable cations consisted of sodium ions. Two things are striking about the results. First, the density of the water very near the surface of the clay appears to have a density about 3 per cent less than the bulk liquid; and as the distance from the surface of the clay increases, the density approaches that of the bulk liquid. Second, the density of the adsorbed water is significantly different at distances of more than 20 molecular diameters from the clay surface.A summary of the results of this experiment is given in figure 7.The effect of the presence of the exchangeable cations on the water density is clearly evident. Notice also that the effect of temperature is quite striking.

Since density measurements are a measure of intermolecular distances and therefore are dependent on the "structure" of the liquid, it is quite reasonable to infer from these results that a definite alteration in the structure of water is brought about by the interaction of the clay and water at the clay surfaces. The question arises, then, what is the nature of this structure and how does it differ from that of the bulk liquid. Unfortunately, x -ray diffraction studies of adsorbed water which have provided the most information on the structure of the bulk liquid cannot easily be made. But, there are some indirect indications.

Two thorough studies of the dielectric constant of adsorbed water on some representative silicates have been reported by Muir (4) and by Palmer Cunliffe and Bough (5). The dielectric constant of pure water in bulk is about 80. It can be regarded as a measure of the ability of the water molecule dipoles to 63 orient themselves in an electric field and, consequently, their freedom of motion. In thin films the dielectric constant is lowered appreciably.Values as low as 10 and 20 for the first few molecular layers were reported.Allow- ing for scme difficulty in interpretation, this may be taken as an indication that the adsorbed water is somewhat less mobile than normal water.

The vapor pressure of adsorbed water is lower at any given temperature than pure water. In fact, it is very much lower for thin films than for thick films; as the films become very thick, their vapor pressure approaches that of the bulk liquid. From the very careful measurements of vapor pressures of water films of different thicknesses made at several temperatures by Mooney, Keenan and Wood (6), it is possible to calculate some of the thermodynamic properties of water in thin films. One property of interest is the entropy of adsorbed water relative to pure water.Entropy is a convenient measure of the randomness of the molecular arrangement of the water. The entropy of water is highest for water vapor, lower for liquid water, and lowest for ice. The entropy difference between ice and liquid water is about 5 entropy units.This value may be used for comparison with the calculated entropy difference between the first molecular layer of adsorbed water which is about 4 entropy units and about 2.5 entropy units for the second layer. Clearly, this is an indication of a more orderly structure in the adsorbed water than exists in the bulk liquid.

Recently, Low (7,8) and Dutt (9) measured the ionic mobility and the diffusion rates of various ions in the adsorbed water films.Their most important finding was that the mobility and diffusion rates were lower in the adsorbed water than in bulk water by about 1/3.They concluded that this was caused by the higher viscosity of the adsorbed water as compared to water in bulk.

It is apparent that gradually our understanding of the properties of water in their films is becoming better. Space does not permit discussion of all the other indirect evidence which has accumulated. Let it suffice to conclude with a very brief statement of our present concept in about as complete a manner as is proper to give at this time.

It appears that for thin films of water on many soil minerals, the water probably exists in a "broken down ice structure" in which there is the tendency, as in the normal liquid, for water to bond itself tetrahedrally on the average to about four other water molecules and also to the oxygen surface of the silicate minerals. The exchangeable ions present probably tend to disrupt the orderly perpetuation of this structure by accumulating their normal hydration envelopes.That water in thin films retains a high degree of fluidity, however, seems to be established by the fact that the exchangeable ions are highly mobile and that other ions and molecules diffuse rapidly throughout the water. Finally, we must conclude that the forces responsible for the alteration of the water next to the mineral surfaces are effective out into the liquid beyond 60 molecular diameters (l0). Other properties of water such as its specific heat capacity, its infra -red spectra, its magnetic susceptibility, and so on are undoubtedly altered also. We know, for example, that its freezing point is reduced and its latent heat of vaporization is increased. Certainly much interesting work along these lines remains to be done. a Calculated 12.5 distance from surface25 37.5 in A. 50 62.5 46 0'97 0 Water content1.0 in 2.0 FIGUREg. 6. per g. 3.0 of Na- bentonite 40 5.0 o o co r. rn m a) o_ ó ó ó oo aad .6 ui .ialDnn pagaospp lo Ansuao 64

REFERENCES

1. Dorsey, N. E. (1940). Properties of ordinary water substance. Publishing Co., N. Y., New York.

2. Hendricks, S. B. and Jefferson, M. E. (1940).Structure of kaolin and talcpyrophyllite hydrates and their bearing on water sorption of clays.Am. Min. 23:863 -875.

3. Macey, H. H. (1942). Clay water relationships and the internal mechanism of drying. Trans. of Brit. Cer. Soc. 41:73 -121.

4. Muir, J. (1954). Dielectric loss in water films adsorbed by some silicate minerals. Trans. Far. Soc. 50:249 -254.

5. Palmer, L. S., Cunliff, A. and Hough, J. M. (1953). Dielectric constants of water films. Nature 170:796.

6. Mooney, R. W., Keenan, A. G. and Wood, L. A. (1952). Adsorption of water vapor by montmonillonite. J. of Am. Chem. Soc. 74 :1367 -1371.

7. Low, P. F. (1958). The apparent mobilities of exchangeable alkalie metal cations in bentonite -water systems. Soil Sci. Soc. Amer Proc. 22:395- 398

8. Low, P. F. (1960). The viscosity of water in clay systems.8th Nat'l Conf. on clays and clay minerals (in press).

9. Dutt, G. R. (1960). Diffusion in clay -water systems. Unpublished Ph.D. thesis, Purdue University, Lafayette, Indiana.

10. Anderson, D. M. and Low, P. F. (1958). The density ofwater adsorbed on lithium -, sodium -, and potassium -bentonite. SoilSci. Soc. Amer. Proc. 22:99 -103. 65

ECONCMIC EVALUATION OF WATER CONSERVATION PRACTICES

Sol Resnick, Hydrologist Institute of Water Utilization The University of Arizona

The Senate Select Committee on National Water Resources recently concluded that by 1980 the number one resource shortage of the nation will be fresh water. The U.S. Geological Survey has estimated that by 1980, the daily requirement for the nation will be 600 billion gallons of water; this will have to be obtained frcm the 4,300 billion gallons of rainfall that falls daily on an average on the United States.

The daily increasing requirements are due to the increasing population, plus the increasing rate at which we are using water in our daily lives. In the arid Southwest, of course, the problem is most serious.While in cities like Tucson, water is thrown around like it were only money, the problem is real to the rancher who has to haul water when his stock pond dries up, or to the irri- gator as he watches his groundwater levels drop consistently.

To provide the required water supplies, development and conservation practices will have to be established.The feasibility of providing water by these development and conservation practices for the various uses needs to be economically evaluated.

ECONOMIC EVALUATION

By noting the value of water in Arizona for various uses in Table 1, one can determine the economic feasibility of the following development and conservation practices.

Table 1 APPROXIMATE VALUE OF WATER IN ARIZONA IN DOLLARS PER ACRE -FOOT

Ave. Max. Lcmestic 100 ? Industry 75 Produces income worth 500 to City Mining 40 Produces minerals worth 3,000 Pulp Produces paper worth 3,300 Livestock 300 Produces beef worth 10,000 Irrigation Wheat 10 S.S. Cotton & Veg. 20 L.S. Cotton - 22 Citrus 25 Recreation 1000 /acre of water surface 66

Weather Modification

Weather modification is in the research stage, but indications are that, if the techniques for example for seeding the clouds to increase the total amount of rainfall over an area are perfected, the cost -benefit ratio would be very favorable.

Demineralization of Saline Water

The chief aim of this research program is to reduce the conversion cost frcm the present approximate 350 dollars per acre -foot of treated water to a cost which more nearly compares with the prevailing price of about 100 dollars per acre -foot (30 cents per 1000 gallons) of domestic water at the tap.

An important factor when considering demineralization of saline water is the cost of lifting and transporting the water from sea level to the point of use.

Artificial Groundwater Recharge

One of the most critical problems in the Southwest today is the diminishing groundwater supply. The average yearly overdraft on the groundwater reserves in Arizona alone is approximately three million acre -feet. Artificial recharge takes flood waters that are normally lost by evaporation and transpiration by non -beneficial plants and places it in the underground reservoir. However, any successful program of artificial recharge must include consideration and resolution of many factors, such as methods of desilting, relative chemical quality of recharge water and native groundwater, algae and bacteria control, and air - plugging of aquifer.

Cost of artificial recharge ranges from approximately 40 dollars per acre - foot using injection wells, to almost nothing where spreading is technically feasible. Research is underway in an effort to develop methods of treatment for well injection where costs will total about five dollars per acre -foot of water recharged. A factor when considering artificial recharge is the cost of back pumping when the water is to be utilized.

Evaporation Suppression

Evaporation frcm reservoirs and stock tanks results in large losses of water in the Southwest. The total yearly evaporation loss frcm free water surfaces in the 11 Western States is estimated at 11 million acre -feet; the loss frcm Lake Mead alone averages about 750,000 acre -feet yearly.

Chemicals like hexadacanol, which form monomolecular films, are applied to water surfaces to reduce evaporation. Savings in water of about 65 per cent are being obtained when the chemical is used on water in pans having a diameter of about four feet. Savings of about 25 per cent are being obtained frcm stock tanks up to two acres in surface area.However, any successful program of evaporation suppression must include consideration and resolution of many factors, such as methods for detecting the film on the surface, technique of application, how to prevent microbiological attrition, and effect on the thermal balance of the body of water. 67

Cost of evaporation suppression in field tests to daterange from approximately 60 dollars per acre -foot of water saved at 2500 -acre Lake Hefner, Okla- homa, in trials by the Bureau of Reclamation, to 10 dollars peracre -foot in about 100 -acre lakes in Australia.

Watershed Management

Since only a small part of the water that falls as precipitation in the Southwest is effectively used, research is being conducted in watershedmanage- ment to determine if eradication of non -beneficial plants will increase timber, forage, and water, and, if so, the economics of the operation.Almost one million dollars is being spent in Arizona alone in cooperative studies involving The University of Arizona, Arizona State Land Department, the United States Forest Service, Geological Survey, and the Bureau of Indian Affairs.

The program consists of block and strip cutting of spruce and fir, thinning of ponderosa pine, and eradication of juniper, pinyon, phreatophytes, and non -beneficial chaparral. The removal of vegetation is being accomplished by prescribed burning, and mechanical and chemical means.

Studies indicating increases in water yield of 5 to 10 per cent ofpre- cipitation and increased plant efficiency in use of water have been reported by researchers from the University of California; however, the program is still in the calibration stage in Arizona.

Treatment of Watershed Areas

On an 18 square mile watershed near Tucson, the average run -off for the past three years has been barely three per cent of the precipitation- the rest of the water is almost all lost by evaporation and transpiration by non -beneficial plants. To increase water yields to provide water for domestic livestock and game animals, small areas are being paved with asphalt, cement, concrete, tar- paper, rubber, or plastics. Chemical sprays, which can be applied from air- planes, are being considered for large areas; sufficient chemical to cover an acre would cost only about 30 dollars.

The research program currently in progress at the Arizona Agricultural Experiment Station is investigating the effectiveness, durability, and economics of the various materials used to seal the surface areas.

Increasing Irrigation Efficiencies

The largest single use of water by far in the Southwest is for irrigation; yet, Dr. 0.W. Israelson, one of the most eminent authorities on the subject, estimates that just 25 to 35 per cent of the water diverted for irrigation is ever used by the growing crop. A 10 per cent increase in irrigation efficiencies in Arizona alone would result in a saving of 600,000 acre -feet of water- this is nearly five times the yearly amount used by both Tucson and Phoenix.

To increase irrigation efficiencies, the following operations are required: Better land preparation; better irrigation systems, including lined ditches, etc.; determine when to irrigate and how much water to apply; measure water being applied; breed varieties of plants which are more efficient water users; 68

and attempt to get maximum yields per unit of water rather than per unit of land. The cost -benefit ratio of these operations would most probably be very favorable.

CONCLUSIONS

Water is the limiting factor in human well -being throughout the Southwest. Better utilization of the rainfall, although it is inadequate and uncertain, holds the best present promise technically and economically of meeting to an appreciable extent the tremendous need for water in this area, and, in fact, all the arid areas of the world.

There are a number of agencies, private, State, and Federal, concerned with the problem. A recent count shows that no less than 26 Federal agencies are involved - every time a drop of water falls, they examine it, name it, claim it, dam it, or fight over it. With this interest, plus the interest of the others, more efficient use, with resulting improved cost -benefit ratios for the operations, of the available water resources in the future certainly is assured. 69

STATUS OF CENOZOIC GEOCHRONOLOGY IN THE SOUTHWEST

Terah L. Smiley Director Geochronology Laboratories The University of Arizona

This report is a summary of the current status of research based on funds primarily derived from the National Science Foundation, the Atomic Energy Commission, the Rockefeller Foundation, The Philosophical Society, the State of Arizona, private individuals and anonymous donors. An extremely close cooperative effort has been maintained with the Department of Geology, and to a slightly lesser extent with the Departments of Anthropology, Botany, Chemistry, Meteor- ology and Climatology, and Zoology; and with the Laboratory of Tree -Ring Research and the Institute of Atmospheric Physics.Other academic and research organiza- tions as well as governmental agencies throughout this state and in other states have contributed certain data. The geochemistry portion of this research is under the direct supervision of Paul E. Damon, the paleontological research is under John F. Lance and Paul Wood, and the palynological research is under Jane Gray- Jacobson, Paul S. Martin, and Gerhard O. W. Kremp.

The present discussion is conferred to the small segment of the North American continent we call Arizona. The Cenozoic Era is the last of the four major geologic divisions concerning the Earth's history. The age of the earth has been recently given the figure of approximately 4+.5 x 109 years. There is no information covering the period between the oldest known rocks and the earliest known rocks in Arizona. The oldest rocks so far found and studied in this state are the Precambrian Vishnu schist in the Grand Canyon, the Yavapai schist in central Arizona, and possibly the Pinal schist in the Pinal Mountains. Although our own geochemical dating program has but recently begun operations, through cooperative efforts with Dr. Bruno Gilletti of Oxford University, several dates have been obtained on these rocks which are of the order of 1200 to 1,550 million years.

The history of this area through the remaining part of the Precambrian the Paleozoic and the Mesozoic Eras is interesting. Because this paper is part of the Arid Lands Colloquia Series which is concerned primarily with the present situation, this long time period will be skipped over with the statement that throughout the intervening billion or so years, this land spent some of its time beneath ocean waters and some of it above as part of a continental land mass. The part of history that is of concern to us in this present study begins with the Cenozoic Era which started approximately 63 or so million years ago. So far no sediments or materials have been found in this state which can be definitely assigned to the Paleocene, Eocene, or Oligocene Epochs, the three earliest portions of the Mesozoic which represent a time interval of approximately 35 million years. Why this is so is a problem which causes us some concern - it would appear that Arizona was in a state of amnesia.There are numerous beds in Arizona which have not yet yielded identifying fossils, and it is possible that some of these undesignated beds are of these earlier periods. 70

The State of Arizona lies in two physiographic provinces. The north and east portion is in the Colorado Plateau province and the south and west pro- tion is in the Basin and Range province with the Mogollon Rim as the dividing marker between the two. This paper is mainly concerned with the Basin and Range province and will refer to the Plateau province only to bring in corre- lary information.

The Basin and Range province probably came into existence sometime during Miocene times approximately 23 or so million years ago. There have been since then numerous changes in the mountain systems, the valleys, and the streams. The present land forms have probably existed since Lower Pliocene times with local volcanism, uplift, erosion and sedimentation altering or other- wise modifying the scene.

The upper part of the Cenozoic, then, which is of concern to us, actually begins in the lower Miocene epoch (figure 1) This portion of the geologic column is called provincially the Arikareean and is characterized by Dicerather- ium fossils and associated types. By the potassium -argon technique, this faunal zone is dated from 25 million to 20.2 million years ago.The Atravesada fauna is found in the Mineta beds in the Reddington hills east of Tucson. These beds are comprised of very steeply- dipping limestones which are almost on edge. They were deposited in the period previous to the formation of the Basin and Range province. The small Arikareean rhinoceros of the Atravesada fauna is in sharp contrast of the huge rhinoceros from the Oligocene and early Miocene of Asia. As a matter of interest, a large Asiatic beast, Baluchitherium, had a head about four feet in length and is estimated to have stood almost eighteen feet high at the shoulders. Near the town of Wellton in southwestern Arizona is a locality containing numerous fossils, particularly significant of which are the rodents. The Wellton fauna has been tentatively assigned to the Arikareean faunal zone.

The middle Miocene period is characterized by the presence of Merry- chippus - a small three -toed horse. This small horse, a descendant of Para - hippus, had high- crowned, prismatic teeth. The hiposdont teeth are generally regarded as being associated with grass -feeding habits. The middle Miocene period is termed the Hemingfordian zone.Although no fossils of this particular period have been found in Arizona, the Crestview Wash fauna frcm localities near Needles, California can be tied into formations across the Colorado River in Arizona. It is because of its close proximity and the fact that it will help in dating sediments within the state, that it is included here.

Of particular concern is that, in all probability, the Basin and Range physiographic province, of which southern Arizona is a part, had its formation during Hemingfordian times. Potassium -argon dates from California sediments indicate that the Hemingfordian zone extended from approximately 20.2 million years to 16.5 million years ago. Shortly after the formation of this province, most of the rivers in the area had a southwest trending direction of flow.

The upper Miocene is called the Barstovian zone and extended (again by potassium -argon dates) from roughly 16.5 million to 13 million years ago. The high mountains which had been formed during the early years of the Basin and Range province, underwent major erosion, and many pediments were formed. 71

The Earstovian period is characterized by an animal called the Monosulax which is a small beaver. There are no localities known in Arizona containing fossils of the Barstovian zone.

The lower Pliocene - the last division of the Tertiary - is called the Clarendonian period, and dates from, roughly, 13 million to 9.25 million years ago. The Clarendonian has as a characteristic fossil the Eucastor, another beaver. The Graywater Wash fauna of east - central Arizona, belongs to the Clarendonian zone. In central Arizona, the Prescott fauna is of probable middle or late Clarendonian age. The Prescott fauna contains but few fossils; however these few are so closely allied to those of the Walnut Grove locality, definitely of Clarendonian age, just a few miles away that we believe they are essentially contemporaneous. Pollen grains recovered from matrix in the Prescott faunal zone have been recovered in considerable quantities.The composition of this flora, when compared with contemporary Arizona vegetation, is mainly of the same genera and possibly some of the same species as those which make up the present climax, woodland and chaparral communities in central Arizona. The presence of this late Tertiary vegetation, so similar to that now living in the area, indicates considerable antiquity for the vegetation. Because of this, it is presumed that the climate of that time was similar to what it is today; and that the composition of present vegetation had its beginning as early as Clarendonian times.

The middle Pliocene period is called the Hemphillian and dates, roughly, from 9.2 to 3.2 million years ago. During this period the intermontane basins and valleys were being filled with sediments carried in by streams from higher mountain ranges. Numerous fossil localities of Hemphillian age have been found within the state. One of these, White Cone fauna locality of Central Arizona contains remains of Dipoides, a small beaver which is characteristic of the Hemphillian zone. The Mammoth and Bingam Ranch faunal localities of the lower San Pedro River area contain remains of Hemphillian fossils as do those of the Pumpkin Center locality in the Tonto Basin. Teeth of Pliohippus characterize the Pumpkin Center locality.Pollen grains, which have been extracted from the matrix in this locality, are similar to those found in higher mountain ranges in the nearby areas today.

The White Cone fauna locality near White Cone on the Hopi Indian reservation in north- central Arizona, is the only locality in the upper Cenozoic which can be tied directly to specific geologic formations. This particular formation is called the Bidahochi and is comprised primarily of lacustrine deposits. The remaining fossil localities in the upper Cenozoic are found in fine -grain sediments often grouped under a catch -all phrase called the "Gila Conglomerate" or the "Gila Series ". The Yepomera fauna, similar to the Crestview Wash fauna, is not in the state of Arizona, but in territory adjacent in northern Mexico. There are many affinities between this and localities within the state consequently it is included here. From what can be determined of the climate at this time, it was probably warm and moist.

The period spanning the time boundary between the Pliocene and Pleisto- cene is lumped into a single zone called the Blancan which is characterized by the presence of the Pleisippus horse. It dates from 3.2 million years ago to a period equal to the Kansan glacial stage. Three major localities in the state belong to the Blancan period, the oldest of which is probably the Comosi Moist CC>- ','8000 BC R E C E N T (HomoPORTALIAN sapian) ModernExtinction fauna of many Modern flora Short periods of sedimentation in valleys erosion and Dry < - - - Lu Lu WISCONSINGLACIAL Sangamon (Equus)RANCHOLABREAN Pleistocene types Rampart Cave flora- IntermittentErosionPresent oftopography Pleistocene lakes sediments - - - Cool-Humid - moistwarm - U LtJzCr V ILLINOIAN GLACIAL Yarmouth Northwest trending rivers H W o, I- KANSAN GLACIAL (Equus)IRVINGTONIAN Tusker fauna End of basin -fill deposition Warm - moist - Or d 4 ,y NEBRASCAN GLACIAL Aftonian BLANCAN Flat Tire fauna SaffordModern flora genera of plants Extensive volcanismlakes in valleys Cool- moist I.emy, (Plesippus) Benson enson fauna Regional tilting - - - >- - - Il.smy 3.5my ------ComosiYepomera fauna fauna( ?) - Warm - moist N _ - HEMPHILLIAN WhitePumpkin Cone Centerfauna fauna (7) -- Pumpkin Center flora( ?) Basin filling -- c 9.Imy` - P L I O C E N E 1--- (Dipoides) WalnutMammothBingham Grove fauna fauna ( ?) Ranch fauna (7) lo.6myIz.omr11.ómII.lmy- - (EucastorCLARENDONIAN - - -- GraywaterPrescott Wash fauna( fauna ?) Prescott flora(----- ?) ------Origin of present mountainsand valleys Y - -Similar --_ to present -H (Monosoulax)BARSTOVIAN r Major erosion of mountains, w 17.3my - Southwest trending rivers valleys and pediments 11s2m- - M I O C E N E (Merrychippus)HEMINGFORDIAN Crestview Wash fauna ( ?) BASINPROBABLE AND RANGEORIGIN PROVINCEOF Wellton fauna ( ?) Hw I23.1my- 2I.6my L -- (Diceratherium)ARIKAREEAN Atravesada fauna ------25.7my I O L I GOCENE C E N E WHITNEYAN(Protoceras) 72 locality from south - central Arizona not far from the Mexican border. Camel and horse are the two main types found in the Comosi assemblage.A new species of Tanupalama has been found here and Dr. Lance prefers to call this the Comosillama. The Benson fauna, found in a locality near Benson, Arizona, is a classic one for the Blancan period and contains a large assemblage of animals. In the Whitlock Hills, near Safford, Arizona, are several localities containing a rich assemblage of fossils.The stratigraphically lower of these is called the "Flat Tire" fauna and contains bones of horses, masta- dons, ground sloths, rodents, and turtles. In the same general area of Saf- ford, deep -well cores have been taken, and from these have been extracted numerous pollen grains which have allowed us to reconstruct a fairly good picture of the plant life during the time earlier than the "Flat Tire" fauna, or roughly speaking, perhaps during the Nebraskan glacial stage. The flora in the cores indicate that the tree -line was approximately 1500 to 2000 feet lower than it is today. This could be interpreted as being a cooler and more moist climate than at present.

During the Blancan time there was, early in the period, considerable regional tilting of the lad masses. This was followed by extensive volcanism, and thick deposits of tufa were laid down in this area. The late Blan- can was also a time of extensive lakes and flood plains in the valleys.

The middle Pleistocene, from the Kansan to the Illinoian glacial stage, is characterized by the period called the Irvingtonian which contains as its most common fossil the Equus - the horse. The only major locality in the state belonging to the Irvingtonian, is the one near Safford which is the upper part of the series in which is located the "Flat Tire" fauna and is termed the Tusker zone.The Tusker contains numerous evidences of horse and camel as well as the capabera. A large number of fossil rodents have been removed frcm this locality and these are being studied at the present time.At least one new species of rabbit will come our of the study. The Tusker fauna indicates a fairly warm, moist condition prevailing at that time.

The upper Pleistocene is characterized by the period called the Rancho - labrean which has as its most common fossil the Equus, the same as the Irving - tonian. This particular period marks the extinction of many of the Pleisto- cene types - such as horse, camel, mastodon, elephant, ground sloth, tapir, and others.

The flora collection from Rampart Cave along the Colorado River in north- western Arizona is the only one known for this particular period. An ecological study made on the Rampart Cave deposits of sloth dung, has given a good picture of a cool area gradually beccming warmer and drier. Radiocarbon dates on the Rampart Cave dung range from greater than 35,000 years at the botton to approximately 10,000 years ago at the top of the deposit.

Early in Rancholabrean times river flow was either changed or was changing to a northwest trending direction.Lakes which had been fairly extensive before, began to dry and became highly saline.At the end of the Ranchol- abrean period, many of these sediments in the higher and middle reaches of the streams throughout southern Arizona were completely eroded and carried downstream to be deposited in large intermontane basins. 73

Approximately 10,000 years ago - a worldwide change in climate occurred in the northern hemisphere. Evidences for this change have been found in the continental areas of North America, Europe, Asia, and Africa, as well as in a number of deep sea cores removed from ocean bottcm sediments. This particular period generally marks the end of the Pleistocene and the beginning of the Recent Epoch. No particular faunal zone has been assigned to the Recent; consequently I have taken the liberty to add one and have termed it the Portalian because it was near Portal, Arizona that the first real evidences of the characteristic fossil of this period, "homo sapiens ", were found. A considerable number of soil profiles and cores have been studied which penetrate the Recent Epoch and from these pollen grains have been extracted. We know, for example, that the modern flora had its origin at least 10,000 years ago - perhaps more. Since then, while we have had no change in the species or genera, there have been numerous changes in percentages of the various species depending upon local climatic conditions.

In regard to the geologic phenomenon during the Recent period, we have had numerous short periods of erosion and sedimentation in the higher mountain reaches, in the valleys, and in the lowlands.

Many archaeological sites, which date through the period of time marking the beginning of the Recent, i.e.,8to 12,000 years ago, are located throughout the general area. Several of these sites, such as the ones at Naco, Lehner Ranch, Ventana Cave, and others, have given us a good picture of the animal life as well as of the human occupation during those periods of time.

Such, then, is the current status of Cenozoic Geochronology in the South- west, specifically in Arizona. Current studies using emission spectrography, the geochemical facies concept, potassium -argon and radio - carbon dating techniques, paleontology, palynology, and paleoclimatology will undoubtedly modify and perhaps even change the picture which has been presented.We are beginning studies on caliche formation and on the aging of underground water, on the pollen sequence found in areas to the west and north, and on the sediments and contained fossil material of the playa lake deposits. These studies, plus the continuation of those whose results I have already described, will add considerable data to our present knowledge.