7.3L 5’ .3, la; .2 \ m V1.5 use sows REL. TIC’NSIEIPS arm" mm V33 A33 SELECTED NATVQAL LAUD TYPGS IV THE L- IAFSIVG REGION OF MICHIGAN Henry Wilford Fairchild A THBQIS Submitted to the School of Graduate Studies of ”ichigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PKIECSCPEY Department.of Soil Science 1950 AC}??? 0".7fli-EDG' Til'lTS Gratification is expressed to all the many people who gave time and information for this study. Appreciation is expressed to the members of the Conservation Institute, the Soil Science Department, the Agricultural Economics Department, the State Conservation Department, the State Highway Department, the county Production Varketing Admin- istration chairmen, the Soil Conservation Service, various county officials and farmers who assisted. The author wishes particularly to express his indebt- edness to Dr. C. 3. "illar, Professor Ledoy Schoenmann, Dr. Raleigh Earlowe, 3r. Lawrence Jitt, Dr. L. A. welfanger and Dr. L. i. Turk, whose broad vision and kindly assistance helped to initiate and carry out this study. Lastly, the author desires to express particular appre— ciation to the man whose thinking, teaching and criticism helped the author understand the nature of the ideas ex- pressed in this thesis, Professor J. O. Veatch. *********** ***#***** ****t** ***** *** 2.: r)? .«rw. q '7'. f~\-,~.. 0px 1 ' .QgflJ TABLE OF CCNTBHTS PAlfl‘ I: :\:TRCDTTC:FOE{YOOI0.00.00.00.00...OOOOCOOOCCOOOOOOOOOOOOO GENERAL INTRODUCTION The Problem.......................................... Purpose and Scope of Study........................... “STHCDS OF STUDY The Land Classification System....................... Statistical Vethodology.............................. Theoretical "ethodology.............................. THE AR3A FEDS? STWDY General Description of the Area...................... Physical Features.................................... Surface Geolo:y................................. Fatural Vegetation.............................. _ Soils........................................... History of the Region“............................... Early Settlement................................ Agricultural Development........................ Population Trends............................... Descriptions of Selected Yatural Land Types.......... Surface Configuration Profiles.................. Roxand Type..................................... Riley Type...................................... DiPlain Type.................................... Stockbridge Type................................ Leslie Type.............. ...................... PE‘XRT II: SrnfLTl-SFICAIJ SSCTICI‘TOQooooooooooooooo.ooooooo000000000 PiESSVT ACRICVLF"1AL LAYD T”-133 Selection of Sample Farms............................ Bias of Production Varketing Administration Data..... TTethods of Analysis of Production "arketing Data..... Total Land Use in the Agricultural Area.............. Farm Land Use........................................ Size of Farn......................................... Farming Patterns..................................... Tillable Land "se.................................... Effect of Farm Size on Tillable Land TTse............. IKTB“SITY OF LASD USE General Discussion................................... Intensity By Land fype............................... "ethods T*sed in Determining Intensity from the Census The TYse of Census as a ”easure of Land T'se by Land TypBOCOCOOOOOOOCOOOOOOOOOOOOOIOOOO0000.......IOCOOOOC' Page 1 03019 10 10 ll 12 12 12 14 15 18 18 22 27 28 32 35 41 41 42 42 43 52 55 66 77 El 61 64 85 I le‘ll‘ I‘I‘III .l‘lllllll'l Land Valuation As a "easure of Intensity............... 88 Land Productivity by Land Type......................... 89 Input-Output Relationships by Land Types............... 91 Channes in Land Use from Year to Year.................. 91 Summary of Statistical Work, Part II................... 94 ART III: A UISCPSSICN OF THE CAPSBS C? DIPFERJ" ES I“ IA". ”83 PA ’- TE} IT'S Introduction.......................................... 96 Factors Affecting Farm Size with a Given Vet Return for Labor and "anagement Per Acre..................... 98 General Discussion of Factors Affecting Intensity..... 110 The Effect of Distance from "arket on Farming Systems. 132 Factors Affecting Enterprises that will be Conducted on a Farm............................................. 140 Effect of Eatural Land Type Patterns on Land Vse...... 145 The Quality of Entrepreneur In Relation to Intensity of Land Pse........................................... lcl Land Vse In a Region With “ore Than One Natural Land Type.................................................. 152 The Dynamics of Land Use.............................. 155 Theory Summary........................................ 157 Relation of Theoretical work to the Statistical....... 158 Suggestions for Further Researcu...................... 160 Si‘.3CLED BIELIC“GTJ§bP—~:YOOCOOCOIOOOOOOOOOOO0.0IOOOOOOOOOOOOOOOOOOOUC 163 IV. V. III 0 VII. VIII. IX. ‘7 as. XI. XIII. XIV. XV. XVI. :LIII o IVIII. XIX. XX. XXI. XXII. XXIII. XXIV. XXV. XXVI. LIST OF TABLSS POPULATION GROWTH FOR LANSING AND TSP LANSING P3GICN PROT 1940-1950.......................................... POPULATION DISTNIDNTION BY POPNIATION C ass IN TN3 LANSING NDOION IN 1940.................................. PLAC‘“"AC3 DISTRIBUTICF OP RURAL NON NARI, NON- VILLACP POPNLATION NITN DISTANCE FRO” LANSING................... PSI ON LAND IN P33 CENT ON A 2560 Acaza SA PLzD rtO' EACH ON PITT SPLSOTED IATWRAL LAND TYPDS (DSTIIAT D 1 ANTNO ).. N33 OF PANN LAND BY NATNPAL LAND TYPD..................... SIZE OP PAD" PNIDNNNNN.LAID TYPE......“HH............... PER CTNT DIS TRIP NTI N OP PART 8126 BY IATNPAL LAND TYPE... POPCPNTAOS OP :AP'S TIPOATINO NATIONS CPOPS............... ACRES P333 PAR? REP STING v SIONS CTOPS.. ... P“C‘VTAC” OP TIL 49L“) AND IN VARIOUS CROPS o.............. P3 R CENT OF TILLAPLT LAND BY C{OP CPONPINOS ONT TTFTTD NATNPAL LAND TYPDS...................................... TILLADLD LAID NS IN DP CTNT FOR DTTTTT‘IT-STV‘D PAI'S ON RILE Y LAND TIPS IN INGHA‘ CONNTY..................... TIILADLD LAID N93 BY PA?.{N ”I ON TV1‘3 NATNI‘ AL LAND TYPfi LA D N73 IT? ANSITY "LASNPDS P T. POND NATN‘LIL LAND TYP S... CO‘PAIISC: OP SSLDCTDD TONNerP LAND NSS DATA NRC 1944 CTN NS OP AGRICULT” 3 AND 1949 P A LAND NS IDATA........ A CONPAIISON OP P A A:D CSNSNS DATA BY USING A TNIPD "TASNID, AOPIONLTNPAL DOONO ICS DATAP ION ‘ATN‘ACCONNT- INO FA1:“S............................................... SOIL PRCDVCTIVITY ON PONP SDLSOTSD NATNBAL LAND TYPSS TRON T3333 DATA SONRCTS...................................... CHANGTD S IN P3 CONT OI TITLLRLC LAND Dz3ch3» D TO IIIIS CROPS PRO” 1950 TO 1940 ON R PTISINTATTVT PAN 8 FOR FONT SD- LSCT 3D NATNRAL LAFD TYPIS............................... ALL PCSSIBLE CAOP SSQNDNCTS PITT TNPECTB D YIEL)S ON A NIPOTNBTICAL PAPN.......... PROFIT CONTINATIONS TITS NATIONS CROPPING SYSTTNS ON A NYPOTNDTICAL FAR"....................................... NET PROFIT PPS ACPD SXPTCTTD FRO"'APPLICATICNS ON A VARI- ABLE INPNT (EERTIIIZSP) ONA POP TDONTNOT OP SEDAN-33D CLOVDR SAND ON A P"PCT73TICAT'“AI ...................... TNT TNPECT OF APPLICATION ON A VARIABLE INPNT (PDPTILIZDP) ON THE YITLD OP SNOAT BNPTS, DNY BEANS AND PABLSY ON A NYPOTPDTIOAL NATNP AL LAND TYPE.......................... FACTOR COSTS AT TNT IAPr AT NATIONS DISTANCIST IN ' TN: CITY.................................................... COSTS OF PRODWCTIOE IN DOLLARS AT Z310 DISTAYCE FRO" THE CITE-VCCOOOCCOOOOOOO00......00......00.. COV"ODITY PRIC 33 AT T“? 1A1” AT VPJLIOVS DIDTANCE PROV THE CITY 30? SVCAR BEETS, BEATS AND BAYLSY.............. CO'PARATIVS 23D CCS TS OF “ILK PRODUCTION ON TWO LAYD TYPLLJSOOOOOOOOOOOOOOOOOO0.00.0000...OOOOOOOOOOOOOIOOOO... 17 19 45 47 53 54 67 71_ 75 75 78 BO 82 86 87 90 93 141 143 144 FIGURE 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 24. 25. 26. LIST OF FIGVRCS PAGE The Lansing Region........................................ 9 Natural Land Type Vap..................................... 20 Comparative Surface Configuration Profiles of Five Fatural Land Types.............................................. 23 Photograph Cf The Roxand Natural land Type................ 25 Soil Type and Slope Class flap of a Typical 89 Acre Farm for the Roxand Natural Land Type........................ 26 Photograph of the Riley Natural Land Type................. 29 Soil Type and Slope Class ”ap of a Typical 80 Acre Farm for the Riley Natural Land Type......................... 50 Photograph of the DuPlain Natural Land Type............... 33 Soil Type and Slope Class Vap of a Typical 80 Acre Farm for the OuPlain Natural Land Type....................... 34 Photograph of the Stockbridge Fatural Land Type........... 36 Soil Type and Slope Class ”ap of a Typical 86 Acre Farm for the Stockbridge Hatural Land Type................... 37 Photograph of the Leslie Fatural Land Type................ 59 Soil Type and Slope Class ”Sp for a Typical 80 Acre Farm for the Leslie Natural Land Type.......u................ 40 Farm Pattern by fiatural Land Type doxand Type.......................................... 56 Riley Type........................................... 57 DuPlain Type......................................... 58 Stockbridre Type..................................... 59 Leslie Type.......................................... 60 Field Pattern by Natural Land Type Roxand and Riley Land Types.......................... 62 Stockbridge and DuPlain Land Types................... 63 Leslie Land Type..................................... 64 Woodlot Distribution Ry Natural Land Type................. 65 Indifference “ap of a ”an Faced with the Decision to Farm or Work in an Alternate Job.......................... 100 Indifference flap of a.1hn Faced with the Choice of full tine farming or full time alternate occupation.......... 103 Determinants of the ”aximum Size of Farm for one Individual.............................................. 107 ”axinum Farm Size by Age For One Individual............... lll ’aximum Size of Farm Selected by one Individual At Various Ages............................................ 112 The Law of Diminishing Physical Productivity.............. 116 Output or Use Elasticity Curves For Four Theoretical Types of Land........................................... 119 Use Elasticity Curve For Potatoes On Natural Land Type A.. 122 Use Elasticity Curve for Potatoes For Each of Three Natural Land Types In a Region.......................... 124 The Effect of the Ratio of Output to Input Price on the Intensity of Land Vse................................... 127 FIGURE PAGE 27. The Effect of a Change of Input Transportation Cost on the Intensity of Land Use with Distance From ”arket... 128 28. The Effect of a Change In Output Transportation Cost on the Intensity of Land Use............................. 130 29. The Effect of a Change in Both Input and Output Trans- portation Cost On the Intensity of Land Wse........... 131 30. Wargins of Transference for Three Crops With Distance From ”arket in a Hypothetical Situation............... 139 31. Figure Showing a Geographic Pattern of Two Natural Land Types............................................ 146 32. Substitutability Curve of Bay For Grain For a Cow Pro- ducing Optimum 3ilk Supply............................ 148 33. Idealized Land Use Yap In a Region with Four Symmetric- ally Arranged Natural Land Types...................... 154 34. Idealized Land Use Vap In A Region with Four Non- symmetrically Arranged Fatural Land Types............. 15o PART I INTRODUCTORY GENERAL INTRO DITC TI ON The Problem This study of the relationships between selected natural land types and land use in the Lansing Region of Michigan is an attempt to measure the quantitative effects of differences in the natural environ- ment (as measured by natural land types) on resulting land use. It is a well-known ecological principle that plants and animals vany in their kinds and distribution in accordance with natural condi- tions of environment. It is in agreement with this principle that the hypothesis is made in this study that change in environment has some comparable effect on man's use of land. If man is a rational creature that acts in some predictable manner to combine the physical realities of his natural environment with the economic realities of the price system, it should be possible to develop a system.or theory to describe this action. It is a second hypothesis of this study that this can be done. Studies of this kind are not new; Probably the German, Von Thunen (41), was the first to theorize as to how man would use land for agriculture within one given natural environment. Weber (43) ex- panded Von Thunen's ideas to all economic activity of man, but still remained within the theoretical framework of a given environment. Aereboe (1) was the first to speculate on how differences in soil affected man's use of land within a given economic region. Brinkman (6), furthering these ideas, arrived at a scientific dilemma by concluding that although a theory could be developed to explain how a given land use pattern could develop, it would be impossible to use the theory because of the impossibility of characterizing natural environments by some scientific classification system. He apparently gave up at this point. With the development of the natural land type concept, the dilemma that had faced Brinkman seemed to disappear. Here was a scientific system of natural environment classification that seemed to satisfy the needs of explaining differences in land use within a given economic environment. This study proposes to discover if differences in natural environ- ment as measured by this natural land type concept are reflected in differences in man's use of land. If they are, it is the aim of this study to develop a theory that will explain how these differences in land use could have come about. Purpose and Scope 2: Study The original purpose of the study was to test to see if land use varied by natural land type and if it did, to explain how and why particular patterns of land use came about. After almost two years' work gathering statistics with the above purpose in mind, it became evident that statistics simply were not available to show'why a particular pattern of land use had developed on one particular natural land type. In order'to salvage what work had been done, the purpose was changed. In addition to attempting to prove that land use differed between natural land types, it was decided that perhaps some framework of studying land use could be developed that would provide a guide for statistical studies of land use so that many of the errors that had been made by the author in his quest to find out why certain land uses existed, could be avoided in the future. Economic theory was selected as the medium in which the out- line for'land use studies was to be written. The thesis as it finally evolved consisted of two parts: one, to show that land use differed between natural land types and secondly, a distinct section having little relationship to the first, showing by theoretical methods how a system of study could be designed to study land use in an area. The scope, then, of the investigation includes two categories of study, statistical and theoretical. The statistical section consists of an inquiry into the relation of agricultural land use to selected land types in the Lansing Region of Nfichigan. Specifically, the statistical work includes the co-relationships existing between five natural land types in the Lansing Region and agricultural land use on each of these natural land types. The effect of farm size on land use is observed. Studies of field pattern and farm pattern by natural land types are made. A discussion of yearly changes in land use is included. The theoretical section of the study includes a review of litera- ture pertinent to land use theory. Economic models and concepts are developed deductively in an attempt to explain such land use phenomena as farm size distribution, land use intensity and land use distribution by natural land type in a given static theoretical situation. 1 I (II i ['1 {II ‘III I I I'll! I. l Finally out of the theory section a program or outline for land use pattern studies is evolved which the author believes can be used as a pattern for studying the why and how of land use in local areas. The scope of the investigation, while seemingly disjointed and uncoordinated, provides some insight into problems facing land use planners and other land utilization workers. No claims are made that this study is complete. The investigation is designed to provide a beginning method of study of land utilization in a local area, employ- ing the natural land type idea. METHODS OF STUDY The Land Classification System The land classification system used to characterize the differ- ences in natural environment is known as the natural land type, and, as used in this thesis, was developed some years ago by Veatch (40). The same concept has been used with some alterations by Milne (26), SHE/6' «lid! (29) and others. Veatch's system of classification, like other genetic systems, rests on the principle that the characteristics used for classifica- tion lie inherently within the phenomena being classified. This particular system.of land classification, the natural land type, employs such criteria as soil type, surface configuration, stream pattern and relief, as defining features. The scale, or order, of classification has been arrived at de- ductively. This appears to be the major weakness of the system for utilization studies. Nevertheless, the system is relatively unchanging. This thesis attempts to test the utility of such a system as an aid to land use studies. Intuitively, it would seem that a natural land classification system would be more valuable in agricultural land use studies than in urban studies.y This can be explained by the fact that the more closely related the particular'use under consideration is to the natural land, the greater the effect will be of changes in the land on the use. In this study the land considered therefore is mainly agricultural land. It is felt that if the natural land type concept has validity for use studies, it should certainly indicate it on agricultural land. Statistical Methodology The usual procedure of studying land use in an area is to sample area blocks of from several hundred to several thousand acres each. All uses are determined in each block and presented in both table and map form. This is an excellent but costly procedure. For this study involving parts of ten counties and about ten thousand farms this process of field mapping was out of the question because of a lack of money to do the field work. A procedure of second best approximation had to be developed. First, a method using mailed land use questionnaires was used. After several hundred returns were received, it became evident that this qystem would not succeed because of so many errors and omissions in filling out the questionnaires. Other sampling systems were tried and abandoned. The one finally selected consisted of using Production marketing Administration.data collected in each sample county, supported by census information, data collected from farmsaccounting farms from the Agricultural Economics Department of "ichigan State College and data from a variety of other sources. In addition to collecting samples randomly scattered over each land type, five areas of four sections each were selected to study farm and field pattern on each of the selected natural land types. The aim.of the methods used was not to obtain as precise a picture of land use as could be had but rather to get as precise and reliable a sample as possible with the resources available for doing the job. The entire section on present land use leaves much to be desired from a research standpoint. It is realized that the data could be imp proved. At the same time a fairly accurate picture of land use is pre- Bented. Theoretical Methodology Deductive theory is used as an analogy to reality in an area where the factors involved are too complicated to be grasped and understood easily. Generally not a completely perfect analogy is obtained or no simplicity would be gained by using theory. Therefore only the more important factors of a situation are studied. A perfect theory of land use would be a regression equation in which all the factors affect- ing land use were included in their properly weighted relationships. This would be an impossible task since the techniques for isolating many of the factors have not as yet been developed. The procedure used to develop theoretical models of land use is one in which, if given perfect knowledge about a static situation of 1 I II I II J' I: III-II- lll'l Ill if I Ir prices, production conditions and pattern of natural land types, the rational land use would be determined, with the explicit assumption that entrepreneurs attempt at all times to maximize profits. This is not a real world situation but is hoped to be sufficiently like the real world to be of value in explaining experimental data for use in the real world. THE AREA UNDER STUDY General Description 3: thg.Area The Lansing region is an economic sub-region of Michigan. It comprises an area about 65 miles long from north to south and 55 miles wide from.east to west at the widest part. The region roughly resemp bles the shape of a pear with the stem to the north. The area enclosed is about 2500 square miles and includes parts of ten counties. The extent of the region, after Thaden, is shown on Figure l. The basis for the region has been given by Thaden (34). He con- sidered the region as the zone of influence of the largest city, Lansing, over its surrounding hinterland. The area of the region corresponds closely to the ”milkshed" of Lansing or to the Christmas shopping area. The sub-region is an order of economic and social unit which lies be- tween the smaller town-community and the larger regional community. For example, the Lansing sub-region is one unit of many comparable units in the Detroit zone of influence. In turn, the Lansing sub-region is comprised of many smaller economic and social units known as town- communities. The M’ason community is one such unit. Lansing, the central city and the large city of the region, lies in the northwest corner of Ingham county. The population for 1950 is given.as 91,678 people. The population of the city with its surround- ing fringe area is about 130,000 people. The Lansing region consists of about fifty-five town-communities, each of these communities consists of the village and the farming hinterland. 'Within the farming area are numerous neighborhoods which comprise for the most part social rather than economic units. FIGURE I. THE LANSING REGION ’ I "“ 4" \ I 1‘ / ITHACA \ \ ,GRATHDT co. . w I, i /f ' ‘ \ . i,___-L4__ .—-—v./———_' -—.— —‘_-T-.—-{ \ / l \ l/ g e ' l“ IONIA . L ST. JOHNS owo‘sso I a \ I a l \ l CLINTON cojIL \‘ N i ' ‘ 1.—-——- —-—-+—-~»—-~-----—;:r--'——--.L-L I © ‘ l . LANSIM - 4g ) I . " |' H EkTON 00. ' INGHAN‘ 0+. \ I ! \\ CHARLOTTE : 'M’Asou [ \ / ‘ l L1 3 \\‘~ 1 ‘/ ' ‘ \\‘ I I’. ——.J—_——-J—-—L?Nt—'~di'— TJF=-::-—-‘———~L 50A E- :IOmiles._ 10 Like its satellite communities, the Lansing region is also composed of economic and social units. In this case, townpcommunities are the building blocks. The structure of the region simulates that of the town-community, having at the center the structures associated'with a city. Beyond the city itself, lies the urban-rural fringe area and be- yond this, the farming community hinterland. Almost the entire region lies in the dairy and general farming area of central Michigan. There is a small section in the northeast part which lies in the cash crop region of the lake bed plain. The Lansing Region itself lies within a larger economic region that has the Detroit Metropolitan area as a hub. This influence of Detroit is particularly noticed in both industrial and agricultural pro- duction. Fluid milk is the most common agricultural commodity affected. Physical Features Surface Geology. The Lansing Region lies in the Eastern Lake Section of the Central Lowland Physiographic Province of the United States. The land surface is composed of glacial till, deposited as a level to undulating or moderately rolling glacial plain. Nest of the area is Composed of undulating till plains with some smooth to pitted outwash plains occurring locally: Moraines are a common feature of the landscape, as are eskers, small potholes and low shallow swells. Rela- tively broad glacial drainageways cross the area at intervals. The only truly level land of the entire region occurs on old lake bed plains in the northeastern part of Clinton County and farther to the north and northeast in Gratiot and Shiawassee counties. Locally, there are flat ll muck plains as around Stockbridge and in Chandlers Marsh, northeast of Lansing. The main drainageway of the area is the Grand River. Tributaries of this river drain almost the whole region, but imperfectly. The broad plainlike divides and the shallow drainageways twine in haphazard fashion. Dissection of the upland has not proceeded to any extent except along the banks of the Grand River and there for a very short distance inland. The undissected condition of the upland caused by the youthful stage of drainage has left hundreds of small to sizeable areas of muck and peat. This is particularly true in the south half of the region. The general direction of drainage is northwest. The maximum relief difference in the region is about four hundred feet with most of the land lying a little less than 900 feet above mean sea level. Average relief differences are not greater than fifty to seventy feet on most land types. Natural Vegetation. Originally a dense growth of mostly decid- uous trees covered the entire area. Local areas of oak openings occurred and in Baton County there were a few scattered prairies. Swamp prairies of tall sedges occurred on the broad muck plains in the area. In general there existed a correlation between soil character and ttmber species. The muck plain occupying the wettest position sup- ported marsh sedge and tall grasses. The drier parts produced red maple, willow; tamarack and aspen. Some marshes sustained a leather- leaf, sphagnum, cranberry-type bog. The lowland mineral soils with poor drainage maintained large individual trees of elm, ash and red maple, with butternut, black walnut and cottonwood scattered throughout. 12 The upland soils supported three types of trees. In some of the very sandy sites, as around Pine Lake, pines occurred. On soils of intermediate texture there occurred an oak-hickory association with a predominance of oak; the heavier-textured uplands produced a maple- beech forest with considerable hickory, oak, black cherry, basswood and ash as associated species. §g$l§. The soils of the Lansing Region belong to the gray-brown podsolic great soil group. They are formed mostly from calcareous glacial drift, said by Leverett (23) to have been deposited 15,000 to 20,000 years ago. The drift varies in depth from.nothing at Grand Ledge where Pennsylvanian rocks outcrop to a depth of perhaps a hundred feet. The soils differ widely in structure, texture, color, amount of organic matter, moisture and chemical composition. The most striking feature is the total lack of uniformity of soil in a small area. It is not unusual for one field to contain several soil types. The upland soils vary from light-colored, loose, dry, incoherent sands of single grain structure, low waterbholding capacity and low supply of available nutrients, to heamy, granular dark-colored clays, high in moisture-holding capacity and available nutrients. In associa- tion with these upland soils are organic soils, ranging from loose, granular mucks to woody and felt—like peats. Histosy of the Region Early Settlement. The Lansing region.was settled later than those regions to the south, east or west of it. This was due to the dominantly low wet swamps and dense forbidding forests. 13 The first land in the region.was sold to land speculators a year or two before actual settlement began. In 1834 the first settlements ‘were made in Ingham county at Stockbridge and Onondaga. It is probable that an earlier settlement was made in southwest Livingston county in 1835. By 1836 there were several settlements in the region, including Lansing. By 1840, many villages were competing for the trade of the area. In 1847, the Capitol was moved to Lansing, a city of less than half a dozen houses. The act of the legislature moving the Capitol to "the city in the forest" determined once and for all the center of trade and political life for central Michigan (12). The pattern of early settlement developed from the southeast and the northwest. This was because the two existing roads at the time were the Ann Arbor to Kalamazoo road, to the south, and the Saginaw to Ionia road, to the north. The south road apparently was favored because of the opportunity to ship heavy goods down the Grand River from."Jackson- borough" to Lansing or Ionia. After 1845, when the railroad was finished through Jackson, the southern route was favored even more. Early agriculture centered around farming the lands which could easily be cleared and planted. This in addition to the belief that the low wet lands caused the "ague" (malaria) led settlers to select the ”oak openings", open park-like lands with scattered burr oaks, for the first farms in the region. The usual pattern of farm settlement was to proceed to the place of settlement by ox team, erect a rude leanto of logs and bark or sod and proceed to break land with a "stumpjumper" and a yoke of oxen. After the corn or barley was planted, the men, who quite frequently had 14 left their families behind in New York State or NeW'England, built a rude log cabin and girdled trees for next year's crop before returning east for their families. The principal farm.products grown were the grains, wheat, oats, barley, rye and corn. Hay was found growing wild. The animals grown were dairy cattle, beef cattle, sheep and horses. Contrary to popular opinion, little of the timber of Central Michigan was cut for lumber by the earliest settlers. After the first wave of settlers had farmed the oak openings for a few years, they grad- ually moved back on the uplands where land with such soil types as Miami and Hillsdale was opened up. These soils supported heavy growths of oak and hickory. It is stated by Fuller (12) that traveling timber cut- ters cut and piled this timber on a custom work basis. The first year, this timber was allowed to lie in the fields and the farmer worked around it and the stumps. The second spring, the timber was burned. While it is true that many of the early roads were plank or corduroy roads, this use accounted for very little of the original timber re- smurces of south central Michigant _Agricultural Development. By 1880, seventy-five per cent of the land in farms was cultivated, virtually the same per cent as was tilled in 1945. In 1880, approximately ninety-six per cent of the land was in farms (35), a higher percentage than in 1945. The major changes in agriculture since 1880 have been clearing more of the woods and develop- ing more rotation pasture on already cleared land. Fany of the farms of 1880 have since been abandoned to either idle land because of low pro- ductivity or converted to urban land. 15 Census figures show that there have been feW'changes since 1900 in the crops grown. Alfalfa has been introduced and many improvements have been made in varieties grown. Generally however, the crop and livestock load per farm has changed little. Size of farm has increased constantly. The number of horses has declined greatly to be replaced by machinery. Investment per farm in equipment has increased with a corresponding decrease in the amount of family labor. ngulation Trends. The Lansing region is comprised of about 300,000 people living in parts of ten counties (37). The population growth has been rapid as can be seen in Table I. This table compares the growth of five counties making up the greater part of the Lansing Region, with the city of Lansing. Table II gives the distribution of the population between four popu- lation classes in 1940 (37). It can be seen that at that time, urban population constituted about half of the total population of the region. About 70 per cent of the urban population was concentrated in the city of Lansing. It should also be noted that the farm population in each of the five counties listed was about equal. The rural farm population constituted about one-quarter of the total population. Village popula- tion, like rural population, was nearly the same in each of the five counties studied. One class of population that has some bearing on this study is the class termed rural nonavillage, non-farm. This group is important since the numbers of this group vary directly with distance from Lansing. From the broad picture of land use this means that as distance from.town increases, this group of population diminishes in numbers, hence in the amount of land use for rural residence and part time farm use. TABLE I POPULATION GROWTH FOR LANSING AND THE LANSING REGION FROM'1840-1950* Year Lansing Region Lansing City Population Population 1840 10,517 ---..- 1850 33,619 1,229 1860 76,858 3,074 1870 121,823 5,241 1880 153,932 15,438 1890 160,022 13,102 1900 164,817 16,485 1910 173,734 31,229 1920 203,052 57,327 1930 207,582 78,397 1940 268,328 78,753 1950 327,217 91,678 * U. S. Department of Commerce, Bureau of the Census (37). TABLE II POPULATION DISTRIBUTION BY POPULATION CLASS IN THE LANSING REGION IN 1940 County Population Class Total Urban Village RNFNV* Rural Farm Clinton 26,671 4,422 4,399 3,094 14,803 Eaton 34,124 12,503 4,016 2,744 14,904 Ionia 35,710 10,481 6,636 5,069 13,702 Ingham. 130,616 87,459 4,696 23,792 14,688 Shiawassee 41,207 17,551 5,801 3,016 15,140 Total 268,328 132,416 25,538 37,715 73,237 Per cent of total 100.0 49.3 9.4 13.8 27.1 * Rural non-farm, nonavillage population fl 18 This can best be shown as in Table III which shows the percentage distribution of this class of population with distance from Lansing. Five townships starting with Lansing township, situated west of Lansing 'were selected to study the distribution of the rural nondvillage, non- farm.popu1ation in 1940. The figures in per cent varied from 96 for Lansing township, which surrounds the city of Lansing to 5 per cent for Sunfield township, situated 24 miles west of Lansing. Descriptions of Selected Natural Land Types The natural land types selected from the Lansing Region for this study are shown on Figure 2. This map was taken from a larger natural land type map of the lower peninsula of MichiganA:;'Veatch (39). The Lansing Region comprises all the area on the map lying inside the heavy- dashed line. Twenty natural land types occur within the boundaries of the region. Only five were sampled in this inquiry. These five are numbered after Veatch with the numbers 27, l, 31, 15 and 11. The author selected geo- graphic names to correspond to the numbers used by Veatch as follows: No. 27, Roxand; no. 1, Riley; 31, DuPlain; 15, Stockbridge and 11, Leslie. It was felt that geographic names might be easier to follow than the number system. In the following section, the five natural land types selected for this study'wdll be described with charts, verbal descriptions, photo- graphs and maps. Surface Configuration Profiles. One of the distinguishing features of a natural land type is the Surface configuration associated with it. TABLE III 19 PERCENTAGE DISTRIBUTION OF RURAL NON-FARM, NONAVILLAGE POPULATION WITH DISTANCE FROM LANSING" Township Rural non-farm, non-village Distance from population Lansing Lansing 96 0 miles Delta 57 6 " Oneida 19 12 " Roxand ll 18 " Sunfield 5 24 " ‘ U. S. Department of 16th Census, 1940: Commerce, Bureau of the Census population (37). See FIGURE 2 NATURAL LAND TYPE MAP 20 .SCALE _ JIO MILES.‘ next page for legend 21 NATUnnL LAND TXPE LaGLND 2 7 nOXnND NanrinL LAND rm; 1 stsx NATUfinL Leno rise :51 DUPLAIN NATURAL LAND Tyre STOCKDnIDGd NnI‘UdAL LAND was I 1 Lesms NATURAL LAND use 22 Figure 3 shows the relative differences between the five natural land types used in this study. It can be seen from the chart that there is considerable difference in the surfaces of the five types. The distance represented on the chart is two miles of surface. There is a vertical exaggeration of 22 times to accentuate the vertical differences in the relief. It should be observed that the Roxand. type (No. 27) which is almost flat land has variations in the surface described as swells. The Riley type (No. 1) is smoothly rolling but can hardly be classified as rough. The DuPlain type (No. 31) has dips in the surface where streams and rivers flow; This type of land has an orientation not asso- ciated with the other land types. The surface profile shown here is a valley cross-section. If the profile had run parallel with the valley it would have been.more nearly flat land. ‘The Stockbridge type (No. 15) represents a complex surface of low hills and valleys with short slopes and low to steep gradients. The relief differences are no greater than for the Riley type but changes in topography in shorter distances give the area a more rugged appearance. The Leslie type (No. 11) has smoothly undulating to rolling topography with relatively broad areas between the swells or hills. The slopes are longer than on the other types represented here. Roxand Typg_(No. 27 on the map). Plains of wet or semidwet, dark- colored soils and low swells of lighter-colored soils make up this group. The flat, darker-colored phase may predominate with the higher drier land forming island-like swells; or the two conformations may be nearly equal in proportion, occurring in a succession of low rounded swells of upland with dark-colored valley flats and shallowhdished depressions. 23 FIGURE 3. COMPARATIVE SURFACE CONFIGUATION PROFILES OF FIVE NATURAL LAND TYPES A L‘ ‘ ‘— A ‘— v— i "V’ '. " ' ROXAND RILEY V —— " r DUPLAIN STOOKBRIDGE LESLIE 24 There are shallow drainage swales and pond-holes but few deep pot-hole depressions or sharp drainage hollows; the few streams present have not dissected the land. There are few lakes, some areas having none, and only a comparatively small number of muck or peat swamps. Elevation on the flatter land varies only from 8 to 10 feet and'the difference be- tween the flats and the crests of the higher swells is usually not more than 30 ot 40 feet; slopes are simple and low in gradient. Silt loam and clay loam coverings of 8 to 10 inches overlap a _ relatively impervious clay layer which is calcareous at depths of 18 to 36 inches. This layer rests on massive, finemtextured till clay which extends to depths of several feet. Stoniness is not a distinguishing characteristic, although locally, boulders may be scattered over the surface and in the till. The darker soils are principally Conover and Brookston; the lighter- colored, Nflami. There are also included patches and small bodies of miscellaneous soil types, classified as Napanee, Crosby, St. Clair, Allendale, Granby, Brady, Gilford and Carlisle and Kerston-type mucks. The soils as a whole are medium to relatively high in natural fer- tility, are productive and durable under'cultivation, and.the land is suitable for dairying and a livestock type of farming. Figure 4 shows a photograph of the Roxand land type. The relatively flat surface should be noted. Some characteristics of use may be ob- served in the picture. The scattered woodlots and fine buildings are commonly seen on this land type. Figure 5 gives the soil type distributions on a typical BO-acre farm of the Roxand type. Conover and Brookston soil types constitute .. .. 44 rr— - FIGURE A PHOTOGRAPH OF THE ROXAND NATURAL LAND TYPE __-.__..._—_-..___ _—_.—-___._ _- . - . - u.- i FIGUhE 5 SOIL TYPE AND SLOPL CLAss MAP OF A TYPICAL so sent; FARM FOR THE ROXAE-ID NATURAL LAND TYPE (courrrsm SOIL CONDERVITION enamels) 6UOA J( 4508 GQOA 0mg SLOPE LEGEND ORGANIC SOIL A O _ 2, .70 [77—5] MIAM': L0An1 B 1-6 % CONOI/E/x’ SILT LOAN C é _ /2. ,4 ‘L8_8_O I BMDKSTON LOAN D ,2 _ M, s ‘E /5- 25' 53 27 the major portion of the farm. The simple pattern of soils of this land type can be compared to the complex soil pattern of the Stockbridge type. Rilgy Type (No. 1 on the map). This division comprises much of the land of the Lansing Region. It is characterized by local areas of roll- ing topography, a high proportion of land area in gentle to moderate slopes, a high proportion of wet lowland including peat and muck and a wide range and complexity in soil and topographic components. The configuration features are mostly rounded, constructional in origin.but having no orientation or pattern of arrangement. The depres- sional features are diverse in kind and arrangement. Glacial bowls, pockets, dips, broad shallow depressions; open, haphazardly arranged drainage swales, filled lake basins; numerous intermittent ponds and swales and occasional lakes occur. 'Upland features include glacial hillocks, ridges, rolls, swells, heads, spurs, and other conformations all rounded in character'and composed of smooth short slopes. Much of the area includes broad swells of upland and sags of lowland, each com- ponent including a variety of configuration forms within its own micro- relief. This land appears on either till plains or recessional moraines. Local relief is generally less than 50 feet but may be as great as 100 feet. The principal soil types are the Miami series, including both the relatively more permeable loam and the heavy clay loam underlaid by tight, impervious clay till. The land when viewed from the air gives an impression that the darker-colored pattern of the lowlands was laid over the relatively 28 lighter-colored base. Various gradations of yellows, browns, blacks and grays appear in the pattern. The browns constitute hilltops, the yellow is found on the relatively eroded upland and the group blends into blacks in the lowlands. Locally, there are inclusions of more friable and pervious soils such.as the Hillsdale. Eskers and sand ridges cross the area as scattered bodies of Bellefontaine or Coloma types. The most common soil association is Miami, Conover, Brookston and muck, which in this case appears also as the drainage catena. The original forest cover on the upland contained much sugar maple and beech. It is doubted, however, if these species were universally present and, if so, in greater number than white, red and black oak, elms, white ash, hickories and basswood. The lowland forest consisted of swamp white oak, elm, black ash, red maple, walnut and butternut. Figures 6 and 7 show photographs of the area and a soil type map of a typical BO-acre farm. Attention should be drawn to the undulating nature of the topography and the more diverse pattern of soil types than was true for the Roxand type. DuPlain Type (No. 31 on the map). The division includes land in narrow’valley plains which for the most part are old glacial drainage ways. They are generally not more than two or three miles in width and are traversed by rivers. The rivers in places have cut a lower plain, a few feet deeper, which is floored with recent flood plain alluvium. One or two terrace plains above the alluvial bottom land may be recog- nized in some of the valleys. The old galcial alluvium.is generally coarse in texture; it is gravelly, at least at the base of the deposition, FIGURE 0 PHOTOGLAPH OF THE BILLY NATUhnL LAND TYPE _--.__,.-_.--_.,._._.l. All - FIGURE 7 SOIL TYPE AND SLOPE CLASS MAP OF A TYPICAL 80 ACRE FARM FOR THE RILEY NATUhAL LAND TXPE (COURTSEY SOIL BONothATION SERVICE) $5” 540 <10 6458 c A 5 g 45;? 4.; g 43 MM 0 4500 A 4558 6 4553 Mo ‘0? A 45° we 55 6x0: 8 456C - MIAMI LOAM A 3|.ng ozecemo - MIAm—Comovw COMPLEX B 2.5, % CONOVER SILT LOAN c 4...,2 0/. CONOVER LOAM D /2-/!% 880 BROorcsTom LOAF! E ”45% F oveR 25’ ”/o 31 and even cobbly and bouldery locally. The plains are nearly flat but are intersected in places by inflowing streams from.the adjacent high- land and contain very shallow dry depressions. The wet components are extensive flats having a high water table, large bodies of muck and linear bodies of swampy land bordering the channels of the rivers. There are a few lakes, but lakes are not a characteristic feature of the type as a whole. The soils are various admixtures of the Fox, Oshtemo, Plainfield and Berrien.types on the drier sites; and Brady, Gilford, Bronson, Granby and Maumee on the wetter. Dark-colored clay soils are a minor component; these are generally underlaid by waterlogged sand or sand and gravel at depths of one to three feet. There are occasional spots of dry prairie soil and marsh, undifferentiated loamy sand and gravelly soil consisting of alluvial wash at the mouths of drainage hollows from the adjacent high land, and yellowish and reddish-brown ochreous spots-- a fraction of an acre or an acre or two in size--which may contain frag- ments of bog iron ore. The wetter land supported a dense stand and tall growth of elm, silver maple, swamp white oak, white oak, ash, shagbark hickory and species such as cottonwood, willow and sycamore. The drier land supported an oak-hickory type of forest, with white pine common on the more northern areas. The more loamy soils on the higher sites, and locally, the wet soils where provided with adequate drainage, are producing good yields of the staple farm crops and truck crops. Much of the land however, remains in woods, or is utilized only for pasture. It has only a low cropping 32 value because of flooding, impracticability of artificial drainage, intimate mixture of wet and dry land, low fertility and durability of soil, and in places, extreme stoniness. Figures 8 and 9 show photographs of the land type and a typical soil type map. I Stockbridge Type (No. 15 on the map). This land type includes lowbrelief plains and mainly sandy soils. It comprises a complex of flat and pitted upland, swampy lowlands and soil variants of both. The topography includes flat, rolling upland, with islands of low, narrow ridges lying from.ten to thirty or forty feet above the swamp levels. The soils are dominantly Fox and Plainfield types with smaller amounts of Hillsdale, Miami, Bellefontaine and Coloma profiles, or vari- ants of these. Various types of peat and muck and miscellaneous sandy, gravelly or clayey soils occurred on the lowlands. The organic soils have been classified as Rifle peat, Carlisle muck, Houghton muck, Kégton muck and Greenwood peat. The mineral soil types which include Conover, Brady, Gilford, Granby, Brookston and Berrien types. Lakes are relatively common in the lowland areas. Some have muck shores and bottoms while others are sandy. Some of the lakes are little more than sedge-filled marshes whereas others have high-banked shores and sandy beaches. The variability of the soils and of the lakes produce rather poor agriculture and recreational land. Locally, drained areas of muck have an extremely high value for agriculture. Likewise, some of the lakes are in great demand for the recreational facilities they offer. In gen- eral this mixed condition of the land favors a conflicting mixed use. FIGUILL 8 PM 'JG'mAFTi OF THE LUPLhIN NATURAL LAND TYPE ~7-_- __ , ———~.~._ i FIUUhL 9 SOIL TYPE AND SLOPE CLAo‘b‘ (in? OF A TYPICAL 8O ACRE FARM FOR THE DUPLAIN NATUIuiL LAND TYPE (COURTSEY SOIL CONSEhVaTION Shh’lICh) 0-2. Yo 2:6 ’70 6-12 “/6 {2"}? ‘5/0 Its-”LS "/0 OVER 25 0/0 OSHTEHO LOAMY' SAND Fox Loan 340 F'CY SANDY Lean IG‘IO I co~ovea sun“ LOAM ‘32-‘31 Fox COBHLY Loam '930‘ SFWWNIALLTEKT‘UKEJ-3 I HiLLSQALF, SANUY Low-1 (7:15- GENESEEJALL TEXF~""'LJ 35 Figures 10 and 11 show photographs of this land type and a typical soil type map of the land type. Leslie Type (No. 11 on the map). This grouping comprises rolling and moderately hilly upland on which the Hillsdale type of soil predomi- nates. The land surface has the constructional rounded configuration features common to moraines and till plains--broad swells of upland in- dented by shallow basins and drainage swales, domes, ridges and valley depressions. In most of the areas, the difference in local relief is hardly more than 40 or 50 feet, slopes are smooth and have gradients less than 10 per cent, although locally some of the land is hilly and broken in aspect. Swampy valleys and basins are widely distributed and comprise 10 to 20 per cent of the total acreage in the larger separate areas. Lakes are present but not numerous enough to be an especially distinguishing characteristic; streams are relatively few; The dominant soil, the Hillsdale type, is distinguished by the peculiar yellowish color and friable, granular nature of the subsurface, clayey horizon of the soil profile. The profile is medium or even strongly acid in reaction to depths of 3 or 4 feet; the underlying glacial drift usually exhibits a strong influence from local sandstone and shale bedrock, but is not entirely devoid of limestone influence and, locally, both sands and clays may be moderately calcareous. The soils are medium to low in organic matter, they are sandy loans and light loams in surface texture, with only medium.natural fertility and productivity. They are adapted to a diversity of crops and, properly managed, are capable of remaining as permanently productive agricultural FIGURE 10 PHOTOGRAPH OF THE STOCKBHIDGE NATURAL LAND TYPE , __ _—_—_. .1 _ . .. __ ._ __—....- _ _. __... _ __ —- =--‘.-— . .- FIGUhE ll SOIL TYPL AND SLOPE CLASS MAP OF A TYPICAL 80 ACRE FARM FOR THE STOCKBhIDGE NATUHAL LAND TYPE (COORTSEY SOIL CONSERVnTION SL’nVICE) SLOW: LEGEND ORGANIC ‘SOIL. o— 'L We SANDY ; cam ‘” ego BRADY s/wox- Loam A 340 FOX AND BELLE FONfAm: B a "b 0/0 C 6 —IZ. 0/0 D 845 GRANB Y SANDY L.{l/..\;.\ [1"6 0/0 92C) VJA$HTENAN ALL foruggg 18"2554 If: 38 land. The land is locally bouldery and in a few places, bedrock of sandstone lies at a depth of a few feet. The Miami, Bellefontaine and Coloma soil types are closely associated with the Hillsdale. Moist loamy soils, the material of which has been largely washed in from adjacent slopes, occur in hollows and drainage swales, but the proportion of wet soils, either sandy or clayey, other than poets and mucks does not constitute over a fifth of the land. Figures 12 and 13 show photographs of this land type and a typical soil type map of the land type. Ficus: 12 PHOTOGnAPH OF THL LESLIE NATUhAL LnHU TYPE ,_ I -- “a FIGURE 13 SOIL TYPE AND SLOPL CLASS MAP FOR A TYPICAL 80 ACRE FAhM FOR THE LESLIE NATURAL LnND TYPE (COURTSLY SOIL CONJLhVaTION SLhVICL) ‘ , D \\ HA “\58 if"? 37:3“77?\£;§)a7 e 9“ ° 0 SLOPE Libthv I 010 l ORGANIC 50”.. A 0"”1. c'/o Few BELLEFONT‘HNE 5mm Loan 8 9,... 5, ~75 HlLLSDALE FINE SANDY LOAN c 43—12. Ccmovaa LMM D 12‘ (s / ‘ ,7““1 .a . ‘ , - a "4 " .1. P/A'x‘r- \L'ufi‘ki AL! ir'r' r 9' 111- 25 m 41 PART II STATISTICAL SECTION PRSENT AGRICULTURAL LAND USE Selection 23 Sample Farms Twenty natural land types occur within the Lansing Region. Because i of the lack of precise land type and land use information, adequate samples for study could be obtained for only five natural land type areas. This number was thought to be sufficient to test whether or not land use varied by land type. The land use sample farms were selected carefully from the records of the Production Marketing Administration in Eaton, Clinton and Ingham counties. Sample farms were selected for completeness of land use infor- mation for 1948, for conformity to the natural land type in which the farm occurred and for distance from the nearest town. In addition, only . full-time farms were selected for the sample. This study is based on the "Land Types of the Southern Peninsula of Michigan” by Veatch (39), in which the scale of the map necessitates the inclusion within an individual area on the map of many smaller areas which do not conform to that land type in soil, topography and other characteristics. (If the map were of a larger scale, these smaller areas would be shown as separate land types.) If a 100 per cent sample could have been used in this study, the bias produced by inclusions of non- conforming land in Veatch's scale of mapping would have been eliminated by the size of the sample. Since, however, a somewhat less than 10 per cent sample of farms in any land type area was used, detailed soil maps Were utilized to check the conformity of each farm. within a land type: 42 to that land type. If a farm did not conform in its characteristics to the natural land type it was supposed to represent, it was discarded from the sample. Detailed soil maps were available for only four counties in the region: Ingham, Eaton, Clinton and Livingston. However, only‘farms from Ingham, Eaton and Clinton counties were employed in land use sample. Livingston was not used since it would not contribute any additional land types, but only to the size of samples on land types al- ready selected from the other three counties. Bias of Production Marketing Administration Data Data for land use information and size of farm were secured from the Ingham, Clinton and Eaton County offices of the Production Marketing Administration (28). Such data may and probably do contain a bias in land use. If the P. M. A. has any success with its allotment program, the land use must be different from that on non-co-operating farms. How much PMA data is biased is not known; it would require a comprehensive study to discover the extent of error. For the purpose used in this study, natural land type differences should reflect in use on PMA farms as well as on non-co-operating farms, so that it is felt that the utility 0f PMA data is just as great as an unbiased estimate of population differ- ences on the various land types obtained through an unbiased sample. ‘Jethods 3.1: Analysis 3f Production Marketing Data In the following pages many statistical facts are to be found about land use. In a good number of cases the statement will be made that no information can be discovered in the data about the causes for certain differences. This is not to be taken as an evasion. It was felt that to 43 integrate theory with statistics in the same section would be a bigger task for the reader to interpret than the effort required would be worth. The hypothetical explanation for some of the reasons behind land use are included in a separate section. Later on, a section integrating the statistical with the theoretical is included. Thus an attempt will eventually be made to state explicitly some hypothesis for causes in dif- ferences in land use by natural land type. Total Land Use in the Agricultural Area In any purely agricultural area, there are always certain other uses complementary to, or symbiotic to, the agricultural use. One major land use is always found associated with agricultural land use in even the most primitive society. This is transportation land use. Roads of some kind are always present as a complementary use to agriculture. In most countries advanced in technology, communication land use such as electric and telephone line right-of-ways occurs in association with farm land. In almost every agricultural area of the humid region, forest land occurs as a symbiotic use. A non-symbiotic use to agriculture which occurs in many farming areas is idle land. This is land that may or may not have been used for agri— culture in the past but at present maintains neither forestry or agricul- ture. Such land may be considered from the agricultural production viewpoint as creating space. In an agricultural community where spaces of idle land intervene between farms, increased production and marketing costs arise from distance of transport alone, without compensating in- creases in value of products sold. This accounts for the non-symbiotic 44 nature of idle land. Water resource land may fall into this non-symbiotic categomy. Recreation land may be either symbiotic or non-symbiotic de- pending on the nature of the recreation offered and the space occupied. If recreational land offers an increased market for farm products locally, the use of space may be offset by the increased value of the farm products. In presenting a comprehensive characterization of land use in an area occupied mostly by agriculture, there are three pictures to view: the agricultural land use, the major uses symbiotic to agriculture and the non-symbiotic uses. In the five natural land type studies, all but one, the Stockbridge type (No. 15), were occupied almost entirely by agricultural uses. Approximately .25 per cent of the land on all types was occupied by roads. This was due to the gridiron pattern of roads, with roads distributed in north-south and east-west directions at one mile intervals, on all land types. The Stockbridge type (No. 15) had only 60 per cent of the land in farms. The remainder was in idle land or in State Game areas. The Rose Lake'Wildlife Area and the Dansville State Game Area occupied large tracts of the Stockbridge land type. The peculiar mixture of soils, mainly wet lowlands and dry infertile uplands, plus the small farm sizes settled in the area, caused uneconomic farms to be developed that had to be later abandoned. Table IV provides a picture of total land use on selected agricul- tural areas of each of the five natural land types studied. State Game Areas did not appear to differ essentially from other idle land in the region; thus, they were included as idle land on the table. TABLE IV USE OF LAND IN PPR CENT ON A 2560 ACRE SAMPLE FRO” EACH OF FIVE SELECTED NATURAL LAND TYPES (ESTIMATED BY AUTHOR) 45 Roxand Riley DuPlain Stockbridge Leslie (27) (1) (31) (15) (11) land in roads .25 .25 .25 .25 .25 Land in farms 99.75 99.75 99.75 59.75 99.75 Idle land .00 .OO .00 40.00 .00 46 Farm Land Use After the broader picture of total land use in the agricultural hinterland is observed, the next aspect to be considered is the broad use of the farmland on the various selected natural land types. Table V gives a view of farm land use in two categories; tillable and non-tillable land. Tillable land as used here includes all land plowed including plowed pasture and idle land. It specifically does not include all land that could be plowed. Non-tillable land comprises land used for roads, fences, ditches, homesteads, barnyards, paddocks, permanent pasture, woodlots, swampfls, lakes, gravel pits and dumps. Land types 27, 1 and 31, the Roxand, Riley and DuPlain land types, all had about 75 per cent tillable land. Stockbridge type (No. 15) had 58 per cent and the Leslie type (No. 11) had 68 per cent tillable land. Surface configuration and quality of soil appear to be factors that cause land to stay out of cultivation on all natural land types con- sidered. Surface configuration includes two factors which cause land to stay out of cultivation: slope and drainage. If the slope is too great, modern machinery cannot be used and thus the land remains untilled. Undrained land such as that occurring in wet swales cannot be cultivated. Quality of soil is an economic factor that may prevent cultivation. If the land is so unproductive as to be unprofitable, it will stay un- worked. Other factors that cause land to remain untilled are 1) type of farming: livestock farming demands more land for fences, small paddocks and lanes than crop farming; 2) habits of the operator: Swiss farmers may use land differently than English American farmers (21), and 3) TABLE V USS OF FAR“ LAND BY NATURAL LAND TYPE 47 Roxand Riley DuPlain Stockbridge Leslie Per cent farm land tillable 78.38 75.71 75.04 57.78 68.06 Crop land 78.31 75.64 75.04 57.67 67.58 Orchard .07 .07 -- .ll .68 Per cent non- tillable land 21.63 24.28 24.96 42.22 31.94 48 size of farm: small farms have a greater per cent of tillable land than large farms. The per cent of tillable land is a relatively stable factor. Al- though the data‘:: not presented here, the census of 1880 compared with that of 1945 indicates that very little change has taken place on the various land types since that time in percentage of tillable land (35). This strongly suggests that the per cent of land tillable on any particu- lar land type is more a function of natural land than it is of economic conditions or whims of farmers. If each land type is analyzed individually, some of the reasons be— hind the percentages of tillable land existing on any natural land type may be hypothesized. If the Roxand type (No. 27) is examined, the reasons for 25 per cent of the land remaining uncultivated may be found. Approximately 10 per cent of this land type remains in woods, but since most of these woods are found on the wet land that comprises about half the area, it is not the woods that prevent cultivation but poor drainage. Probably the main factors preventing cultivation in addition to wet land are in- stitutional factors. Such uses as determined by the type of farming and the habits of the farmer are in this group. It is probable that the longer this land type remains in farming, the larger the percentage of tillable land will be up to the point where virtually all the land is tilled. There are some areas that are diffi- cult to drain. Few areas however are impossible to drain. As the land becomes drained, the woodlots will disappear in favor of the cultivated crops. As an example, in north Eaton county, it is not unusual to find 49 many individual farms on which 78 out of 80 acres of each farm is under cultivation. This happens to be in a well drained section. A few miles from here are farms that are not properly drained as yet that have only 40 to 50 acres out of 80 drained. From observation and the history of drainage of this land type, it seems likely that the census of 1880 listed too much land as being till- able. If Roxand township in Eaton county is taken for an example, the land in farms has not increased since 1880. Neither has the per cent of land tillable increased according to the census of 1945. Even to the casual observer, the land that is tilled can be seen to be increasing due to increased drainage produced by ditching within.the area. The only con- clusion that can be drawn is that the estimates of tillable land were too high in 1880. The Riley land type (No. l) is comprized of undulating to moderately rolling topography. The slopes are short and frequently steep. In local areas slope too steep to easily till with modern equipment may comprise up to one-quarter of the land area. This is enough to explain the 25 per cent of non-tillable within the land type for these areas. In addition to the slope factor, the presence of swamps, swales, pond holes, and wet muck pockets accounts for some of the untillable land. It is unlikely that many of these wet areas will ever be drained due to the technical difficulties of drainage. At the present state of land use, woodlots frequently occur on upland soil that could be tilled if cleared. It ap- pears likely that these woodlots will be cleared and the percentage of tillable land increased a few per cent. 50 Land type 31, the DuPlain type has 25 per cent of its land out of cultivation for several reasons. Much of the land within this type lies along the banks of streams and rivers. Such land is flooded every spring and frequently is wet or water-logged all summer long. Back from the streams a short distance are short steep slopes of former streams that flowed in this land type area. The upland terraces comprise soils that are sandy in texture and relatively unproductive. Locally, areas so covered with stone and boulders occur that cultivation has not been attempted. This combination of wet land, local steep slopes and unpro- ductive landwill probably act in the future to prevent much of an in- crease in tillable land in this area. The Stockbridge land type (No. 15) is composed of such a mixture of low producing soils, amt difficult to drain.muck swamps and local steep slopes that it is not difficult to see why only 58 per cent of the land was tillable in.1948. an a large number of farms in the area, a much smaller per cent than this is farmed. If individual farms on the land type are studied, the great diversity in land use from farm to farm can be seen. One farm may have 80 per cent of the land in muck that has been drained and tillable, the next farm down the road may consist of 10 per cent muck and 90 per cent Coloma and Plainfield sand. The first farm may be prosperous and well kept, and the second run down with broken down fences and dilapidated buildings. On other land types in the study, one can expect on individual farms to see some resemblance between the average of tillable land on the farms and the average for the land type. 0n the Stockbridge type there is seldom a farm.that corresponds to the land type average. 51 During the depression of the thirties, much of the land was repos- sessed by the state for non—payment of taxes. If a farm had enough productive land to pay the taxes and maintain the family, the farm did not revert to the state. After the depression, the State went into areas of Stockbridge land and purchased farms for state game areas. The effort involved in.breaking up farms so as to sell the poor land to the state and keeping the productive land in agriculture was apparently too great, because only in a few instances was this done. It appears probable that if some other system of land description were used than.the rectangular system that much of this land would have never been taken into the boun- daries of farms in the first place. On those farms that have developed a farming system to take advantage of whatever kinds of land they happen to possess it seems unlikely that the per cent of tillable land will be increased much. The Leslie type (No. 11) comprises land that varies from gently undulating to quite strongly rolling. Slopes are long and locally may exceed 15 per cent ingredient. The more common gradient is 6 to 8 per cent. Locally the land may be stoney and cobbly. Frequent inclusions of less productive soil occur within the land type. Broad drainage valleys of low lying sandy soil are common. In the more rolling areas of the land type it is steepness of slope that prevents cultivation. On the undulating areas the broad swales of wet land are the limiting factor. ‘Woodlots commonly occur on the hilltops of this land and on the wet swamps of the lowlands. It appears likely that in time considerable of the low wet land will be cultivated but probably the steep slopes will remain untillable. 52 Size 2: Farm Farm size is an aspect of land use that differs from one natural land type to another. Table VI shows the average size of farm and aver- age acres of tillable land per farm on the land types under consideration. The smallest aVerage size of farm occurs on the DuPlain type (No. 31) with 97.3 acres of land. This type also has the smallest number of till- able acres per farm.with 73.0 acres. The largest farms both in total acres and tillable acres occur on the Leslie type (No. 11) with 132.1 total acres and 89.0 tillable acres on the average farm. The remaining land types fall between these extremes. The greatest difference in total acres between.any two land types is 34.8 acres. The greatest difference in tillable acres between any two types is less than half that of total acres, 16.0 acres. It would appear from this that farmers attempt to purchase a rather definite number of tillable acres to farm. In order to obtain this number of tillable acres, the size of farm will vary from one land type to another. There is, however, a considerable difference between land types in the acres of tillable land per farm. This differ- ence cannot be explained from.auch simple statistics as these. The dif- ference is very probably due to changes in types of farming and quality of farmer from one land type to another. This relationship is not a clear one, however. Table VII gives a breakdown of the percentage of each of several farm size on the various land types. A general characteristic on all. land types is the high proportion of 71-90 acre farms. On the average, about 1/3 of the farms are in this group. Most of these farms are 80 acre farms. The next important group is the 111-130 acre farms, most of TABLE VI SIZE OF FARM BY NATURAL LAND TYPE 53 Roxand Riley DuPlain Stockbridge Leslie Average size of farm in acres Average crop land in acres 107.9 99.4 97.3 128.9 84.5 75.2 73.0 74.3 132.1 89.0 TABLE VII 54 PER CENT DISTRIBUTION OF FARM SIZE BY NATURAL LAND TYPE (P. M. A. DATA) Acres Roxand Riley DuPlain Stockbridge Leslie 31-50 12.5 14.6 20.0 14.7 9.8 51-70 7.3 7.7 5.0 10.0 6.5 71-90 36.0 33.2 38.0 25.1 17.0 91-110 9.0 10.7 12.0 7.1 11.1 111-130 12.8 13.6 5.0 14.7 16.3 131-150 3.1 2.8 1.0 4.7 6.5 151-170 5.8 7.9 9.0 4.3 7.8 171-190 2.8 2.8 2.0 3.3 5.2 Over 190 10.7 __6_:_6_ __8__0 1_5__§ 11.3 Total 100.0 100.0 100.0 100.0 100.0 Per cent under 130 77.6 79.8 80.0 71.6 60.7 -__ 55 which are 120 acre farms. Third are the 31-50 acre farms, mostly 40 acres in size. The high proportion of small farms on all land types is evident when it is seen.that on the various types, the percentages of farms less than 130 acres in size are: Roxand - 77.6%; Riley - 79.8%; DuPlain - 80%; Stockbridge - 71.6% and Leslie - 60.7%. About half the farms on all land types are less than 80 acres in size. The small size of farm is even more evident if it is observed that in calculating the percentage figures, all farms less than 30 acres in size were eliminated. It was felt that farms less than 30 acres in size were part-time farms and were omitted for this reason. There was a sizeable proportion of farms larger than 190 acres on only two of the natural land types, Stockbridge and Leslie (Nos. 15 and 11). These types had 15.6% and 19.6% respectively of farms greater than 190 acres. It might be said that on theoretical grounds Stockbridge type should have larger farms than the Riley type. The less productive soil would be expected to have larger farms. But when it is observed that the Leslie type has a greater percentage of large farms with its more productive land than the Stockbridge type, this argument seems to have little basis. Undoubtedly there is a reason for the distribution of farm sizes but this reason is not at all evident in the statistics. In the theoty section of this study, some light may be thrown on this problem of farm.size. Farming Patterns The areal distribution of farms by land types is given in Figure 14 to indicate that no particular pattern of farms exists on any land type EIGURE 14 man PATTERN BY NATURAL LAND :YPA ROEAND TYPE 56 80 A. no FIGURE 14 FARM PATTERN BY NATURAL LAND TYPE RILEY TYPE -‘ 57 f». FIGURE 14 FARM PATTERN BY NAmRAL LAND TYPE NPLAIN TYPE ' - l ,\:I {I} e : (+Q 1:50 3x} xx. FIGURE 14 FARM PATTERN BY NATURAL LAND TYPE STOOKBRIDGE TYPE 59 [J1 FIGURE 14 FARM PATTrLRN BY NATURAL LAND TYPE LESLIE TYPE .-- . . . _. _ - 1+0 A. 80. A. 160 A. 61 area (28). The pattern is apparently randomly determined. Farm boundaries and size are found to be a function of the land description system set up for the Northwest Territory under the Ordinance of 1787. Square and rectangular farms are the rule because of the way in which land was laid out and sold under the ordinance. Figure 15 gives the field pattern on an average 80 acre farm for each of these land types. Surface configuration and wet land patterns apparently alter field arrangements on the Stockbridge type only. Square to rectangular fields are the rule wherever conditions permit. It can be stated as a generality, however, that field pattern was more affected by natural land character than was farm pattern (28). Figure 16 gives the woodlot distribution on the land types (2). It can be seen that a greater proportion of the Stockbridge and Leslie types are in woods than the remaining three natural land types. If the woodlot distribution map could have been made on a larger scale, it would have been.evident that wherever flat land occurred, the 'woodlots occurred mainly on.the wet lands, and there only away from the roads. In north Eaton County, in Roxand and Sunfield townships in some areas it is almost impossible to find a woodlot lying adjacent to a road. Apparently farmers cleared the land closest to the road first because their houses were located by the roads. It required less time and effort to work the land close to the buildings than the land farther away. On the rolling land, the pattern of woodlots is more closely asso- ciated with rough hilly land and wet undrained land. This is true of Leslie, Riley and Stockbridge types. Even here, there appears to have been a tendency to clear that land closest to the roads first. 62 FIGURE 15 FIELD PATTERN BY NATURAL LAND TYPE ON A TYPICAL 80 ACR£ EAR! ROXAND TYPE G PP C 6 G C o C N C~~Cropland PP-tPermanent pasture W~~Woodlend RILEY.TYPE C If C. u 6 R 63 FIGURR 15 FIELD PATTERN BY NATURAL LAND TYPE on A TYPICAL 80 A083 PARK OF STOCKBRIDGE TYPE - _ .— ._._ _ _ ——v.-. G—-Cropland PP--Permanent pasture ‘I‘O’ - ~ 7:. O C d 8. DUPLAI“ TYPS (1 C (2 [ Pl FIGURE 15 FIELD PATTERN BY NATURAL LAND TYPE on A TYPICAL so‘Acaa FARM or LESLIE TYPE (I C-mCrOpland ifmnlknwranmxt gesture W-- floods 64 65 FIGURE 16 IIOODLOT DISTRIBUTION BY NATURAL LAND TYPE ROXAND RILEY QM” M.- 1. ' 2&5;- . 1‘ ' J ‘ ' 9: "-5 “if 55' ".P "i”; ;’. w\ 1‘, . I 1 :0 «e ' ' .q o “‘51 {A I" 3.; . i o '1 at? I M fi'h. 2‘“ “'11:: .3. J [Ir .0 "cut!" , O 9 5T0 CK BRIDGE. DU PLAIN Mk ._ .,\. ‘ ., ‘-.,. “a-” . ..'. ‘V :"' l . . . ‘ - ‘ w s ' L". 3 H . ‘ “A ‘u 3 I a \ l ‘ ‘ fl 5. m , . x w 'v. ‘ \lfi ‘ . I * ‘ ' .' 'k 5.. ‘ u ' “' '“. ‘. ,"' x 0' Ice - . A I ' UN (lg. n - .' ‘ .. ya) 1., , .. ‘ V I . a I ' ‘ I d ' H J . t J L...‘ ' ”I. ' i A . LESLIE .3" ' ‘5 . ‘1 ' 5 g .. .’ + SCALE '/zmcu=l mu: .2 5 I ' .. ' . ? \.' ‘ A f f II 3' i» a .0. _ I I I ‘ . M. I' n’ 1 7 I 1 66 Tillable 3.3112 _l_l_s_e_ The per cent of tillable land devoted to any crop grown in an area is determined by the per cent of farmers growing the crop and the number of acres of the crop grown per farm. Attention is drawn to this relationship because it points out the two decisions made by farmers which cause land to be devoted to a certain crop within an area. (The first decision is to grow the crop, the second, how much of it to grow.) It is possible for various combinations of these factors to cause perhaps 50% of the land within an area to be devoted to corn. For example, if all farms were of equal size, half the farms in any one year might devote 100% of their land to corn and the other half grow none, so that 50% of the land area in all farms would be in corn. Another combination producing the same result would'be for all farms to devote 50% of their land to corn. Numerous such arrangements could be listed. Table VIII gives the per cent of farmers reporting the various crops in 1948 by the selected natural land types. Wheat, corn and oats 'were reported as the crops grown on the greatest percentage of farms on all land types. If the per cent of farms growing any kind of tame hay could have been obtained, it probably would have been reported on as great or greater per cent of farms than corn, oats and wheat. As it was, there was no way of knowing, of the farms reporting alfalfa, red clover and other tame hay, which farms grew two or more kinds of hay. It is clear that the dominant type of farming is centered around the production of the staple feed crops, oats, corn and hay and the cash crop, wheat. TABLE VIII PERCENTAGE OF FAR‘S REPORTING VARIOUS CROPS (P. w. A. DATA) 67 Roxand Riley DuPlain Stockbridge Leslie Wheat 92.9 88.9 78.0 66.8 84.4 Corn for grain 89.5 85.3 86.0 78.3 87.5 Silage corn 19.0 23.1 21.0 25.3 20.6 Soybeans 4.8 .5 7.0 1.8 Potatoes 1.4 1.2 2.0 4.6 5.6 Dry beans 28.2 19.9 9.0 7.8 5.6 Sugar beets 3.7 2.2 7.0 1.4 Oats 85.4 83.5 75.0 66.8 81.9 Barley 3.4 4.9 7.0 .5 1.9 Rye 1.4 2.9 14.0 6.0 5.0 Vegetables .7 4.7 3.0 6.5 3.8 Other cash crops 5.8 1.7 3.0 12.9 1.9 Alfalfa 38.1 44.5 45.0 49.3 53.8 Clover 48.3 61.9 47.0 66.8 59.4 Other Tame hay 23.8 15.5 11.0 12.5 20.6 Temporary pasture 20.4 19.7 4.0 18.0 14.4 Rotation pasture 46.9 47.7 56.0 32.7 54.4 Summer faIIOW' 11.6 12.0 12.0 15.7 6.3 Idle crepland 10.2 15.5 9.0 46.5 18.1 Green manure 2.0 2.0 1.0 3.2 2.5 68 The per cent of farms reporting a crop within an area is, in a measure, an indication of farmers' opinions regarding the possibility of economic success with the crop. If 100% of the farmers of an area grow”wheat, it can be assumed that wheat is adapted to the land of the area and to the farming systems. If wheat were reported on only 25% of the farms, it could mean either that wheat was not adapted and that 25 per cent of the farmers had made wrong decisions, or else that it was adapted to the land but only 25 per cent of the individual farmers could use wheat in their farming systems. A measure of this latter conclusion can be determined by means of a historical study. If 25% of the farmers report wheat consistently through the years, the crop can be assumed to be adapted to all the land of the land type, but not to all the farming systems. If the land types are compared, some comments can be made about each crop. Wheat was generally reported on all land types. Large dif- ferences existed between land types, however. Only 67 per cent of the farmers reported wheat on the Stockbridge type. This was 11 per cent less than the next lowest, the DuPlain.type and 26 per cent less than the Roxand type, with 93 per cent of the farms reporting wheat. Since wheat is a cash crop in most of the Lansing region and is not necessary to the farming system, it can probably be assumed that the per cent of farms reporting the crop is a measure of true comparative advantage of the crop over others in the farming system.on any individual natural land type. This comparative advantage is a function of crop adaptation to the land and the price relations of wheat to all other crops that could be grown. Although it is not shown here, study indicates that 69 wheat has been reported consistently on the various land types in essen- tially the same relationship between land types as reported here. This would indicate that wheat is better adapted to the Roxand, Riley and Leslie land types than to the DuPlain or Stockbridge types. Farmers' opinions and the yield figures available confirm this conclusion. Most farms that reported wheat also reported corn, oats and some type of hay. A study of thirty-five farms selected at random from each land type gives the percentage figures for farms reporting all four of these crops: Roxand - 80 per cent; Riley - 83 per cent; DuPlain - 68 per cent; Stockbridge - 57 per cent and Leslie - 63 per cent. This in- dicates the prevalence of this general farming type of farming system on all land types. An even more general relation on all land types is the almost per- fect correlation between corn and oats reported. Practically every farm that reported corn also reported cats on all land types. The reason for this is that oats is almost the only cr0p that will follow corn in the crop rotation. Barley is not very well-adapted to the land types listed here and the table bears this out. It and spring wheat are the two grain crops that wdll follow corn. In cash crop areas, beans and sugar beets follow corn on some farms. The prevalence of this corn-oat crop sequence points out the importance of considering crop rotations rather than individual crops when attempting to explain farm crop ecology. Such crops as soybeans, potatoes, sugar beets, barley, rye, vegetable crops and other cash crops are reported on such a small per cent of the farms that irrational behaviour of farmers rather thfi‘n any particular adaptability of the land for the crops might explain their presence. 70 Even with these crops there are differences between land types, however, that lead to the modification of this conclusion. If sugar beets are taken as an example, it can be seen that on the DuPlain type of land, 7.0 per cent of the farmers reported the crop. 0n the Leslie type, no sugar beets were reported, even though a market for sugar beets existed in the area where this type land was located. Apparently there were no farmers willing to grOW'sugar beets, indicating that irrationality among farmers extends only so far in explaining crop incidence. The per cent of farms reporting ensilage corn, rotation pasture and temporary pasture measures in.a rough way the presence of livestock en- terprises. It is unfortunate that livestock information was unavailable from.PMA data. The picture can be pieced together-with the figures on these crops in addition to the information given about the corn-oats- 'wheat-hay relationship. All in all, there seems to be no great amount of difference between land types in the per cent farms reporting live- stock. From.the study of individual farms on the land types, it was dis- covered that the Roxand and DuPlain types had more cash crop farms than the other three types. The Riley and Leslie types had the most dairy- general farms and the Stockbridge type had more of a dairy-specialized crop type of agriculture. The decision made by a farmer after he decides to grow a crop is how much he shall grow. This is presented in.Table IX. The acres grown on any farm is detennined by at least two considerations: the size of the farm and the relative importance of that crop in the farming system. The acres per farm resulting from.each of these two forces cannot be determined from.this table. The size of farm.factor was eliminated to TABLE IX ACRES PER FARM REPORTING VARIOUS CROPS (FROM P. 7!. A. DATA) 71 Roxand Riley DuPlain Stockbridge Leslie Wheat 21.3 20.5 18.7 15.7 20.9 Corn for grain 17.3 13.7 15.9 14.7 17.9 Silage corn 9.1 8.9 8.7 9.9 10.1 Soybeans 8.6 28.0 7.4 5.6 ---- Potatoes 5.5 3.4 3.5 2.3 9.1 Dry beans 13.8 12.8 8.3 8.9 9.8 Sugar beets 8.4 12.8 7.4 4.0 ---- Oats 15.6 13.2 15.0 14.1 15.5 Barley 11.7 9.9 10.7 3.0 7.7 Rye 9.8 8.2 10.9 7.1 7.2 Vegetables 8.0 3.7 8.3 6.4 4.3 Other cash crops 6.7 18.0 12.7 17.1 8.0 Alfalfa 14.7 11.7 11.6 13.4 18.0 Clover 13.1 14.3 14.8 12.3 15.2 Other tame hay 13.4 9.1 22.6 11.1 12.7 Temporary pasture 16.7 10.6 9.3 13.6 13.6 Rotation pasture 16.1 14.8 15.7 21.0 32.6 Summer Fallow 11.3 11.4 11.7 9.3 16.7 Idle Cropland 9.8 11.9 14.4 17.4 18.1 Green Manure 9.5 11.9 24.0 21.4 14.5 72 study further the relative importance of the various crops in the farm- ing system. The conclusion from this study was that generally, the size of farm affected only the acres of crops grown without affecting the ratio between.crops until a small farm size was reached. 0n farms smaller than about 60 acres less individual crops were grown with a relatively larger per'cent of the farm devoted to each crop. Table X, the per cent of tillable land in the various tilled crops, presents an integrated picture of the per cent of farms reporting various crops and the average acres of the crops grown on the farms reporting them. This table provides an over-all view of the use of tillable land but is not a very useful tool for analysis. Changes in land use from year to year on the natural land types could be used in a statistical study of the elasticity of supply of crops. In this way aggregate pictures such as this could be used. While not a very good tool for analyzing land use, this table does show the resultant of all the interacting factors producing land use. Historical study bears out that this picture is to a great part, the function of the natural land type. While land use may vary on any one land type from year to year, the differences between the types persist. In reconnaissance studies of the census extending back to 1880, the author found that more wheat was grown on the Riley and Leslie types than on the Stockbridge type. It is through experience of many years that farmers of the Lansing Region have come to recognize that only certain land types are good land for beans or corn, sugar beets or peppermint. This thesis set out to prove whether land use varies by natural land type. This table provides the evidence for tillable land. It should be TABLE X PERCENTAGE OF TILLABLE LAND IN VARIOUS CROPS (P. n. A. DATA) Roxand Riley DuPlain Stockbridge Leslie Wheat 23.4 24.2 20.9 14.0 19.8 Corn for grain 18.2 15.4 18.7 15.5 17.5 Silage corn 2.0 2.7 2.5 3.3 2.3 Soybeans .4 .l .7 .l Potatoes .0 .0 .l .1 .5 Dry beans 4.6 5.5 1.0 .9 .0 Sugar beets .3 .3 .7 .0 Oats 15.7 14.6 15.4 12.6 14.2 Barley .4 .6 1.0 .0 .1 Rye .l .3 2.0 .5 .4 Vegetables .0 .2 .3 .5 .1 Other cash crops .4 .4 .5 2.9 .l Alfalfa 6.6 6.9 7.1 8.8 10.8 Clover 7.4 11.7 9.5 11.0 10.1 Other tame hay 3.7 1.8 3.4 2.6 2.9 Temporary pasture 4.0 2.7 .5 3.2 2.2 Rotation pasture 8.9 9.3 12.0 9.2 12.8 Summer fallow 1.5 1.8 1.9 1.9 1.1 Idle cropland 1.1 2.4 1.7 10.8 3.2 Green manure .2 .3 .3 .9 .3 Number of farms 294 407 100 217 160 r - 74 stated that a difference between any two percentage figures as small as 1.5 per cent mathematically is a highly significant difference, with the sample sizes used and for percentage figures between 10 and 30 per cent. It is seen that the per cent of wheat grown on the Stockbridge type in 1948 differed significantly from the Leslie and DuPlain types which in turn differed significantly from the Roxand and Riley land types. To take a specific example, the Stockbridge type had significantly less land than the Roxand type devoted to wheat, corn for grain, dry beans and oats. It had significantly more land in other cash crops (mainly peppermint and spearmint), alfalfa, clover and idle cropland. There was no significant difference existing in the other crops grown. When analogous comparisons were made between all combinations of land types taken two at a time, it was found that the two land types 'which were most nearly alike (Roxand and Riley) varied significantly in at least three crops grown. The two land types varying most (Roxand and Stockbridge) showed significant differences in seven crops grown. In previous studies in Michigan (17), (38), the Roxand and Riley types have been consolidated along wdth several other land types into a land group called Class 1 land. While the information given here shows that there is some basis for this classification, a better characteriza- tion of use could have been made if the natural land type system had been used to stratify land differences. A summary of tillable land use is shown on Table XI. Here the twenty crops grown have been grouped into six classes of crops: corn, small grain, tame hay, plowed pasture, cash crops and other crops. Here again a difference of slightly less than 1.5 per cent is a highly PER CENT CF TABLE XI TILLARLS LAND BY CROP GROUPIYCS ON ELECTED NATVRAL LAND TYPES 75 Roxand Riley DuPlain Stockbridge Leslie Number of Farms Sampled 294 407 100 217 160 Corn 20.3 18.2 21.2 18.9 19.9 Small grain 39.8 39.8 38.6 27.3 34.6 Tame hay 17.8 20.5 20.0 22.5 23.9 Plowed pasture 12.9 12.1 12.5 12.5 15.0 Cash crops 6.0 4.6 3.4 4.8 1.5 Other crops 2.9 4.5 4.0 13.7 4.8 Ayerage acres per farm 107.9 99.4 97.3 128.8 132.1 Average acres cropland 84.5 75.1 73.0 74.3 89.0 76 significant difference in use, employing the method of the least signifi- cant difference between percentages. After comparing all land types as before (two at a time), the following observations can be made: The Roxand type differs from the Riley in the per cent of land in corn, tame hay, cash crops and other crops. Roxand differs from DuPlain in the per cent of tame hay, cash crops and other crops grown. Roxand and Stockbridge differ in corn, small grain, tame hay, cash crops and other crops. Finally, the Roxand type differs from Leslie in the use of its land for all classes of crops given but corn. Other combinations are: Riley and DuPlain which differ in the per cent of corn grown, Riley and Stockbridge which have differences in corn and tame hay only and Riley and Leslie which differ in small grain, tame hay and cash crops. DuPlain differs from Stockbridge type in corn, small grain, tame hay and other crops; from Leslie in small grain, tame hay, cash crops and plowed pasture. The final comparison is between Stockbridge and the Leslie types. These types differ significantly in the per cent of small grain, cash crops, plowed pasture and other crops. All these combinations are given to indicate that even with a crop classification as general as given here, that tillable land use varies significantly from one land type to another, in at least one class of crop. From.this table, it can be observed that corn, for example, appears to be a crop more adaptable to all land types studied than the small grains. There was a difference of 3 per cent between Riley and DuPlain 77 in the per cent of tillable land in corn. At the same time differences in small grain were as great as 12 per cent between Riley and Stockbridge land. The per cent of land in other crops varied almost as much between individual land types as did small grains. The Stockbridge type had almost 11 per cent more tillable land in this class than the Roxand land type. This class of tillable land consists mainly of idle land. Effect of Farm Size on Tillable Land Use **~—-_ In the above section it was seen that land use varied by land type. Why it varied was not explained. This problem.will be taken.up here. It was seen in a previous section that size of farm varied signifi- cantly from one land type to another. It was also found to be true that land use varied by land type. The question which now arises is whether the difference between land types in land use is due to a difference in size of farm. This question was studied first on 94 farms of the Riley land type in Ingham County to discover how much change in land use resulted from a change in size of farm. Table XII gives the per cent of land devoted to six classes of crops for all 94 farms, for 17 small farms less than 50 acres in size, for 17 large farms of more than 190 acres and for 24 eighty acre farms. The conclusions that can be drawn are these: as farm size decreases, the per cent of tillable land increases, the per cent of land in small grain increases and the per cent of land in tame hay decreases. Cash crops are almost absent on the farms smaller than 50 acres. TABLE XII TILLABLE LAND USE IN PER CENT FOR DIP BRENT-SIZED PARTS ON RILEY LAND TYPE IN INGHA“ COUNTY 78 All farms l7 farms under 24 farms of 17 farms over 50 acres 80 acres 190 acres Corn 18.3 18.7 14.4 21.0 Small grain 40.0 50.6 40.9 37.2 Tame hay 20.8 17.2 19.4 21.3 Plowed pasture 9.3 9.3 7.8 9.7 Cash crops 5.1 .2 6.5 6.0 Other crops 6.2 3.8 10.8 4.7 Average acres per farm 100.6 28.4 80.0 221.7 Average tillable acres per farm 74.4 23.2 60.9 159.9 Per cent tillable land 74.0 82.0 76.1 71.3 79 Table XIII is an extension of the idea just developed, but for three land types. Here comparisons of land use are made of 80- and l60-acre farms on each of the three land types. The same relationships mentioned for the Riley land type farms from Ingham county were found to be true for Roxand, Riley and Stockbridge types in Table XIII for the entire region. It can be generally concluded that although land use dees vary by size of farm on the same land type that for the same size of farm on different land types, land use differ- ences vary significantly. Using only 80 acre farms, a comparison can be given to show the following: land in corn varied from 20 per cent on Roxand type to 15 per cent on Riley land; small grains from 29 to 39 per cent respectively on those two types; hay from 20 to 24 per cent; cash crops from 2.8 to 5.6 per cent; plowed pasture from 8.7 to 14.5 per cent and other crops from 3.6 to 17.9 per cent. If further stratification of the statistics could be made, factors such as age of farmer, tenancy grouping and type of farming could be isolated and tested for their effects on land uSe. The information for such division of land use data was not available for use in this study. Thus the short discussion of the effect of size of farm on land use on a given land type can be taken as an example of a method for the study of the effect of any factor on land use. Studying the relation of a factor to land use still does not answer the more fundamental question of what causes the particular factor to be as it is. For example, size of farms within a land type is a factor affecting land use, but the question remains of what causes the farms to have the particular sizes they do. This study of cause and effect cannot be answered in this section. The problem.will be discussed in Part II under the land use theory sections. TABLE XIII TILLABLE LAND USE BY FARM SIZE ON THREE NATVRAL LAND TYPES Land Farm Type Roxand Riley Stockbridge Size BO-acre farms 19.7 14.9 17.4 Corn Large farms 21.1 20.7 21.6 S 38.2 38.8 29.3 Small grain N 38.9 39.0 29.1 S 20.8 20.3 23.9 Hay L 17.6 19.3 23.7 S 5.6 4.9 2.7 Cash crops L 5.4 4.7 2.2 S 12.0 14.5 8.6 Pasture L 13.1 13.9 12.0 S 3.5 6.5 17.9 Other land L 3.6 2.2 11.1 S 80.0 80.0 80.0 Acres in farm L 220.9 195.1 214.8 S 63.7 63.2 51.7 Acres in crops L 166.9 149.7 126.2 8 79.7 79.0 64.6 Per cent land tillable L 75.5 76.7 58.7 81 INTENSITY OF LAND USE General Discussion Von Thunen (41) has given the basis for the belief that intensity of land use should vary with distance from market. Klimowski (22), mak- ing further study of Von Thunen's conclusions, has shown that intensity of use should vary also with the productivity of land. Thaden (32) in his excellent study of the Lansing region demonstrated that use intensity varied with the distance from Lansing in 1930. This has been shown in other studies (6), (10), thus it can be taken as a demonstrated fact. No figures relating to intensity of use with distance from.town are in- cluded in this study for that reason. Brinkman (6) has stated that soil productivity differences often exerts a stronger influence on intensity than the distance from.market. He was careful to point out, however, that this influence of soil dif- ferences became relatively more important as distance from.market de- creased. This being the case, the greatest differences in land use be- tween land types should occur close to market. Intensity By Land Type The sample farms for this study were taken on an average of about fifteen miles from Lansing. It is possible that there may not have been full opportunity for influences from the market to be asserted in land use intensity on the various land types. Furthermore, it is probable that the intensity measures employed, may not have been the best measures to use. Table XIV gives the measures that were used with the resulting intensity figures. Data for this table were secured from the Minor Civil 82 TABLE XIV LAND USE INTENSITY MEASURES FOR.FOUR NATURAL LAND TYPES Roxand Riley Stockbridge Leslie Data Source A.E.1 CEN.2 A.E. CED. A.E. CEN. A.E. GEN. Animal Units 12.80 13.80 12.54 11.77 9.34 9.23 15.44 12.31 per 100 acres of farmland. Animal Units 17.18 18.80 15.77 16.58 16.93 17.20 19.77 18.00 per 100 til. acres 0 Productive days 2.93 3.10 3.28 2.74 2.94 2.84 3.00 2.63 work per til. acre 1. Agricultural Economics data (25). 2. Census of Agriculture, Minor Civil Divisions, 1945 (36). 83 Divisions tables of the 1945 Census of Agriculture for selected townships by natural land type (36). Comparative data are given from farm account- ing farm records of the Agricultural Economics Department of Michigan State College (25). It appears from both the census and farm accounting records that intensity, as measured by the number of animal units and productive days of work, varies from one land type to another. There appears to be about one more animal unit per 100 acreas of farmland on the Roxand and Leslie land types than the Riley and Stockbridge types. If tillable land is considered, the livestock intensity on the Stockbridge land type is con- siderably less than on the other three types. Productive days of work does not appear to vary significantly from one type to another. The dif- ferences between land types were not measured statistically for signifi- cant differences because of a lack of reliability and validity measure- ments for such data. The interpretation of Table XIV is a rather precarious undertaking. In the first place, the concepts of the animal unit and the productive man work units is not as rigid and reliable a concept as one might be led to believe. In the second place, the sample farms from which the samples were drawn were not the most reliable samples that could have been taken were the data available. Reliable information was not available in the sense that samples were too small in the case of the farm accounting farm data. In the case of the census data the townships selected only approxiwated the natural land types they were selected to represent. In general all the land types except the Stockbridge type probably have about the same intensity of use. If some better measures of B4 intensity were available for this investigation such as gross income, net income or amounts of physical output per year, the differences be- tween land types would probably assert themselves to a greater degree. ”ethods Used in_Determining_Intensity from the Census The method used for determining land use intensity from census data was worked out by Mr. John Doneth of the Hichigan State College Agricul- tural Economics Department and the author. It consisted of estimating the numbers of the various kinds of livestock per township and assigning a productive man work day factor and a productive animal unit factor to each grade of livestock. The number of animal units and productive days of work was calculated for each kind of crop according to data published by Hill and Doneth (18). The total number of animal units was calculated per township and per acre by land type. Similar calculations were made for crops. Total productive days' work per acre on crop and livestock were added together to obtain total productive days' work per acre. As a more specific example of how this process was used, suppose it is assumed that there was listed in the census for township X, a figure of 10,000 sheep. In the next column there was listed a figure of 2000 ewes and lambs kept for breeding. Although it did not specifically state that most of the other 8000 animals were lambs, this could probably be safely assumed. To obtain the animal units of sheep in the township, the figure of 2000 sheep for breeding was multiplied by the factor..14. This figure was then added to the figure obtained by multiplying 8000 lambs by the factor .05. The number of productive animal units of sheep in township X, therefore, was found to be 280 plus 400 or a total of 680 85 animal units. This number of animal units was added to the number for all other kinds of animals kept to determine the total number of animal units in the township. Although this method appears to measure intensity rather well, the specific reliability of the method needs further investi- gation. alleysgofCemusagai'easuregflgnigfibyflm One of the reasons that the natural land type was selected instead of soil type for this study was that it was felt that census data might be used as a measurement of land use by land type. In the scale of mapping used by Veatch (39) for delineating natural land types of Vichi- gan, it was possible on many of the more extensive types to find that entire townships were covered by one natural land type. This would never be possible in the case of the soil type. If a land type completely covered several townships in its occurrence throughout an area of several counties, it should be possible by the use of township data to character- ize land use by natural land type. Thus, if several township samples of land types could be found, it would be possible to measure the differ- ences is land use by land type. The idea was tested by selecting four groups of townships to corres- pond to natural land types 1, 2, 4 and 5. Table XV shows land use by land type obtained from the census data. P. W..A. data were not avail- able for 1944, the year on which the last census of agriculture was based. A devious method of comparison was used in Table XVI to show that census data probably wmuld agree well with PTA figures for land use, if 86 TABLE XV COMPARISON OF SELECTfiD TOWNSHIP LAND USE DATA FRO“ 1944 CENSUS OF AGRICULTURE AND 1948 PMA LAND USE DATA Source of Roxand Riley Stockbridge Leslie Data 1948 1944 1948 1944 1948 1944 1948 1944 PTA census PMA census PTA census PEA census Crop class Corn 20.3 20.2 18.2 19.0 18.9 19.2 19.9 15.4 Small grain 39.8 29.7 39.8 27.4 27.3 20.9 34.6 30.3 Tame hay 17.8 20.4 20.5 18.8 23.5 21.4 23.9 23.1 Plowed pasture 12.9 11.4 12.1 14.7 12.5 14.1 15.0 18.9 Cash crops 6.0 10.6 4.6 7.2 4.8 5.3 1.5 1.6 Other crops 2.9 7.5 4.5 12.7 13.7 18.7 4.8 2.7 TABLE XVI A COMPARISON OF PMA AND CENSUS DATA BY USING A THIRD MEASURE, AGRICULTURAL ECONOUICS DATA FROU'FARH ACCOUNTING FARWS 1944 Comparison 1948 Comparison Census Ag.Econ. Ag.Econ. Census Corn 19.0 24.0 22.0 18.2 Small grain 27.4 26.0 34.0 39.8 Tame hay 18.8 20.0 20.0 20.5 Plowed pasture 14.7 17.0 16.0 12.1 Cash crops 7.2 5.0 2.0 4.6 Other crops 12.? 8.0 6.0 4.5 88 the two sets could have been obtained in the same year. It is seen that in the census figures for 1944 and for PMA figures in 1948, each agree fairly well in their respective years with the figures Obtained on land use from an average of about 150 Farm Account farmers for Type of Farm- ing Area 5, the area covered by the Lansing region (9). Since things equal to the same thing are equal to each other, there is at least some agreement between PMA and census data for land use studies. There is a need for further testing of the use of census information for land use studies of a more or less broad nature. If it is found that census data are fairly reliable, then it would be possible to use the vast store of historical land use data available from the census for studies in land use trends by land types. The value of data such as these, if reliable, is obvious. Land Valuation As a Veasure of Intensity —_—————*—-‘- Land use intensity is the result of a historical process. For this reason it would be well to include investments per acre over a number of years. Unfortunately, census data appear: after testing to be of no help here. There is as much variation within land type as between land types. The only other source of investment figures on individual farms appears to be the tax rolls. While tax assessment figures often leave something to be desired as to accuracy, there is sufficient homogeneity of the data within land types and heterogeneity between land types to give some aid in investment studies. On four of the land types on which tax data were available, the valuation varied from a low of $25.58 on Stockbridge land to a high of $59.99 on Leslie type. Roxand and Riley types had valuation of $53.67 and $51.65 respectively (20). 89 Since a valuation figure on the tax roll in Michigan includes the total valuation of land and buildings for the farm, there is no way of knowing the valuation of land as separate from buildings. From observa- tion it appeared that quality of buildings followed the same trend as tax valuation per acre. Land Type 3 that was not included in this survey of valuations probably would fall next to the bottom in both valuation and quality of buildings. Land Productivity by Land Type The aim of the whole study included here was to study use and not productivity. Since, however, land use is partially a function of soil productivity, it was considered that the inclusion of such soil produc- tivity figures as were available would be proper. It should be stated here that such figures as could be found are probably not as accurate as the estimates that could and have been made by experienced agronomists. It is regrettable that there are not more soil productivity figures avail- able by either land type or soil type on which to base such studies as this. Unfortunately it is impossible to characterize yields by soil type in less than eight or ten years of study. Productivity figures were taken from three sources for the land types. Census figures for three census periods comprised one estimate (36). Figures from the records of Soil Science Department experimental fields comprised another (31) and the third source was from the FarnlAccounting farms of the area (25). Table XVII gives an estimate of the yields of the major crops by land type from each of the three sources. TABLE XVII SOIL PRODUCTIVITY 0N FOUR 33130-50 NATURAL LAND TYPES FROM .sRss DATA SOURCES Roxand Riley Stockbridge Leslie Corn for grain Census 32.4 31.1 31.7 31.3 (Bu.) Ag. Econ. 39.7 39.6 33.4 41.0 Soils Dept. 47.0 42.0 36.0 37.0 Corn for silage Census 8.8 7.9 7.8 8.3 CT°"’7 Ag. Econ. --- --- --- --- Soils Dept. --- --- --- -—- Cats (Dd) Census 34.3 30.5 25.1 24.0 Ag. Econ. 43.8 39.4 31.0 32.6 Soils Dept. 61.0 56.0 43.0 44.0 Wheat(80) Census 22.1 21.0 18.1 18.2 Ag. Econ. 25.8 26.1 19.1 19.2 Soils Dept. 28.8 29.0 27.0 28.0 Alfalfa CTN”) Census 1.5 1.4 1.3 1.5 Ag. Econ. 1.8 1.7 1.7 1.7 Soils Dept. 2.8 2.5 2.6 2.8 Red clover(TOOg) Census .8 .8 .7 .8 Ag. Econ. --- --- --- --- Soils Dept. 1.9 1.8 1.3 1.2 91 As a general productivity rating for the five land types studied, Roxand type is probably highest for cash crops and grain crops. Roxand is followed by Riley, Leslie, DuPlain and Stockbridge in that order. Hay yields seem.to be best on the Leslie type. Input-Output Relationships by Land Types Little experimental work has been done to characterize soil types and nothing to characterize land types as to the ability to absorb in- puts of labor'and capital. A few fertilizer studies have been conducted by not generally at more than two levels so that efficiency curves for the use of fertilizer have not been developed. Watson (42), Smith (30) and Cook (8) have studied several soils of Michigan with fertilizer input- output studies with certain special crops. No conclusions can be drawn for the land types in this study however insofar as physical capacity or efficiency of the land is concerned. If agricultural land use is ever to be seriously studied, there will have to be more attention given.to characterizing land by its physical input-output relationships. Changes in Land Use from Year 32 Year Nowhere in this study has there been a detailed history of land use on any of the natural land types. This is because there is an almost total lack of such land use history. The P. M..A. has history running back to 1946. The census of agriculture is available for the years 1935, 1940 and 1945 by minor civil divisions. Farm.accounting records are available on a few farms for the years 1929 through 1949. There are not 92 enough farm records available to characterize land use by years on the natural land types. In spite of the paucity of information, something should be said about changes in land use from year to year. Table XVIII is given to provide some insight into this aspect. This table is provided to show changes in land use on representative farm-accounting farms on four natural land types by two-year periods from 1930 through 1940. The pic- ture is given to ShOW’hOW land use changes over short periods on single farms. All farms selected were larger than 160 acres. Thus the oppor- tunity was available on each farm every year to grow all the crops in a rotation. About the only conclusion that can be drawn from.the'table is that land use changes considerably over short periods of time. Since only single farms are given, no land type comparisons can be made. If the crop, tame hay, is taken as an example of how greatly land use changes, it is seen that on the Roxand farm, the land in tame hay varied from a low of 21 per cent of the tillable land in hay in 1936 and 1940 to a high of 30 per cent in 1930, a change of 9 per cent. The Riley land type farm changed 16 per cent, from a low in 1930 of 8 per cent to a high in 1932 of 24 per cent. The Stockbridge type farm changes from 21 to 48 per cent over the same period and the Leslie type from 12 to 40 per cent for changes of 27 and 28 per cent respectively. These figures point out the difficulty of interpreting land use data for individual farms. In averages of several dozens of farms, the indi- viduality of land use on single farms is lost. 'While this points out trends in use on a single land type, it frequently leads the researcher to believe that the average represents every farm on the land type. TABLE XVIII CHANGES IN PER CENT 0F TILLAsLs LAND De'orsD T0 FIVE CROPS FROM 1930 TO 1940 ON EPRTSSENTATIVE FARMS" FOR FOUR SELECTED NATURAL LAND TYPES 93 1930 1932 1934 1936 1938 1940 Tame Hay Roxand 3O 27 29 21 26 21 Riley 8 24 10 10 11 19 Stockbridge 28 21 40 21 33 48 Leslie 15 20 19 40 12 36 Plowed Pasture Roxand 9 15 12 28 8 9 Riley 12 10 ' 21 10 19 10 Stockbridge 10 16 13 8 23 1 Leslie 21 8 l9 7 34 8 All Corn Roxand 22 28 25 20 15 23 Riley 12 10 17 13 17 16 Stockbridge 31 29 26 30' 17 30 Leslie ' 25 28 22 l6 13 16 Cats Roxand 19 18 18 17 10 11 Riley 0 0 10 24 7 24 Stockbridge 20 23 20 21 17 10 Leslie 9 8 10 8 11 23 Wheat Roxand 0 0 7 0 23 4 Riley 22 20 16 17 27 13 Stockbridge O 0 O 19 0 0 Leslie 14 14 7 21 13 ll * Michigan Farm.Accounting Farm Records, Agricultural Economics Department, Michigan State College, East Lansing, Michigan, unpublished (23). 94 Sumna of Statistical'Wbrk, Part II ____’2__ __ Agricultural land use from a sample of almost 1300 farms or Pro- duction Marketing Administration cooperators were studied to discover differences in land use between five selected natural land types in the Lansing region of Michigan. Within wholly agricultural areas of the land types studied, only the Stockbridge type of land had less than 99 per cent of the land in farms. Only 60 per cent of the land of this type was estimated to be in farms. The remaining 40 per cent was in idle land or state game areas. Seventy-five per cent of the land in farms was cultivated in 1948 on the Roxand, Riley and DuPlain land types. 0n the Stockbridge land type only 58 per cent of the farmland was tilled and on the Leslie type, only 68 per cent was under cultivation. The non-tillable land on all land types was devoted mainly to woodland, pasture and swamps. An analysis was made of the land tilled in 1948 on the five natural land types selected for twenty crops. Differences were discovered to exist between the land types for a considerable number of crops. The most often reported crops on all land types were corn for grain and silage, oats, wheat and tame hay. The twenty crops studied on the five selected natural land types were grOuped into six crop groups, corn, small grain, tame hay, plowed pasture, cash crops and other crops. Even with this general crop classi- fication, differences were found to exist between all natural land types considered. The effect of farm size on land use was studied. Land use was found to vary by size of farm. Differences in use due to size of farm were 95 less on any given land type than between use on the natural land types. Intensity of agricultural land use was studied. Animal units per 100 acres of cultivated land were calculated. Productive man work days per 100 acres of farmland and per 100 acres of tillable farmland were found. The use of census information as a tool in determining land use intensity was discussed. Land valuation on the various types of land was given and discussed. Few conclusions about the intensity data were drawn because of a lack of methods for evaluating the intensity measures used. Estimates of land productivity by land type for several crops using three sources of data g: included. Finally changes in land use from year to year on selected farms on four land types were given. The major conclusion that could be drawn from Part II is that land use varies significantly from one natural land type to another. This proves the hypothesis that the natural land type concept measures differ- ences in natural enwironment which within a given market environment will be reflected by differences in land use. Practically no conclusions about why land use varied between natural land types could be drawn from Part II. This was because of the lack of precise information on soil productivity and input-output relationships on the various land types. 96 PART III A DISCUSSION OF THE CAUSES OF DIFFERENCES IN LAND USE PATTERNS Introduction In the preceding sections, statistics were given for various aspects of land use on several selected natural land types of the Lansing region. The facts presented proved that land use varied from one natural land type to another. This proved the first hypothesis of the study, that differences in use were associated with differences in natural environ- ment as measured by natural land types. The second hypothesis given was if land use did in fact differ between natural land types, that a method could be developed to explain how these could have come about. In the statistical sections where facts about existing use were presented, little if anything about the reasons for an existing use could be discovered from the data presented. In other words, the statistics were not arranged or presented in such a way that cause and effect relationships could be discovered. To have drawn conclusions why land uses existed or how certain uses came about would necessarily have rested heavily on valuations and pre- conceived notions of the author. The facts are of such a nature that it is nearly impossible from observation to determine how or why land use AREA. in any particular«comes about. In studies of land use employing the statistical method, there are often blank spots in the information. That is, a complete picture of land use is not presented. Inferences about reality cannot be drawn without information from.which to draw them. For example, in this study, yields per acre were little better than guesses. Input-output relationships 97 on the land types were completely lacking and farm incomes could not even be approximated. It can readily be seen that with such lack of informa- tion, it would be virtually impossible to discover the reasons why land use differs between natural land types. In such a complicated field of study as land utilization, it would be an aid to research if some method or framework of study could be developed to study the cause and effect relationships which exist. One such system of study is available. This system is known as the analogy. If by use of certain assumptions, a system can be developed synthetically that is logical within itself and resembles the realities of the world, the hypothesis can be made that the theoretical system developed is the one that operates within the real world. This hypothesis can be tested experimentally and statistically. After trial and error with several such theoretical systems, one can probably be found that when tested fits most of the facts of the real world. When such a system is found, rapid strides forward become possible in research techniques. The following sections are an attempt to synthesize such a system. Only a beginning is made. If, however, only a part of the system holds when finally tested against the realities of the world, the effort will have been worthwhile. In the synthetic method used, the assumption is made that every farmer has perfect knowledge about his land and the price system in which he operates his business. Such physical realities as vagaries of the weather, and risk and uncertainty of the economic world are ignored. Institutional factors and other irrational behavior are ignored also. 98 In a complete explanation of land utilization within an area, all the realities of the real world would have to be included. For a begin- ning explanation, however, only the more important factors affecting land use should be included to gain clarity of the land use process with- out undue complexity in the explanation. After a beginning hypothesis has been developed and tested, then, other assumptions and refinements can be added to the system. These additions will not invalidate the beginning structure, if it has been tested, they will only alter and re- fine it. Throughout the sections on theory, one main assumption is made. This is, farmers attempt at all times to maximize profits and do this under conditions of perfect competition. A second assumption is made that only full-time farmers working on one—man farms are of any real im- portance in determining average farm size and land use wdthin an area. A third assumption states that conditions of production are static and knowledge of production is perfect. Further assumptions are included in the main body of the theory. Factors Affecting Farm Size with 3 Given Net Return for Labor and "anage- ment Per Acre Choice of size of farm is an interesting phenomenon. Although it varies considerably from farming system to farming system under different states of technology, there is one characteristic that appears in much of the world's agriculture. This characteristic is the association of family with farm. One man with the help of his family operates the farm business. A notable exception to this is found in collectivist societies. 99 In the United States, China and most other countries the family farm is the rule. In this section only one-man farms will be considered. The factors that determine net returns for labor and management per acre will not be discussed. The problem to be considered is, with a given net expected income per acre, why does the size of farm vary so much in a given farm- ing system area? Under a given proportion of factors of production in- cluding one man as the labor factor what will be the minimum and maximum size of farm.under a given farming system? If land use is assumed to be a rational procedure, the problem of finding the determinants of farm size in a given type of farming area can be approached with economic analysis. This rational process presup- poses that farm operators attempt to maximize net profits over the long run on their:farms both in the selection of farm enterprises and in their everyday farming operations. It is assumed that in the long run, econo- mic forces of competition and Igternative opportunities will eliminate those farmers who do not in fact maximize net returns. It is recognized that in the short nan, capital rationing, various institutional factors and combinations of circumstances wdll often operate to sustain those operators who operate their ferns at less than maximum net profit. A fundamental reason.why people farm can be shown with a system of indifference curves. Figure 17 shows an indifference map of a man faced with the decision of whether to farm or not to farm. For example, the indifference curve 11 shows all the combinations of dollars earned in a highest alternative occupation and dollars earned in farming to which the man would be indifferent. In other words, any lOO tf'}"!{rf l" INDIFFTRENCE AP OF A MAN FACED hITH THE DECISION ' ‘..~_ Y" .' I '3' IN"? K T.- ' ‘\| . f " : :' " "’\' r“ ' I.) (”nigh K'IX S‘w/‘uA L . :lli ’Li1r_,'.(:\. ,1 ,1)» r . ' i i\ J‘) LJL‘A rte) DOLLARS IR FALMIHG 101 point on the curve I is as satisfactory to the man as any other point. 1 The situation presents two commodities, dollars in one job and dollars in another job, that can be purchased with labor. Two more concepts are included in the Figure as constructed. The first is that each higher indifference curve is more desirable to reach than the previous one. 12 is a more desired position than 11 because of a psychological fact that most men prefer more money to less money. The second concept is that of increasing marginal disutility of labor as more work is done at any wage rate. This is shown by increasing in- tervals of space between indifference curves away from the origin 0. To review, each point on an indifferencc curve is equal in desir- ability to any other point on the same curve, a higher curve is to be desired over a lower one and there is an increasing physical and psycho- logical difficulty involved in reaching each higher curve. If a man is indifferent as to any point he may reach on any one curve but would rather reach a higher curve if he could, how does he decide what to do? Fortunately this is decided for him, not by what he wants to do but by what he can earn in each job for his labor. If the indifference curves as drawn are compared with the opportunities avail- able, the decision is made as to the best way to spend time working between the two jobs. If FG represents the opportunities available for buying dollars by spending labor, and PC is tangent to 12 at E, then the optimum way to spend time working is to earn 0D dollars farming and 0C dollars in the other job. That point 8 represents the point of maximum utility that can be reached with the conditions given can be explained by moving away from point 3 in either direction on line FG. If any other 102 point on line FG is selected than E, a lower indifference curve will be reached and less utility for the same expenditure than at 3 will be re- ceived. In the case of choosing between two occupations, the shape of the indifference curves may well be different from those drawn in Figure 17. Hicks (16) has pointed out that very probably the curves will be convex to the axes due to the principa of diminishing marginal substitutability. The only argument that this principle holds here is to say that a man would rather work at a variety of jobs if given free choice. This may or may not be true. Figure 18 shows an indifference map of a man drawn in such a way that for all but one condition of choice, namely, where the slope of the wage opportunity line is parallel to an indifference curve, a van will either farm full-time or work full-time in his highest alternative occupation. At the same time under the special conditions where the way opportunity line is parallel to the indifference curve, an opportunity for {art-time farming exists. Figure 18 shows with wage opportunity line AB that touches 12 at B that the individual for which this indifference map is drawn would rather earn zero dollars in a non-farm alternative job and OB dollars farming. For every person farming such.a decision as is discussed above is constantly before hinlat every change in wages in either farming or his highest alternative job and at eveny change in preferences, that is, his indifference curve. This idea can now be generalized to say that for all people presently farming, there are alternative occupations. The reason they are farming is that the alternative jobs are not sufficiently lucra- tive to induce them to change. It was conclusively shown during WOrld Lillllefilfin‘ IN ’ ”I ‘x‘v . '- “LIT 1:1." nT C. O C(‘UP A‘I‘IUN ° Fr" I .L 1 AW T‘Y? IYFLI' :v‘.‘-Y(‘1.‘ “. D erg-V“ ‘. \- *1.‘.' A '-' lifi\./._ AURI‘A' ‘..A :1 lA—L'Hak“ A'- vwv" T ”‘ .'-;vm -~vv rJAJL TAJ‘IL PAIL-u 5 J i/.L .4"th ’F u " . l v m "I . w.» \1 when; [BI FARE: ‘7/‘ uh) 103 105 war II that many farmers would leave the farm if alternative jobs became sufficiently enticing in relation to agriculture. The next important question to consider in.studying the theory of farm size is not one of logic but of belief. Does a man work for a daily wage or a yearly wage? If he works for a daily wage, the problem is one of finding his daily supply schedule of labor. In other words, what wage‘ will induce a man to work one more hour? This involves a choice of one more hour of working versus one hour of leisure. In agriculture, the choice is a free one; a man can loaf most of the time or work all the time, depending on his ambition. In industry the choice is more restricted. He can work eight hours a day or not work. Occasionally he may have the opportunity to work overtime. It is believed that a yearly wage determines whether a man will farm or'not. If an opportunity exists in an alternative occupation of earning $2500 per year at eight hours per day, the farmer will have to earn some comparable wage at comparable hours tosifford to stay on the farm. Let it be assumed that $2000 per’year at an average of nine hours per day will be sufficient incentive to keep the farmer on the farm. The number of acres of a given kind of land that will produce $2000 and still keep within the average of nine hours per day determines the minimum size of farm, If it is impossible to farm enough acres with short enough hours per day and high enough wages per year to compare with the highest alter- native occupation, the farmer will not farm. If for the moment it is assumed that the farmer can earn $50.00 per acre with nine hours or less effort per day, then the absolute minimum size of farm for this man would be forty acres. 106 The determinants of the maximum size of farm can be shown on Figure 19. As additional acres are worked by one man, after a point the law of diminishing physical productivity sets in so that each additional acre earns less than the previous one. At the same time, the leisure given up by working the additional acre becomes more and more valuable as each additional acre is worked. Finally a point Y is reached where the value of working the additional acre is equal to the value of the leisure time lost. This is the maximum size of farm under the given conditions exist- ing on the farm. For eveny farmer the minimum and maximum size of farm will be different. In general it is thought that since wage opportunities in alternative occupations are generally on the same wage level for un- skilled or semi-skilled people of the farm class and that ambition is probably distributed normally in the farm population, that for any given farming area, size of farm ought to be distributed much like the curve of ambition to work additional hours or to give up additional hours of leisure. This theory explains how it is possible for good agricultural land to remain idle several miles from town. There is simply not enough oppor- tunity wage from farming with the type of farming and the working condi- tions available for anyone to farm it. This applies also to poorer land much farther from town. The logic is this: if a man with ability enough to farm successfully in the outlying areas now moves near town, his opportunity wage in indus- tny increases faster than his opportunity wage in farming under comparable ‘working conditions; thus he cannot afford to farm. The opportunity wage in farming near towns used to be high enough to make farming worthwhile; FIGURES}? - ‘ “'35. l i" . . .p , I \1‘ 'r V .:-, v .A.d ‘ '6 107 '5 108 however, conditions in town are now SUCh that the opportunity wages are relatively higher than those in farming. This explains also why former- ly good farmers are now living on their farms and driving to town to work. At the same time, these men cannot find anyone to work their fully equipped farms. If they find a man with ability enough to farm, he, like his boss, can make more money working in town. If the farmer hires a man of low ability and low opportunity wage in town, he probably will not succeed in farming for the same reasons he does not succeed elsewhere. This theory also indicates why people of low natural ability migrate to poorer farming regions. Suppose there are two men, one of whom has high.ability with an alternative wage of $3000 per year, the other of whom is lower in ability and has an alternative wage of $1500 per year. Further assume there are two land types, one earning $30 per acre, the other, $15. Both men are willing to work ten hours per day. If the farmer of higher ability works the good land, he will have to work 100 acres. This he can do in ten hours or less per day. But to work the poor land would require a 200 acre farm, one too large for one man to farm. The man with lesser ability could farm the good land and have only a 50 acre farm to earn his alternative wage. He could also farm 100 acres of the poorer land, something the more capable'farmer could not afford to do. If the aSSImption that the good land earns $30 per acre and the poor land earns $15 is put in more realistic terms, it seems likely that on either grade of land the good farmer could earn more money per acre than the poor farmer. Thus he can afford to pay a higher rent or price per acre than the less fortunately endowed farmer. Thus he tends to acquire o. 109 more of the good land than the poor farmer for two reasons. The good man will not bid on the poor land and he can afford to pay a higher price for the good land than his competitors. This does not eliminate people of low ability from good farming areas, but it does tend to eliminate the better class farmers from poor land. It should be stated that if a good farmer does farm poor land, it is apt to be larger acreages than on better land with more labor and machinery. This explains the very large farms run by excellent farmers on poor land. If the land is poor enough, even the poor farmer cannot afford to farm there. In such cases the land is not farmed and may remain as idle farmland or go into some other use as forestry or wildlife reserves. In the preceding analysis, each individual had different indiffer- ence curves and different wage opportunities than every other individual. The same is true for any single person over a given period of time. Since the size of farm selected by any one individual to farm is chosen only once or twice in his lifetime, this question deserves some attention. The minimum size of farm is determined in the preceding analysis by dividing the alternative wage by the net earnings per acre. If then the 'working conditions on the farm are comparable with alternative working conditions and the marginal disutility of labor is low enough to allow production at all, the farm will be farmed. There are two essential factors to be studied in finding if different sizes of farms will be selected by the same persontlt different ages. The first factor is the net earnings per acre available to the individual. Figures from the Uetropolitan Lif: Insurance Company (24) show that farmers incomes do increase with age up to about age fifty. Part of this effect is due to llO added acres and part to added income per acre. The other factor to con- sider is the alternative wage per year available at different ages to the same individual. If a man is farming, his alternative wage will in- crease in an alternative occupation only as his productivity increases. Very probably not only as his productivity increases but as productivity for men of his age increases. Productivity per man in such skills as factory work increases until about age thirty-five and then decreases. An example of the minimum size of farm that a person at different ages would select is shown in Figure 20. If the alternative wage is divided by the net returns per'acre at any age, the minimum size of farm can.be found that that person would select. The size of farm that any individual would select would be dif- ferent from all other individuals. It is felt that there is probably enough similarity between individuals as to alternative wage and physi- cal productivity that some general curves could be drawn to show the minimum size of farm that persons at different ages would buy or rent. The maximum size of farm would also change with the age of the in- dividual entering farming. In addition to (net income) productivity per acre increasing with age as given above, probably there is a decrease in the marginal utility of leisure from age twenty-five to perhaps age fifty. After that added leisure becomes more desirable. These relation- ships are shown in Figure 21. In this table the maximum.number of acres that a man would work increases to age fifty and then decreases sharply. General Discussion 2f Factors Affecting Intensity In the previous section, average profit per acre was assumed to be due to some type of farming prevailing in an area. In this section the 111 v‘ :" If! 1"» . .:' .‘T:,"‘ \Q"" '7""" \“J 1., ‘/"' h" V" ' ’ 4' full“ [.m’lid :‘ ’a‘I‘_-‘-.{ 1.); j. t, L/L‘ALAFAJ ; Isll‘i 1.1 I, )‘u I. ‘ III i, 31 3 , DOD DOLLARS #0 20 30 [+3 F: i) ("l-Q) ACE OF INDIVIDUAL more. NET DOLLARS 5 112 I ,0 MAXIMUM SIZE OF mm encore). BY'-'iQRF‘ PRODUCTIVITY PER A 113 factors dealing with the farming types, that is, differences in farming due to differences in intensity of production will be discussed. To understand the changes in intensity from place to place and from farm to farm requires the knowledge of the nature of production, and particularly agricultural production. The principles of diminishing marginal substitution and of diminishing marginal productivity prevent all the food and other products of farms from.being grown adjacent to the areas of consumption. Von Thunen.and weber in particular have shown that there is a system to land and natural resource utilization. In general this system consists of producing products in one location and consuming them in another. Frequently the consumption location will be several thousand miles from the place of production. "Weber (43) showed that in the process of production and consumption that weight moved in production added to production cost while weight moved in consumption detracted from the price of the product at the place of production. Because of the law of diminishing physical productivity the total pro- duction that can be secured in any one area is ultimately limited. In the attempt to produce all that consumers want, production spreads out over space so that production cost constantly increases as a function of the cost per mile away from the market for production goods and that price received for the product decreases as distance from.market diminishes. Bach enterprise in production thus is forced to compete with every other enterprise so that each will be located at the point of greatest net economic advantage, that is, where the economic rent for any one enterprise is greater than for any competing enterprise. 114 If it were possible all production should take place in the same place as all consumption so that there would be no transportation costs. To put it another way: if there were no costs of transportation, pro- duction could take place wherever it suited the fancy of the people. Economists for years have shown that there are definite patterns of production. Von Thunen in 1826 showed this as a series of rings around the central marketplace laying on a broad homogeneous plain. By deduce tive analysis he showed that there are areas of greatest net economic advantage for each enterprise and that these are detennined by costs of transportation. Alfred Weber improved on Von Thunen's notion by includ- ing more variables in his analysis. He worked out mathematically a system showing that points of net economic advantage of production depend on the weight loss in processing a material for consumption. Aereboe and Brinkman worked more specifically with agriculture and advanced theories of broad land use based on the theories of their predecessors. In the United States, Gray (13), Taylor (33) and others at'Wisconsin did considerable work in the'broad theory of land economics. Other economists contributed along the way; Black (4), Bunce (7), Hoover (19), Hammer (14) and others made their contributions. Even with all this work, there remains practically no literature that deals specifically with causes for land use in local areas. In the pages following, an attempt is made to bring the contributions of past workers into a logical scheme to show why natural environment affects social and economic structure within a region. The first idea of importance in the study of land use intensity has probably been known for centuries. The idea is attributed to 115 "itscherlich (27) in Germany. This idea is the law of diminishing physical productivity. Baule (3) and Willcox (44) expanded on the law and gave it the graphic form of Figure 22, Baule was responsible for transposing the verbal principle into mathematical terms. Baule developed the so called "Baule unit" so that any attribute of plant growth that affected yield could be expressed mathematically in terms of physical effi- ciency. Baule derived from the experimental work of Nitscherlich the mathematical law of plant growth and found that if any element of growth such as nitrogen were studied, it was possible to find in a soil a minimum amount of the element that would produce the maximum amount of growth. Once the minimum amount necessary to produce the maximum growth were found, it was possible to divide the amount of growth factor into ten equal parts and assign a percentage of growth term to each of the ten units of the factor. For example, Baule found that following the asymp- totic growth.curve that the first of the ten.units of the growth factor (Baule units) would produce 50 per cent of the maximum growth possible for any plant with the element nitrogen. The second Baule unit would produce an additional 25 per cent of the maximum growth possible, or 75 per cent of the maximum growth. The third unit would add 12.5 per cent for a total of 87.5 per cent of maximum growth and so on. maximum growth for any plant could be experimentally determined. Uitscherlich found that the nitrogen content of a plant determined the maximum growth possible and that if a constant which had been experi- mentally determined for all plants were divided by the nitrogen content of a given plant that the maximum yield per acre could be determined. TH? LAB 0F PIMINIJHINU 130 75 F:U“RV 116 1-5;. DHV"~TI"‘.~‘ '1'. " "n'\”"‘" :’ ',."‘I’i * ~ imas ”I'll. f .-:'[u;i 1;- \. ni .l‘ f.11.'. ‘ A H V f E H17 ' 1%.) “f C\ {b UNITS OF IN? T Lu 117 This mathematical expression of physical efficiency made it possible to estimate yields of any plant if the nitrogen content of the plant and the Baule units of plant growth factors were known. The method for determining physical efficiency is the per cent of yield produced by each growth producing factor such as water, stand count, nitrogen, climate or phosphorus, (infitschlerlich claimed that any of these could be characterized by Baule units.) multiplied by each other. If there were 2 Baule units of Nitrogen in the soil, lO Raules of water and l Baule of potassium the per cent of maximum yield that would be produced under such conditions “muld be the percentage of yield from each factor multiplied by each other. In this case it would be 75 per cent times 100 per cent times 50 per cent or a total yield of 37.5 per cent of the maximum possible yield. If this were a crop of sugar beets with a maxi- mum yield of about eighty tons per acre according to Hitscherlich then .the yield on this field would be approximately 30 tons to the acre if all other growth factors were optimum. The above example is given to show that plant growth factors act in the same way in combination as economic efficiency factors as described by Black (4).“ Furthermore it is experimentally possible to characterize land in terms of Baule units of plant growth factors. Since this is true, there is some point to discussing physical efficiency and capacity of land. The work of Nitscherlich and Willcox is by no means ended. Bray (5) working at the Illinois station has recently employed a modified Baule system for predicting yields in Illinois. Although there is considerable controversary over the work of Mitscherlich and Willcox, there remains no question that the law of diminishing physical productivity acts in some way comparable to the description given by "itscherlich. 118 The law of diminishing physical productivity operates in other fields than plant growth. Spillman (51) working independently of ”itscherlich discovered the law about the same time in the United States. He found that the principle operated for animal growth, machine operation, human labor application and in many other activities. Economists have shown that there is a like law operating in the substitution of one factor of production for another. This is termed the law of diminishing marginal substitution. The two laws combine in practice to give rise to the gen- eral concept of diminishing marginal returns. In regard to characterizing land as a factor of production, Black (4) and Hammer (15) at least have applied the law of diminishing physical productivity to land and have concluded that insofar as physical capacity and efficiency are concerned, that there exists four kinds of land. The four kinds are land with high capacity, low efficiency; land with high capacity, high efficiency; land with low capacity, low efficiency and land with low capacity and high efficiency. These four types of land are shown in Figure 23. The construction of the figure should be explained. Capacity is measured horizontally. It is the total physical ability of land to return a yield for inputs to land of capital and labor. The table is based on the known fact that some soils have greater ability than others to absorb inputs. Efficiency as a general term is the ratio of outputs per input. Narginal efficiency is measured vertically in the figure, it is the marginal output per mar- ginal input. If inputs are considered as completely divisible, a line drawn vertically from the horizontal axis at any input gives the marginal output for that unit of input. Total yield can be determined by adding OUTPUT OR USE ELASTICITY CURVES . FIGURE 23 119 FOR FOUR THEORETICAL TYPES OF LAND ( AFTER BLACK.AND HAEMAR ) EFFICIENCY EEFICIENCX High efficiency High capacity &:, CAPACITY Low efficiency High capacity CAPACITY CAPACITY High efficiency Low capacity L: CAPACITY Low efficiency Low capacity 120 the marginal increments of outputs to the left on the diagram plus the increment of the input in question. The curves in Figure 23 are derived from.the work of Mitscherlich mentioned before. A term that can be applied to such curves is use elasticity curve or input elasticity curve; the rate of change of marginal output in response to a small change in input. For every piece of natural land, every soil type or every soil type phase, there are literally dozens of physical use elasticity curves. There is a different curve for each use on each land type and a different Curve for every association of uses on a given natiral land type. Thus a soil might have high efficiency--high capacity for beans in one farming system but low efficiency-—high capacity for the same crop under a differ- ent rotation system. Generally, however, soil types are considered to retain their capacity-efficiency relationships over a considerable range of land use. Thus muck soils are considered to have high capacity--high efficiency for a considerable variety of vegetable crops. These same soils have low efficiency--high capacity for timber crops. Economics cannot change the capacity of land. This is because capacity is a physical concept. It may never pay to expand production to the physical capacity, nevertheless, the opportunity for physical ex- pansion exists. Black (4) has a different concept of economic capacity. He clains that economic capacity is the sum of all marginal positive economic efficiencies, that is the highest profit combination. If Black is correct, and if his is the right concept, it is not a very useful con- cept, because, for every change in price of either input or output, a different capacity of land results. Since capacity is generally a measure 121 of size, it might be more useful to think of capacity as the ultimate physical yield possible. Then, for any ratio of output-input prices, at any positive marginal physical output, there will be some positive rate of return on input. This positive rate may be ten cents return on ten dollars input, it is a positive rate of return, nevertheless. Only when the marginal physical output is zero or when one of the prices of either output or input is zero will the rate of return on input be zero and the physical and economic efficiencies differ. In the former case when the marginal physical yield is zero, the physical capacity is reached. In the latter case with one of the prices of either input or output is zero a non-economic situation is the result. Thus physical and economic capacity are always measured at the same point. Economic efficiency, on the other hand, is a different concept than physical efficiency. It is measured by economic output per economic in- put. Uarginal economic efficiency is calculated by multiplying the mar- ginal physical output by the price of output. The figure obtained is divided by the marginal physical input times the price of input. For example, to show the concepts of capacity and efficiency that have been discussed, suppose that land type A has a certain input elastic- ity curve for-growing potatoes (”N). If the ratio of prices is l bushel of output equal in value to one unit of input, then a price line AB1 can be drawn on Figure 24 to indicate the point B1 where the value of the marginal output B101 is equal to the value of the marginal input Cl, that is, the highest profit combination.under free competition. Thus, by mar- ginal analysis using the physical use elasticity curve and the ratio of prices of output to input, it can be seen that CCl inputs will be used ID"? "I; I v 9 A 1‘. {FV\V"F‘ I u .1 I‘ll U.) a CA L Yul i r’: Ofi ‘e‘J P‘LGEIEJ‘L 4L, Pl.."~.i3TliJlT'Y CIR-NJ; F31”. 'r‘DTuTCb.) Oil uT l.. i. TYPE? A. Price I‘_.-t‘l.O 4.1110, out; 1:1 \ Lzl .x' \ - \ f“, ? ’ ‘. x, q I T Y‘J'T'T‘C iN. .. 1 123 and that the total yield of potatoes on land type A will be the area OUBlCl at the highest profit combination. The usefulness of plotting ratios of prices on a physical marginal productivity curve is immediately evident. If, instead of a ratio of prices of output to input of 1:1, there is a ratio of 2:1, the new price line will be 082 and input use will increase to 0C2 and total production to the area OVBZCZ. Since curves for land such as given in Figure 24 do exist for land types and soil types, it would be well to investigate some theoretical aspects of use and particularly of soil and land productivity. Suppose that there exists in a region in a virgin condition, three land types. The first is composed of soil type A with use elasticity curve AA' in Figure 25. In the same region there exists soil type B with curve AB' on the figure. In the same region there is a land complex com- posed of 50% of soil A and 50% of B in such a complex mixture that the soil types cannot be separated for use by the farmer. This complex can be termed A/h and its use elasticity curve will be AC' or the average of AB' plus AA'. If the ratio of pricefl of output to input were the line “N, the amount of input used on land type B would be OP, on land type A/B, OQ and on land type A, OR. If production were to continue for perhaps fifty years on virgin land it is unlikely that the component A of land type A will remain the same as A of land type A/B, even though in the original state they were identical components. This is due to the fact that com- ponent A of land type A receives more inputs every year per acre than A of land type A/B. PHYSICAL OUTPUT PLR INPUT 124 FY”UZH ”r Luflwin’. I». ) I "' ‘“" \‘Wfi‘v f‘Tm '\| . -- y. 1.- m don handlltlll tantao P)” Di »T”‘ .‘- A .‘i . ‘L FOR EACH OF Tfihhh NATWVAL LAHZ TYPES {J A Y317'N A! INPUTS 125 This analysis of the components of land types according to the manner in which land use takes place provides a logical method for conducting soil productivity studies. It can be stated rather definitely that land used such as A/b in the example given, for several decades is not the sum of the weighted averages of land type A and land type B. It is as illogi- cal to believe that economics plays no part in soil productivity studies as it is to believe that soil type or land type plays no part. The conclusion that follows from this is: soil productivity studies must be based on fide'units of land that farmers actually use in applying inputs if research in this field is going to be realistic. The next important step to consider the land has been characterized is to review the work of Von Thunen and Weber (41) (43) in regard to land use intensity. Von Thunen worked out in his "Isolated State" the nature of the geographic incidence of land uses around a central market city. Weber has pointed out that Von Thunen was essentially correct in his analy- sis of the distribution of agricultural production except he forgot one factor. This factor was the cost of transportation of inputs to the farm. Von Thunen had considered only fine marketing cost of transportation in his analysis. weber barely mentioned this omission of Von Thunen's; he did not elaborate. If the same assumptions are employed that Von Thunen used, that of a city in the center of a homogeneous plain, it may be possible to add something to Von Thunen's work. Perhaps it would be well to start with one product such as fluid milk, assume a market demand schedule and consider how intensity of pro- duction would vary with some factors affecting intensity. Later the 126 limitation of one product will be eliminated to include many products. First, it will be necessary to characterize the land for milk pro- duction. This means that a use elasticity curve for fluid milk produc- tion at the highest profit combination on the land used will have to be assumed. There are a number of variables to consider in this problem of vary- ing intensity besides the use elasticity curve of land. These are: the prices of input and output, transportation cost to and from the farm and the demand schedule for milk in the market place. Other variables will be added later in the analysis. The effect of these variables can be shown graphically. Figure 26 shows the effect of price of input and output on intensity of land use. If a use elasticity curve is given and price ratio lines are drawn on the graph, the relationship can be seen clearly. A given relationship of output to input exists physically as is shown on the curve. At a low price of output per input, production will be at 0A inputs with an output equal to the area ORSA. With a higher price, inputs will increase to CB and output from the area ORSA to ORTB, an increase of ASTB. Finally with a very high price for output in rela- tion to input, OC inputs will be used, causing an increase in production of output of the area TBUC or a total area of ORUC. hfith any curve where diminishing physical productivity is the rule, there is a point of maximum.physical returns, 00, where regardless of the ratio of prices of outputs to inputs, no further production can take place even though inputs can increase. PHYJICAL OUTPUT PER INPUT FIUUhh £5 a? lib FLTld a? JVTrVT TO IN ”T fltICF 127 ON THE INT SSITY wt unfit U31 3:1 Ratio ii 1:1 T ‘- {:1 l, A r f D IN” :m‘» s v' A p . Q ~ \ I A U ,_ mi. I 1 VP; (..A \ .1 e .. if: 3 in. N e n . T.. h, . . T4. . I rL .I ..H .rt . . . . - - .Y t : vi r . I Ail a ~.\ N ‘5' L .u .. V\J r . . \ \ ,. v!. I: h . h. . x Y} A\ 'I 1‘4 ll \V l a h b 11,0 » ¢ ’1 4 l- in h a . 4 i4 vha L .J V fir \(nv A .\ 1. lb "It; . ..\YJ h, \nlb /\ Ti \) . \ "it .i‘ . v. .. \ ... a“ ffi c 1.‘ A ‘,L a yr TIA ..0 ML . L P v. \s ,.5 PI?) n1 1,11‘5 A<..J;* 4-J\411~”§H #511 1.. i... >irL.... . a .. " r < f Figure 28 shows that where input involves bying and hauling equip— ment from market such as feed, fertilizer and machinery and even food for the family, that ultimately there would be a limit to production spatially, even if it cost nothing to carry the produce back to town to sell. The effect will be the same as transportation cost of input, but more pronounced generally since most inputs are derived at the point of production and are not hauled to the farm; whereas all the output is hauled from the farm. In other words, even if input transportation cost were the same as output transport cost per unit of weight, the effect on the area of production would be much less from input (an increase in the cost of production) than to output transport cost (a decrease in the selling price of the product). This would be true in most cases of agricultural production, at least. When transportation costs change, they change in the same way for both outputs and inputs. This produces an accentuated effect on the intensity of production in a given place. The changes in intensity asso- ciated with both changes in output and input transportation cost at the same time are shown in Figure 29. These examples are not the only factors affecting intensity of pro— duction. Others such as the personal qualities of the entrepreneur, “and“, :cd institutional factors of production, irrational behavior, " changes and other factors could be included. Enough of the important factors affecting intensity have been given to understand the nature of the problem. THE EFFECT OF A CHANGE IN OUTPUT ThnhLFQthTlWH Casi CH TH? IHTTNSIIY SF LAN; fink a" MILES FROM lhhAET 131 FIG’JRE_;29,L,H- - I “first EFFECT or A CHANGE IN BOTH wreath? r .3": ‘I‘; a l 3' ON THE INTEJSITX, swings ' .2 ~' ‘ Va. _ O 3_ «4"; .. A Imn‘gsfl‘i' ‘ k , , PBS. :4 132 The Effect of Distance from Varket on Farming Systems Essentially there are two elements in calculating where production of farm products will take place. The first lies in discovering physical characteristics of the land, its capacity and efficiency, under all pro- duction systems physically possible. Secondly the economics of production must be known. It is the interaction of physical characteristics of production with the economic characteristics that produce apparently hap- hazard land use patterns. If a theoretical example is taken, the factors that determine the location of various land uses can be seen more clearly. In discovering where production will take place, it is necessary to know the physical limitations of land for all possible agricultural products. Let it be assumed that there is a.city lying in the center of a homogeneous plain and that the land of the plain has the characteristics for production of its three possible products, sugar beets, dry beans and barley as given in TablelQflIat the various levels of a variable input, in this case fertilizer. It should be pointed out that if the land had either different pos- sible uses or different use elasticity elasticities than given in Table XXII, that this would affect the whole land use picture in this problem. After all characteristics of physical use are known, the answer to the problem of location of agricultural production lies in economic con- siderations. All the economic facts that need to be known are given in Tables XXIII, XXIV, and XXV. The implicit assumption is made in the tables that capital cannot be substituted for labor. This is not true, of course, but can be assumed in this problem. TABLE XXII THE EFFECT OF APPLICATIO"T OF A VARIABLE INPUT (FERTILIZ R) ON THE YIELD OF SUGAR DRY BEAVS A30 BARLEY ON HYPOTHETICAL NATURAL LAND TYPfl fl 3.: 133 Yield Amount of fertilizer Sugar Bee Dry Beans Barley pounds pounds pounds pounds 0 10,000 1,000 800 100 12,000 1,300 1,300 200 13,800 1,500 1,750 300 15,300 1,600 2,150 400 16,500 1,650 2,450 500 17,400 1,650 2,650 600 18,000 2,750 700 18,300 2,800 800 18,500 2,800 900 18,600 1000 18,600 134 TABLE XXIII FACTOR COSTS AT THfl FARE AT VARIOUS DISTAFCES FROM THE CITY Miles Cost of fertilizer Cost per hour Cost of per 100 pounds or labor Vachinery O 3.00 1.00 1.00 10 3.10 .83 1.05 20 3.20 .67 1.10 30 3.30 .50 1.15 40 3.40 .50 1.20 50 3.50 .50 1.25 60 3.60 .50 1.30 70 3.70 .50 1.35 80 3.80 .50 1.40 90 3.90 .50 1.45 100 4.00 .50 1.50 135 TABLE XXIV COSTS OF PRODUCTION IN DOLLARS AT ZERO DISTANCE FROW THE CITY Crop Fertilizer Labor Machinery Sugar Beets 21.00 60.00 40.00 Dry Beans 12.00 20.00 20.00 Barley 18.00 10.00 20.00 136 TABLE XXV COVNODITY PRICES AT THE PARK AT VARIOUS DISTANCES FROH THE CITY FOR SUGAR BEETS, BEATS AND BARLEY Crop _— Distance Sugar beets Dry beans Barley in miles Cents per lb. Cents per lb. Cents per 1b. 0 1.2 6.0 4.0 10 1.1 5.9 3.9 20 1.0 5.8 3.8 30 0.9 5.7 3.7 40 0.8 5.6 3.6 50 0.7 5.5 3.5 60 0.6 5.4 3.4 70 0.5 5.3 3.3 80 0.4 5.2 3.2 90 0.3 5.1 3.1 100 0.2 5.0 3.0 137 The price of labor was assumed to vary as the cost of transporting a worker to town every day of the year out to a point where the wages per hour became so low as to cause a worker to move into town rather than to drive to town each day. The point at which a worker would move to town was assumed to be the floor for agricultural wages for the manager- worker combination found on a one-man farm. The cost of capital goods was assumed to increase for'at least two reasons. There is a transport cost in moving production goods such as tractors and tile. Secondly the interest charges on capital itself would increase with distance due to an increasing scarcity of capital for production purposes. Lastly, the price of fertilizer increased in price according to a set transporta- tion cost and product prices decreased for the same reason. With the information at hand, the location of production can be discovered. The yield of a crop on a given soil is the function of the amount of variable inputs such as fertilizer, increased tillage or more intensive spraying, if all their conditions for growth are optimum. In this problem the only variable input included is fertilizer. Von Thunen listed two costs in the production of rye, that dependent on the price of rye and that independent of the price of rye. In more modern terms these would be variable costs and fixed costs. An economic principle which states that production will be pushed to the point where the marginal unit of the last variable input is equal to the value of the marginal increase in yield from the variable input, is the determinant of the level of production that will be used on any given soil. As an example, if barley sells for four cents per pound and ferti- lizer costs $3.00 per input, that is 100 pounds, it will take 75 pounds 138 of barley to pay for the fertilizer. In the example given, the sixth bag of fertilizer increases yield by 100 pounds and the seventh by 50 lbs. The sixth unit would be used rather than the seventh and if sacks of fertilizer could be split up, a little over 600 pounds of fertilizer would be used. Thus 2750 pounds of barley would be produced per acre. It is the variable input that determines the optimum point of pro- duction. The fixed inputs when added to the variables determine whether a farmer should attempt production or not. If when the variable costs .and fixed costs are added together and the sum is less than or equal to the total revenue from selling the product, production should be attempted if the use in mind is the highest economic use. To recapitulate, there exist in the central market, prices for fact- ors of farm production and prices for farm products, which are determined by forces of supply and demand. There is a price for transporting goods. All these factors interact with the physical characteristics of the pre- vailing land type to give rise to an economic rent at any point selected. Figure 30 gives all the economic rents possible for sugar beet, dry bean, and barley production for all points from zero to 100 miles from the central market. In the example chosen, all crops could be economically grown at all points included on the chart. However, the one that would be grown is determined by the law of comparative advantage. This prin- ciple says that the use producing the greatest economic rent per acre will capture use of the land. In the example given sugar beets would occupy distance 0A. Point A would be the margin of transference to barley. Barley will have the use of land from A to B and from B beans will occupy the land to some next margin of transference. [00" so~ 60* qo-« T? ,‘VV”"'I .30 1 Inuit; _j MALGINS OF Tn JSFhBLHCEQFOH THREE ChOVS KITEi DISTJILCE FEOLI 3:121}:an IN A EIIPOTEKETICAL SITUATION SUGAR ‘V SUGAR AR EA SEETS lfiEET Glljew'zn (3 -_-_ -.-.1. --_..-.-..1-.- .-._ __.--...’ 61Qo AR __L WIN/'1 EA 3.4 C A ‘0 § y..._....1. . ORV BEA N AREIA 65 7o tthKET 140 Factors Affecting Enterprises that Will be Conducted on a Farm —--*-—--——-_ The analysis of the best combination of inputs to result in the highest profit can be found in any textbook of Production Economics. The purpose here is not to solve this problem but rather to indicate how certain enterprises tend to be used by a farmer. The size of farm and habits of farming were originally determined in any area by a combination of factors. Among these were the opportunity for selling farm products, the price at which they could be sold, the products that could be produced and their costs of production and finally, the Physical and mental capacity of the farmer and his alternative wage. It can be assumed here then that the size of farm has been deter- mined by historical events. If the size of farm is taken as given for any particular short-time production period and it is assumed that one man is available for labor and management, the factors which determine “mat enterprises will be undertaken can be studied. With the amount of land and labor given, the problem of the best combination of inputs and what inputs to combine can be found by discover- ing the combination that gives the greatest net profit per acre. This combination can only be found by analyzing all available combinations of enterprises. As an example, suppose the opportunities for growing only wheat, cats or red clover seed as cash crops exist on a farm. What are the fac- tors that determine which of these crops will be grown and in what pro- portions? It is known that crop yields vary with crop rotation or crop sequence on any individual natural land type. All possible sequences of crops are given in Table XIX for the three crops: wheat, oats and red CLo¢¢8. ALL POSSIBLE CROP TABL E XIX 141 SBQUTVCES WITH EXPECTED YIELDS ON A HYPCTEETICAL FAR? Crop Yield 'Wheat alone 12 Cats alone 15 Clover seed alone 100 Wheat 15 plus Oats 20 Wheat 25 plus Clover 200 Cats 35 plus Clover 200 Wheat 20 plus Oats 25 plus Clover 150 Oats 30 plus Wheat 18 plus Clover 150 142 With the yields given in Table XIX, if costs of production were as follows: 0ats--320.00 per A., Wheat-~320.00 per A. and Clover seed-- $20.00 per A., if grown alone and $10 per A. if grown with a grain crop and.....Suppose wheat were $2.00 a bushel, oats, $1.00/bu. and red clover seed, $.30 a pound, which combination would the farmer use? Wheat-red clover seed with $40 per year net profit would be the win- ning combination for this farm and half the farm would be devoted to wheat and half to red clover. This problem can be sunwarized to state that the enterprises grown on this farm depend upon the price of input, the price of output and the physical productivity of soil. All the crop sequences are for static production conditions. If this limitation is eliminated, hOW'will this change the crops grown? If only one example is chosen, for example the crop sequence wheat-red clover seed, how will the adlition of fertilizer affect the cost combi- nations and how much fertilizer will the farmer employ at the highest profit combination? Table XXI gives the yields the farmer could expect for each increment of fertilizer added. The most profitable amount of this fertilizer would have been 500 pounds per acre. If the complete analysis were done, the effect of ferti- lizer on all crop sequences would have to be studied and perhaps the highest profit combination.would then not have been the wheat-clover se- quence. Such analysis could be carried out for use of any other input such as irrigation water or additional tillage. The final combination of enterprises then is determined by the cost of inputs, the price of output and the physical productivity of soil at each input level plus the farmer's problem in allotting his labor supply between enterprises and using his crop rotation scheme. TABLE XX PROFIT COWBINATIONS WITH VARIOVS CROPPIKG SYSTE”S ON A HYPC'I‘ET‘ZTI CAL FARM 143 Net Profit/Year Wll e at Cats Clover ‘f-Jh eat Wh 8 a t Oats fiheat 12 bu. 15 bu. 100 lbs. $30 ~20 -20 plus plus plus plus plus oats ClOVer clover oats Wheat : $12 2 ' 3‘80 2 ' 935 2 plus clover $45 Plus clover -10 $35 :45 ~10 {35 $24 — 20 = $4.00 :15 - 20 = -5.00 $30 - 20 = 10.00 3 5.00 $40.00 ”32.50 360 = $20.00 "'7? {61 : $20.33 3 144 .aoNWHfipaou mo mmcsom 00m scapmcHQFoo pwmoum pmogmHm * 0m.m> mvfi baa moH mm we own He w *om.m> swa sma mcH mm ow 0mm He m oo.ms wea wma moH mm an own aw m 00.05 owa mua ma Ow on 0mm oe w Om.mw smH Hwa pm we ¢n 0mm an m Om.nm 50H mma we we mm 0mm mm H oo.oew 0m OHH om om on 00m mm o R w m Hence w meow w pdoxfi w poem mpmoo Hapoe pobofio w m ao>oflo wow two» you was“? was pwoafi mensom wHogmsm memo oped nod oscobom opo¢ pom oped pom voom noboflo asofifi Lomfififlppoh pwmoam pop proe ossobom Hdpoa pmoo Hapoe new we @How% we ufloww mo mesmcH . .I ii 0.! LE; Q¢oHemn§om>m 4. 2.0 Qmm‘m mmbugo amalefimm .. ho moumtcmw @039 < 20 AlmNHAHHmwev thLH mAm