| Hi lh ll fl 110 im 684 oT HS im j aM 4} a Y dato ye ro i THE RELATION BETWEEN ‘ i RL eee OSS ul ! PORTS NTE TLA? | Waki SNe | | aaa saa . RST GE Gn chia 2 1S Bas) P. a: at a ta ee ce OO nO, a a emai ie THESIS y rc ‘see C. ~ ‘ee Le PU oO al © Keceveeln, PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE | DATE DUE DATE DUE DEC21:8 004 6/01 c:/CIRC/DateDue.p65-p. 15 THE RELATION BETWEEN THE CHEMICAL COMPOSITION OF THE SOIL, AND ITS PRODUCTIVITY thesis for Degree of X.S. Oscur Kdwin harrington 1916 TIYE SIS Tuble of Contents. Page. Introduction------------------+------- -- - + ee ee eee 1 Keview of Literature ----------------------- ee ee ee 4. Description of the Areas Studied -------------------- 12 Description of Samples Used ------------------------- 14 Ixperimental e----e------ ee ee eee ee ee ee ee ee Le SUMMNATY ence ee eee ewww mee meee ee eee ewe eee ee eee ee 45 General Conclusions ------------ eee ne eee meee ee eee wee 47 Bibliography OE505 1. Introduction. There has been a considerable amount of discussion during the past few years as to the value of a chemical analysis of a soil as a means of studying its producing power.Many opinions have been ventured,a limited amount of data has been compiled,and conclusions have been drawn which vary between two extremes:tnre one held in the past by the United Stutes Bureau of Soils that the productivity of soil is minly dependent upon its physical condition,water supply,etc.,(23) the chemical condition as related to plants being essentizlly the same in all soils;the other held by Hilgara(6),late of California, Hopkins,etal,maintaining that the productivity of a soil is almost directly proportional to the amount of plant that it contains. The effect of cropping upon the chemical composition of a soilis also of considerable interest to both the theoretical and applied phases of soil investigations. To what extent the total ampunt,as well as the amount of readily available,nitrogen, phosphorous, potassium, sulfur and calcium are decreased or increased by cropping and in what way,if in any,the acidity of a soil is affected by cropping ,are questions of interest and practical value. Each of these questions has been dealt with in this paper. There has also been an effort made to touch upon the variations in the relation between composition and productivity »caused by differences in type.Certain more or less arbitrary standards have been suggested by various workers ,notably that of Maerckerof Halle Station,Germany, and that of Hilgard.These standards are of a limited value, 26 however,because of their failure to ~ze into proper consideration variations in composition,texture,and structure due entirely to type. Itis hoped that this paper may throw some additional light,of interest and value,upon this phase of the work. The work reported in this paper includes both field and laboratory stuviies of the soils under consideratim. It was realized at the beginning that laboratory studies would be of little or no significance ,unless as exhaustive knowledge as possible was obtained concerning the field conditionse The field studies include a consideration of the climate,geolory,physiography,cropping systems,and soil manarement in all of the areas sampled,as these factors all have a direct bearing upon the solubility of the plant food and consequent loss by percolation,removal by erosion, crops,and the like.They also inciude a careful considera- tion of the physical conditions of the soil as atfecting bacteriz«l and chemical activity and plant growth. The laboratory studies include a determination of the total amount of nitrogen,phosphorous, potassium, sulfur,and calcium supplemented by a determination of tne readily available phosphorous and potassium in at least one sample of each type.The lime requirement of each soil was also determined. Finally ,an effort has been made to correlate these laboratory and field studies,and to compare them with the producing power of the soil. It is believed that such a correlation is the most important part of a line of work of this nature,and as such correlations have not been extensively reported in the literature,it is noped that they may prove to be of a certain interest and value. 5 4. _Review of Literature. For purposes of review,the literature has been divided into three groups:that dealing with soil depletion by cropping,the factors affecting the availability and solubility of plant food,and finally miscellaneous articles. the literature which bears directly upon the matter of soil depletion by cropping is rather Limited. In\Snyder (1)reported the results of his investigations upon some soils of Minnesota.He determined the amount of nitrogen, phosphorous, potassium,sulfur,and calcium,and the readily available phospnorous in both cropped and uncro:red soils. The greatest decrease was in the nitrogen content which was about 40% during the period of cropping.Similar but smaller decreases were found in all the other elements. The amount of soluble phosphorous was noticeable in the presence of a large amount of organic matter.There was often a slight gain of calcium in the sub-soil of the cro: ped soil,due to percolation.The length of time the solls were cro:rped was not given. VanSlyke discusses in his text(2) various sources of loss of piant food from the soil,and gives in the following table the approximate amounts of nitrogen, potassium, and vhosphorous removed per acre by corn,wheat, and oats wnen yielding 25 bue each per acre: Crop Lbs.Ni. LbseP. LbseK Corn 39.2 13.8 27.6 Wheat 42.5 16.6 21.0 Oats 24.0 9.0 20.4 be He summarizes the factors governing the loss by leaching as rainfall,adsorbing and holding power of the soil,organic ma'ter,and tne form of ‘.1° plant food, King reports (7) in 1905 the results of some pot experiments conducted to show the depletion of plant food by crovping;ske found the available potash and humus to be depleted the most rapidly. In 1911 Hart and Peterson (4) reported the results of their ivvestigations into the relat- ion of sulfur to crop production and its loss from the scil due to cro;:~ing.Tney Bound that the cereals and grasses remove about 2/3 as much sulfur from the soil as phosphorous and thet legumes remove more sulfur than phosphorous. They also found that continuous cropping for a period of fifty or sixty years reduced the sulfur content by about 40%. From these results they concluded that the sulfur content of the soil ,the same as that of phosphorous,might become dangerously low,and might become an important factor in soil fertility work.A similar work was reported in 191% by Shedd (5).He had worked on some Kentucky soils wnich showed the sulfur content to have been depleted as much as the phosphorous.The per cent loss of tne sulfur varied from 3% to 407,and that of the pliiosphorous from 5% to 40%. An important of exch of the two last named articles is the high loss of sulfur noted.qris point will be dealt with more extensively later. The literature having to do with tne various factors and conditions which affect the solubility of the mineral plant food is somewhat more extensive.One of the earliest 6. investigutions is that of Kilgard (6).His work wi: general and very extensive,end he drew tie following conclusions: 1.Al]1 vircin soils having high per cents of ylant are highly productive unless physical conditicons are adverse. 2.The reverse is not always true because some soils having Low amounts of plan® food are highly productive and are lasting.Therefore unalysis alone cannot be relied upon to determine crop production.3.Tne coarse textured soils seem to be able to produce with lower zcmounts of plant food than the finer textured ones.This he believes to be due to tne fact that in coarse textured soils the plants could aevelope a more extensive root system than in the fine soils and therefore come in contact with move plant f od. 4. He believes that lime is often a controlling factor in the producing yower of a soil;that in very sandy soils o.l. is often enough,while in heavy clays 0.6% is often inudeyuate. A very considerable amount of work has been done on the factors influencing the solubility of prosphatic comzounds.Most of the studies have been on pnosphatic rocks whicn are used as fertilizers,but the facts thus obtained should a ply well tothe phosphatic minerals of the soil. Soffre (7) in 1096 tested the solubility of tricalcium phosphate and apatite in pure water and in water charged with carbon dioxide. Ee found thet the water churged with carbon dioxide dissolved on an averasce of fifteen times as much of the tricalcium phos;rhate and eight times as much of the apatite as the pure water,thus emphasizing the role of carbon dioxide in the soil in connection with ve the availability of phosphatic minerals. Statstrom (3c) in 1904 Obt:ined results with humus,sour milk,etc.,which also point to the solvent ac ion of cur: ondicxide wnen Gisscolved in water. in 1°0O'' Cameron and Be?1 (9) published what was largely a review of much of the work previously done on the solubilit:, of rhnos:rhates. This included the work of Maley and Donath,Wurrington,Erlenmeyer,Terril, and otnerseAll of the results by tne authors in this vaper, like the ones mentioned above,bring out clearly the value of carcon dioxide as a solvent of soil minerals. Perotti (11) concludes from his investigations with different s>5il organisms la culture media containing nitrogen as asparagin, tartrates,sulpnates, and urea, that in media containing a physiclogically acid salt there is a uniformly impoftant solution of tne phosphutes. Likewise,Sackett,Patten,and Trown (12) concluded in 1°CS that when baectertal growth i8 accompanied by acid formation,there is a decided solution of the insoluble phosphates and that the presence of carbo- hydrates increases the solvent acticn because of tiie in- crease in the amount of carbon dioxide formed. Prianishnikov (1%) found in 1909 that the presence of calcium carbonate in the solvent had a decidely depressing effect upon the solubility of tricalcium pnospnete, due to the presence of the common ion,Ca, in both the solvent snd solute. Mitscherlich and Simmermacher (14) found in 1915 that the same princiral holag true when any calcium salt is presentj;but that the presence of other salts as as onium sulphute,amnonium chloride,sodiium sulpnive,and 3 magnesium sulphate,increased the solubility of the piuos;hate Inl91%,Jordan (15) reported that the solubility of rock p.ospnete increased with the fineness of the material and Peck (16) found the same thing to be true with bone eal.It vould seem tht the same princifals snould hola tru: with thie phosphate minerals of the sgoil,assuming that the composition of the minerals is the same. In 1914 Truog (15) reported results obdtuine: by Compoustincs row rock with manure and grass.ye found that there vas slignt,if any,solvent action of the fermenting material.ithis he tnougnt to be due to the fact that the water cm reed with curbon dioxide had became saturated with the pnos nates and tiat if it could be removed,as under field conditions,the results would svuow a sreater solvent acticn. In 191% Tottinshum ard loffman (19) reyorted results very similar to those of Troug.They carried their work further,nowever,and found thet the rnosy;.crous in the cell material of: the bacteria wus in an insoluble form,and concluded that the consumption of the pliosphorous by the bacteria after it hai become soluble rendered it once more irsoluble.It is very possible that, cad their results been taken after u longer pericd of time,civine the bactericl celis time to become broken down egain,thit they would have been different.ifcDowell (10) obteined results similar to these in 190%. T ie work just reviev.ed all deals with tne solu- bility of phosrnorous;the principals should,hovever, a ply more or less directly to all the mineralconstituents of the soil.eThe next work to be revieved treats of tne avail- ability of an organic constituent of the soil,nitrogen. Winter and Kobson (20) made investizat:ons along this line 96 and concluded trat ormgunic com: ouncs of nitrozen a-com:ose ard become svauilable more intensely in sands than in clays no loums when the vater content is low,out if the water S content is nigh the difference is much less marke:. Thus it will be seen that the factors which in- fluence the solubility of the mineral constitu-nts of th- Soll uuve teen rather extensively studisdgthe exect reasen “or tne relative insolubility of these constituents has received much less a*tenticn,oweverelt is taken for granted usuclly,th.t it is due to the chemical composition of the minerals in which the elements occur.Th- rext ork deals witn this subject.In 191% snpeared the results (71) of Prv,s work on the condition of the insoluble phosrhorous. Fre of "ered to hyvotheses as possible explanations of tne 4 insoluble coniition of the phospnororas ani worked on tnese. Yirst he succested that the pospvhoréus mignt be in in- soluble coupounds;but upon investivzation he found thut these insoluble compounds occur in only small .moints,not enoush to account for all the insasluble phosphorous present.ne that it misht be present in soluble com- es (9 nN ct a ~ C4 Co 69) cr Cy Cu poun 3 but proteeter in some wuy from the soivent action of the acids.This ne found to te true.Lergse quantities of phosthneatic minerils v-re protected by insoiable giurta Sralins. A work some.hat gimilar in nuture vas reported in 1915 bs, Plusuner (22). He ca ried on investisations on the relation between the mineral com:osition of some typi- cal soils and their fertilizer requirements.After mineral ° ’ e ; . 72 . 2 s a4. 15 4 analysis of the soils and corresponiing fertilizer estaiens LO s ® . he decided that there is a definite relation between the mineral composition and the fertilizer requirements. Other literatur2 move or less misvellaneous,but having a bearing uson the problems of tnis ruper,wilil ve taken uv at this point.iIn 1915 Whitney and Cameron (23) published results of very extensive stuiies ani reseuarcnr upon tne relation of tne chemistry: of the soil to its prodguctivity,and tnesr concluded tnat but very little can be determnied concerninzs tne productivity of a soil by its chemical conunosition.thnoy concluded tnat all soils contain enougn pluint food and ure pr..ctically inexiaustubdle?: ani thet the greatest Linitins factor,next to temneraturs, is water supply.They consider tnat texture is a very inm- portant factor due to its close relation to water nolding vowere Alimitei amount of valuable ana sisnificant work nas been done on tne relation wnhiocn sulfur bears to soil Fertility,and some of tnat work will be reviewéd here. Honokins (25) believes that sulfur is of much less impor- tance in soil fertility than many otner workers have velieve Ke contenis that in as mucn as there is a considerable amourt of sulfur a-des to tne soil each yeur by rainfill,tners is but slisnt possibility that it will ever become a limit- ing factor in crop productio:.ie seens,however, to nave neglected in nis criticisms tne fact,as estublisned by the Rotnamsted workers,thnat more is lost in drainage water tian is added voy precivitation. Shedd (26) nas curried on pot excerinents in an e‘fort to determine whetner or not ya sulfur is an inportant element in plant food.He found tnat lle it increeged the vield of certain of the levumes bat not of clover.ie could find ro rela*ion between the protein content of tne plant and tne sulfur content. Kossovitcn (26) has made extensive studies on the sulfur circle in soils ana concludes -hut *he smount added oy rainfall wiil not revluce 11 that is removed by plants and percolation.tnis leads nim to believe tn:t sulfur may in tine oecone as imnortunt a factor in soil fertility as chosphorous.Peterson (235) in 1914,reported that he found an increase in sulfur in plant tissuss,with an increase in the soil.enart and Tottinchiam (29) carried on a similar work ana found ti.t hne beneficial results with sulfur seemed to be with plants nizn in protein.Brown and Kellogg (30) mae several analsyses and found abdou*+ hulf as much sulfur as phospnorous ani concluded tnat it mi;ht in tine become a limitinz factor. Kuprecht (470) did work on the toxic effect of iron and aluminum salts on clover seedlings.This és a different phase of the question of the relation between chemical composition and productivity,but is of more or less importance.Ee found ferrous sulphate to be toxic if present in as small amounts as four parts per million, and aluminum sul;nate to be toxic if present in as small emounts es forty parts per million.He found that this toxic action could be overcome to a great extent by the addition of calci:mcarbonate. 12 Description of the Areas Studied. The soils used in this work were taken from two separate areas ir. Michigan:The Wexford County Area and the Allegan County Area.Both of these areas have been surveyed by the United States Burezu of Soils. Tne Wexford County Area is located in north western ilichigan,at approximately latitude 44 deg. 20 min. N.1It is avout twenty-five miles from the Lake Michigen shore and consequently is not materially affected by it.The climate is rather cool,ranging from 18 deg.F to 6& dueg.F in Auguststhe mean for the year is 42 deg.F. The rainfall is fairly well distributed throughout the year and is sufficient for good agriculture;the mean for the year is 42.4 in.Tne greater part of the area has good natural drainage ,and most of it is level to gently rolling ,although there are some very hilly sections.The soils are of glacial origin and were origin- ally derived from the breaking down of igneous and metamorphic rocks. Tae soil is nearly of the Coloma series;there is a small amount of Dunkirk and Clyde. The oldest parts,apriculturally,have been under cultivation only about “wenty-five or thirty years; other parts ten to fiftesn years,and a large amount of tne area is still unimproved.tThe chief crops grown are potatoes,corn,some oats,rye,wneat,and hay.lo very definit® system of crop rotation or soil management is followed except in a few isolated cases.Originallr very gcoda crois of potatoes could be raised,but continued cropping L436 has greatly reduced the yield. Tne Allegan County Area is located in south western Mijchigan,about latitude 42 deg. 30 min.N.It borders Lake Michigan and consequently the climate , especially that of the western part is modified.The mean annual temperature is somewnat higher here than in the Wexford Area,being 47.6 deg. F.The mean annual precipitation is 40 in. and is fairly well distributed throughout the year.The surface is rolling to hilly in some parts.There are considerable areas of poorly drained - land, including mucks and soils of the Clyde series. Some of the upland soils are characterized by numerous rettle-holes. The soil is of glacial origin and has been altered to some extent by water action.ihe soils are quite varied and the largest areas are of the Miami and Clyde sreies.The area has for tne most part been under cultivation for forty to sixty years.The chief crops are fruit and general farm crops in the western part and wneat,corn,hay,oats,and some truck in the eastern part. In many parts of the county very good systems of crop rotation and soil management are in use;in other parts but very little effort is being made along tiar 1 ine e 14. Descripticn of Samples Used. ae tne next few pages will be devoted to a des- cription of the soils used.This description will in- clude source,type,length of time crorred,chief crops erovwn,and any other points of value tnat could be learned wore or less difficulty sus experienced in getting a reliable history,especizlly in the wexford Area,.as most of tne farmers had been on their farm but a short time or ned kept little or nce account of the managemente Only those soils ure included, however, of which the history is considered fairly reliable.Less dificulty was met «10 ® e — mp am TO = ee lee we te ‘9 e e e le e 14,2 —_ wm am =" |=—* "Se "-m -8¢ -8B - onmlCcrllCCUCCelUClC eC Cc CS cl ss — a — > (13 : 9 ° =~ @e awe “3 “= @e “SB =~ = gw =H -@ =~ sw ww wR om Ss -—" Ss ww > 14 : e e om ame ow et a =e od ow ew wR ew oR ee TE ee -e ° 5 ° e _S ° mpness ww ew ae oe Ss Se a ee -_ s— 8 “= BF we Se ° nm ° e ‘ eo ™ mw «= ow = = we 8 eR RO Ke ee ~~ > 0 6G e \w e -- =m an oe ow en am mm =e me ee ell lel > 619: L? e 4 e i mes we = w = 2: a mee eee le — => <-> e 20 ° e — e am wes |] “SBS “HB OS HS SS Bee ee = eee ee e ~ 7 e e e es ate em ae ew em oem we ew ~e o—w oom we ws wt we eee ee e cs e e qa — e meer wT ST ae eme eeee eee eee e ~ 7 e ”~ ° —y e Cane _—me pa ee oe Te FEB ee lll -=- 2 @fe:fa™m"lNeU 6S rs. ce. 2 Lh e S( e <= e -—) nwpeanpeewns = SS =e we “SS wesw FF PS] we e — e e a e me a om se =O —-— —. ~~ =) we oe - «= oe e m~ e e ad oe = pp e@ eR eB eR =e we eB =e 8 eR oe TE Eee e ye e ® om f e oD ap ewe ww ww ese ewe ST eee eel eS OU OUD elle CLS _ 7 ° Ca ° are e be ww e -" =m @ of ee ee Ge & a a aw es of fn e — , oy, Oe l40 3: GO 0 “~ : ° a ° ny ‘ ee O pS CN }J o> el 22 hn: ew = «= 7 ew == — —-s ee ee ed v. \. ee q rN @ J e qc) ~ J © X > ee e j ©) ~ J AN ee ©) i i e C. &t ©) cy t 3 ~é ) Saorile Vov:Virs'4 of: ° Dp ¢ vee os ~~ mi ww @ © ~- ,~-* od a e /~ ats / AL i ee ee ee ee ee) -—@ | . an e ° ~ ~~ oa os e ~ 7 ‘ e _— / e e Oe re | e ao @ ames apap eanp ware +s Fs FB <« =“ WS =P we —“ Se wR eee eel Kelle le -_ —_=eP => ab aD —_ «= == e “? ~ e e oN mm“ ™ e o~ ‘ e J e e ~/ @ “yy - ° v7 e@ wx Dm a am =e -—8 oS eB ow —w ==] - 2 owe BOK TllUlUle le lela lee lf _ = = «=e =e = =a “== -2 «aw @fg-s SE vad e > L ° o - “7 e ~ ‘eon ° / * ° we ~FdA e ~ © — 8 e sf e . Dt e e >. ° e y e ena es == == ~~» —@3. = «2 oD s e 7 A e e Jj” e e a ap a ws oe oe = = om «om aD e yo _7-o-- or) e Bee eet eee e +f e ° QO ° oN e ao e e , _ e e 7 e es as eo @emenpeeem SB WZ WS «8s ws os oe : 4] 3 “4 =m—a_ nw ee -e8elUlU SUS UCU la nN s ~ WeVu yy e /@ Cell5 3 %. me OTD 8 oe G 7? = Ce LOD : O.% 9. Osu ° 0.9 0.309 : 0.0 QL e O.C9o e O.C : G8 > O.155 3: +«.0 : 45 8 2 020''9 3: O20 : 44 3 42 > 0.155 : 0.0 a Te ee Ne : GZ) 3 > 046.005 2 O68 : 47 3 > 0.145 : O.C > 50 ft > 51 : _ 5a 8 : 54 8 0.05% : O.C¢ : Oe 104 : Ie C > QO.0-9 =: OC 50 0.167 : O.C¢ > 02040 : OC : 55 3 © O45 3 O64 : 56 54 2 0.109 : O.C Table: I. (Cortinucd) | :Sample of :Virgia of: p~p NX : 7 P. : 57 : > 0.105 3 U0" : 52 : 0.065 ° CO. one ap @e w= - =SB= we 8 CBOs OS aeae ss ws ew ee ee =e MS CR we Ke TR eee elle ele lee | eo eee en wasesa se Be BP Vw we ews FT Pewee OwWw ee SBS ee Ss DR wea we ee ee ee le Ue le Bane wma wT we ws we weewmeemeese ae em epee ss HF as ww TB TH ow ae ew we Ee ee Se ee ee mannan @ wwe se ee VS ww SP Se se ee es ew we ew ee ee ew et ee ate = eS ee Ow om om lt = @R ape @eweem® 2 =s ase ewes SH SS DP SDP ew WDB ewe =— «Ss se Be =F 8S 2 = = Re le lew apenas Se we wZ BO SP Se wes SF ewes =e wea ase ww =) 8 SOS DR eH 8D 2! 8S 8 oe me Qn =e 2 ew ses Ves ee SB wee wee ewe ee ele ee eel ee ee ee ee es nana wewzwee ws —s = w= web ws = awe wee aw se = =e wee eee eee ee ee le le la lt eomaenwraenna Sse we Be eSB SSP SH SPewse Bese ese mas aww es ww ow PF ee se ee eae weeeewvwese ewe eer eaeeeweaewewes 3 |= Ss @wwWesw_ WSs wSewew =| Ws 8 sw ew es = fh On e A comparison between the per cents of the vafious plant food elements,and total plant food of productive anda unproduttive soils,of the same and of different types,brings out some very interesting points. First a comparison might be made between soils 12 and 25.The former is a virgin medium sand and will produce from 40 to 50 bu. of wheat per acre,and other ceneral farm crops in proportion.It contains fairly large ampunts of all the various food elements with the exception of phosprorous; in this it is deficient.It has a total of 1.965;5 of plant food. 25 is a cropped fine sandy loam wnich will produce only 20 to 25 bu. of wneat and other farm crops in proportion.It contains large amounts of all tne plant food elements ,and a total of 5eAZL2 of plant food.or nearly twice as much as 12,fet will produce only about half as much.That is to say 06161, N,0.037% P,1.591j2 K,0.0677% S,and 0.259;5 Ca will produce as large a crop in a virgin medium sand as 0.240% N,0.4567 P,2.231;5 K, 0.103% S, and 0.6115 CaO in a cropped fine sandy LoOeme We have a similar condition in th- relation between soils 19 and 48.19 is a virzin medium sand which will prosuce 40 bu. of wneat per acre;it is low in no essential element with the exception of phosphorous.lIt has a total of 1.066. of plant food.35 is a cropped clay loam which will produce about half a norm.1l crop.It is very deficient in pnosonorous but in no otner element,and has a total of 2.454. of plant food.Or in other words 0.1657 N, 0.051% P, 1.752,%, 0.04055 S) 0.426% Ca0O,is producing about one-half as much in a crooped clay loam as 0.178% N,0.086% P,1.365 K, 0.059. S, 27. and 0.550. CaO,in a medium virgin sand. Otner similar comparisons mignt be made;32,a virgin medium @dand with a total of 1.646. Of plant food,produces more than 21,a cropped fine sandy loam,with 2.2035 of plant food. 50,a virgin medium sand with a total of 1.051% of plant food, produces more than 26,a cropped clay loam with a total of 2.2455 of plant food. Thus we have a series of comparisons between produc- tive and unproductive soils of different types.It would also be of value to compare productive soils of one type with productive soils of another type.Soil 1l2,with a total of 1.955, of plant food,19,with a total of 1.868% of plant food,are productive medium sunds;24 with a total of 2.75155 of plant food,24,with a total of 4.550%. Of plant food,are productive fine sandy loams;28,witn a total of 387955 of plant food,40,with a total of 5.655% of plant food,and 44 with a total of 2.662% of plant food,are productive clay loams. Here we have a series of soils of very nearly e}uul productiv- ity, buy of different types,and varying in total amount of plant food from 1.916% in medium sand to an average of 3043500 in tne neavy loams. To make tne comparison complete two or three more soils should be mentionedwhich are virgin and contain fair amounts of clant food,bout which are not productive.56 and 59 are medium sands,uncropped,and witn a total of 2.2825 2.310% of plant food respectively.Tnis is considerably above tie averase for tne productive sands men’ ioned above./74 is a medium sand witn 1.204. of plant food,only a trifle under the average for the sanis but is very unproductive. We now lave compurative duta on productive and un- productive soils of tne same and iifferent ’:;pes.One of the most obvious facts brousht ont by these data is the great aifference in the relation ovetween comsosition and produc- tivity ,due at least apparently,to type.eA mucn lower amount of plant fcod is needed in a sandy soil to produce a given crop tnun is needed in a finer,loam soil.This might be due to a ditference in tne availability of the nutrient salts present,or to a difference in tne pnysical conditions.Poor tilth and drainage misht easily account for the difference between 19 and 33.lfuch of the organic has been removed from 56,leaving a poor condition of tilth and the drainage is poor.eBoth of tnese coniitions are dertrmental to aeration and bacteriul activity,and conse juently the plant food might become available more slowly,and the roots of the plants might be hindered in their development.Poor tiith misht also explain the differences between 50 and 26.1: would not,however explain the difference between 12 and 25 or 32 and 21.The possibility of availabillty being tne determining factor will be discussdd later;but it is very probable that the different ce is due in a large measure to type and consequent differst ces in texture.As pointed out by Hilgars,the coarser textured soils are more easily penetrated and allow a more extensive root development,and tnus the roots may come in contactb witn a larger amount of plantfoode One further comparison of variations within the class should be made.l2 has already been discussed3;56,59,and 43 are of tne same class as 12.The typemnis different;1l2 is a Miami medium sang @4d 56,59,and 3 are Coloma medium sznds. 296 The tnree latter are coarser,mote open,and are underlaid witn a coarse and very open sub-soil.Tnese physical character- istics of the Coloma soils coupled with a probable difference in tneir geolozical origin ani mineral composition so reduce their water holding capacity tnat their pdoductivivty is very Low. As a result,tnen, of a comparison of the producing capacity of various soils we find that an interpretation of such analyses must be modified by a considereation of the class,drainag2,sub-soil,organic matter content,and all other conditions which may in any way :ffect the bacterial condizions and the availability of the plart food. This brings up the question of the relation between the amount of avaioable plant food and the producing power of the goil.Owing to the limited amount of time available for the work,only one sample of each type was used,and only the available phosphorous and potassium were determined. The digestions were made with N/5 HCl and the determinations made according tc the official method. Table II gives the results obtained.The soils are arranged according to type and the results given in parts per millio:: (ppm) Table II .Sho-inz tne ppm. of Av:silable K,Oand 506 P50 in the ¥arivcus Ti m23 ofSoil. >: Sample No.: Class : Virgin of: ppm. Ko0 =: ppm. P05 : : l. :MediumS: ~~: 170 =: 103 =: : Bootleg t~) O54 tO ”—~—C QC Do 5 tC UCU tCOCHSC< SC‘id ti‘ OH : 2: UU 144.6: #130 +: 397 ¢ : 40: mp 160 : 360 3. : 45: : : m0 : 7 of : aps me 270 2=OtidOHt : mo: UMUC 2 73: " : 7 +: 100 +: 179: : 24. :FineSL : 25 : 290. : 210 ¢ : Pe a © Co : 33 :ClayLb oo: +: 20 +: £61 ¢ : 40: " : 38 +: 290 3 80: : 42: * : 3 10 : 45: >A tt 90—iti(‘C™S# 64: : 6: " : ¢ 20 +: 263, 670 HCtCOSCi‘aESSSSO en eeeeeeFe@aaep@e@e@meewmwe--- ar iaeaeiagwess Se Fs Fe ePesewseeeersesteaerwnw ws Fe se nseneeieeewe ws @ weeds = = =F = oF oD oF -— 41. Areview of this table wil” show that it is more or less difficult to draw from it any conclusions concerning the avallability of phnos:norous and potassium in relation to productivity. It is not certain tnut the quantities inaicuted in kh the table really indicate the amounts cctually in the soil because considerable quantities of each of these elements might go into golution,but be readsorbed by the soilj;in this case the filtrate would: contain less after filtering than had previously been in solution.This vould be especially true of the finer types of soil. Using tne results given in the table as a basis for discussion,it would seem that in no case should the amount of soluble material be a limiting factor except in the case of tne phosphorous of the clay loams and. 45. But the mere fact that certainamounts of these elements are soluble in weak acid,doesn,t necessarily mean that they will be available to the plunts under the conditions as they actually exist in the field.¥or example in the soils 12 and 14,the former virgin and very productive,the latter cropped and very unprodustive,the amount of soluble phosphorous and potash has not been chunged enough to cause such a marked difference in productivity.But the organic ma'ter in 14 is so much lower than it was originally that: the water holding Capacity iias becnbgreatly reduced and one would not expect to find so much of the nutrient salts actually in the soil solution as in 12,altnough it might be in 2. soluble form. Also there may be enough soluble, plant food in the soil for good plant growth,yet the structure of the soil 426 and other conditions of tilth be such that the root system cannot come in contact witn sufficient quantities of it to give the best results.Thus,in 25 and 12,we find only a small difference in the amount of soluble material present,yet soll 12 is much more productive.It is more open and permits of a better root development and consequently the soluble m material may be more com-leyely utilized. Another factor not brought out in the table,but which might partially explain the lack of correlation between the dissolved material and the amount of plant growth supported,is tne availability of the nitrogen.If the con- clusions of Winter and Robson (20) be correct,that nitrogen- ous compounds become available more rapidly in sands than in finer classes of soil,then a smaller amount of nitrogen in a sand would produce better results than a laregr amount in a loam soll due to its greater activity.It is also conceded that the nitrogen of a virgin soil is more .ctive than that of a cropped soil;consequently,although there might bea large amount of nitrogen in a cropped soil,as in 25,than in a virgin soil like 12,yet its inactivity might be a limit- ing factor. Thus we find difficulty in interpreting results ob- tained in availability studies withoutza careful consideration of all other limiting factors. The ohter pnase of the subject,the effect of cropping upon the chemical composition, will next be taken up. The croy;ped soils have been arranged in Table III according to type,and the per cent loss of nitrogen,pnospuorous, potassium, gsulfur,and calcium recorded,together with the number of years of cropping. 350 Table III.Showing the Per Cent Loss of the Various Food Flements and the Number of Years Cropped. :Sample:Type:N Loss:P Loss:K Loss:S Loss:Ca Loss:Croyped: 5 4 311.64 4212.2 429.20 £:58.6 %: 730.—Cié«*‘S > 7 et "210.6 4:28.84 )=—or4aa ds 20°" > 10: " 250.9 £:20.6 £:7.60 431.3 &: iOS : 14: " 234.1 %:20.6 £:10.8 4:59.6 2: 360 ‘ Se 2S Feet aee@aees Beraeaee @eeegeeeret eg ae see ew werern ese ase BP ewe es eee wee es @® ws @ 8 @ & & nwenmnewneneeeeitiws we SB ee weeeaeene —- Bete wee eee eee aecwewwa SS ae aw 8 eo we we a aw = «= = mm eae ow = as GS eo GE GE a 54-6 Only discussions of 2 rutner senere-1 nature will be possible of Table III for the following reasons:certain loss es have ocurred inthe virgin soils during the time wnich tke corresponding cro:ped soils have been under cultivation ; consequently it 1s impossible to know the exact original composition of the cropped soil. Also it has been entirely impossible to determine the exact number of the various crops that have been removed from the soil during the pericd of cropping sand as there 1s so great a variation in the compo- sition of the crops removed it is impossible to more than acyroxiniate the amount removed by the crops. The losses from the soil are brought about through percolation,surface washing,some loss of nitrogen into the air,and by the removel of crops.All of these losses are of course greater in cropped than in uncropped soil,.althoughn there are some losses from the uncropped eoil due to per- colatione | The hignest per cent of loss seems to be in the case of nitrogenein practically all cases, judgins from the length of time cro };ed,the loss of nitrogen is greater by far than that brought about by the removal of plants alone.this is especially true of the sands.Although there is no great dfffer- ence in the averare loss in the sands and the loams it must be kept in mind that the amount of material really removed by croyping from the sands is much less.Soil 10 has been cropped the same length of tie as 21 or 33 ana originally produced as much, yet during the last half or more of the fifty or sixty years of cropping the production on the sands has been far behind tnat on the heavier types and at present practicelly mothing is being taken from the sands,and the loams,esrecially 21,ere still producing crons thet mean a distinct loss in plant food. Thus the loss in other ways must be greater in the sands.Soils 10,14,17,and 50,are open soils and consequently the oxidation of the oreanic matter would take rlace very rapidly,and altnhoush the sub-soils are not excessively open, vet the loss by percolation would be great during cropping. Oxidation would be much slower in the loams. in the soils,1,4,7,34,54,anda 57,ve have not only avery open surfuce soil but a perfectly pervious sub-soil, both of which conditions greatly promote the loss of nitrogen Consequently,although none of these soils have produced a large ageregate of crops yet the nitrogen loss is comparati- vely,high,considering the time of cultivetion. the loss of phosplicrous averages considerably than thet of nitrogen.eThis is probably due to the fact thot plans remove less as food ana tnit there is less loss by percola- tion, due to its insoluble form in the soil.It is notice:ble that the losses are somevnat higher in the loam soils than in the s.nds,due to the fact,no doubt,that the removal by cropping is much greater and *he possible loss by percola- tion is small.65 and 71 are striking except ons to the above stu’ement.It is impossible to explain the greut 10ss of phospho ous from71 without further study into the form in which it occurs as this is in unusual soil.The gain in 65 is undoubtedly due to imperfect samrling. The losses of potash,1in percents,are considerably lower than those of phosphorous or nitrogen.This is because of the much larger amount of potash in the soil,and the 366 actual loss in pounds per acre is creater.e There is a somewhat greater loss in the loams than in the saunas aveé,as in the case of phosphorous,to the greater total amount of production bzr the loams and the comparative- ly smail loss due to percolatione In a consideration of the loss of sulfur,two points are of especial interest.The first of these is the fact that tnere is a decided loss of sulfur from the soil due to cultivetion.this fact is worthy of notice at this time becamse of the discussion in recent literature as to tne improta- nee ot sulfur as a factor in soil fertility.These results corresrond very favorutly to those obtsined at otner station And in as much «: sulfur is considered a necessary element of plant f od =nd is a constituent of all protein compounds, it would seem that sulfur should be consicered a factor in fertility problems. The second point of interest is the relation bet:-en sulfur ena nitrogen. Although it cannot be said that there is a distin: t relation between total sulfur and total nitrogen,yet it is evident from Table I that there is some relation and those soils tha* are especially highin nitrogen are likewise hisl. in sulfur.Tne reverse of this,however, isnot elwaz trise but the striking relation between the two elements is thut of the losses due to cro;ping.It will be seen from Table IIi that with but two or three exceptions a high loss in nitrogen is acconpcnied with a nicn loss in sulfur,and a loss in nitrogen by a low loss in sulfur.This misht be accounted for b7 the close a sociation of each element with the protein compounds.Plants vnich are hirsh in protein of course take up large amounts of nitrogen and the same thing has been found to be true with sulfur(29). Thus a large removal of nitrogen by plaints should be accom- panied ty a large removal of sulfur.Likewise when the plant tissues in tne sc 1 decay and break down into scluble com- pounds the nitrocen and sulfur ».ill be associated in the accoripanying loss by percoiatione The last element to ce considered in connection with its loss due to cropping is calcium.A considerable diffi- culty was met with in gettigg consistent data on calcium convent wend loss due,probably,to sampling in the field,as it has been found that a much larger number of samples is necessary to make up a representative composite than is the case with the other elements.The loss is in all cases,of course,far beyond thit which could be caused by cropping aloneeTh2 loss by percolation depends upon the nature of the comrounds in which the calcium occurs.The highest losses recorded o.cur in the loam soils in which the calcium is probubly in the form of a carbonate. One of the most interesting and practical phases of soil depletion by cropping,is the difference in the lasting powers of theddifferent types of soil;that is the difference in the length of time which the various tyres of coib will _ produce profitable cro})s.This subject,with the relation t the original vegetation,will be:taken up at this -oint. Soils 12,19,23,24,56,59,and 67 will be used in the comparisonsel2 and 19 are Miami sands covered with oak timber. These soils are at present producing no profitable crops, and have not for the past twenty years.Soils 23 and 24 are Miami loams covered with hard-vood timber including beech 58.6 maple,anc some hicory.There nas been only a gradual decrease in the productivity of these soils,and even at present,afeer fifty years of continuous crcpping,they are producing crops at a profit,even though it is smal1.56 is a Colome medium sand,covered chiefly with pine,with some hemlock and cherry. 59 is essentially the same type,but is in a region of trans- ition between the Coloma ana Miami series,and is covered with small. maple and beech timber.These soils produced fuirly good cropu for the first two to five years,then decreased very rapidly in productivity,and no profitable crops can be grown after ten years.67 is a Clyde loam covered with elm,maple,and a greet deal of oak.Its productivity,like thit of 24 and 24,has decreased gradually and it will still produce a somewnat profitable crop. Tnus we find tnet the uplund,oak soils are much less lasting than the low land oak soils,or the up land typical maple and beech soils.The pine up lands are much less lasting than the hard-wood up-landsésthis is duevery largely to their coarse texture and open structure,which decreases their water holding capacity and incrszses oxidation to a great extent. As a summary of this point, it is apparent that oak launds,if low,are very lasting in productivity;iand if high are only moderutely so.eThe maple and beech up lands campare very favorably with the oak low lands in this respect.The least valuable from the standpvint of ubility to produce profitable crops over a series of years,ure the Coloma pine soils. 39. The next and final phase of the work to be discussed is the chzunges in the acidity of the soils due to croxping. The Veitch method was used in making the lime requirement determinations. The results are recorded in Table IV. 40. Table IV.Sho.;ing Changes in Lime Revuirement .Nosign before CD the Number in the Last Column Indicates th- Chunge to hav Eeen un Increase.eA - Sign indicates a Decrease. : Sample : Virsin of 7:CaO Te-vuirement: Chunge : toll "22500 pbs. per A: 2500 Lbs. DOR 1 oo: wone Ot : 4CU””””””CS OOSs=“‘<‘ 14 t”~<—s”:—C CG a >: 42: 14 71000 " " ©; > 47 oo: GOO Bag > 19 00Clt” 17. :0240 > 2k ot 1600" * 8 800) > 28 : 21 32400 " * Hp > 2 oo: s1l00 223300 > 2 : 260870 > 30 ¢: None tno change 2 B28 > 34)~C~C~ 40 +: 38 None : | D4 BQ Do 44 42:00 * *§ ® : | > 5OO™C*~C~Sé‘<=<=CS*é‘é‘é‘éit wR RCC = Eenwe ee Sesser e SF FP BF OFT TBE FT Fr eeeag2”2#x weer qneG Baste eweFeeeae@ges BP Bex Ff SG FS &S @& ao ea we wmweeewewroeeerer anaes SF SF PMP esse VBwaraswewFZF F SF Oewwwrwoero OQwerww es eeeuae @& £2 Fw @ 41. Table IV. (Continued) : Sample >: Virgin of 2CaO Recuirement: Change meee Ferme ene 2 @ eee eeee Ber @ewewew Fee @ wD & S&F aww wow TB www wee ww eS SS eS eS oe Ss ee SE as = es : 54 : : 2000Lbs.per A : 400 Lbs. : 56 54 2=~=C<( GOO OM : 56 + $3400" * ® 2 -5600" > 6=— 59s s— it and wnich correct it ,are more or less limited,and that the metnods of determining the lime require- ment are somewhat erbitravy and may not be accurate. For pur:oses of dixcussion,the acidity will br con- Sidered as positive and nezative.eThe positive acidity being due to tne actual presence of acid and the nesative acidity being due to the lack of sufvicient bases to satisfy tne adsorutive power of the soil for bases.An accumulation of orgenic matter, poor dralnage,or other conditions detrimental to oxidation,.would undoudtedly increase tne positive acidity, and anything th: t woula increase the oxidstion or improve tne Gruinage w.yuld have a tendency to decrezse it.Any fuctor vnich removed the basic muterial of the soil would undoubtedly increzse the nesutive acidity. In five of the crozved soils studied,three sands and two loams,neither an increase non ae decrease in the acidity is noticeable.These ure 10,14,%0,60,and 65.12 is the virgin of 10 and 14,end 432 is the virgin of 40.These virgin soils estreciully 12,contain-a considerable amount of orgenic matter and vere covered with quite a large accumulation of leaf mola etceAltiough this accumulation was entirely removed before sampling,it no doubt has had a marked effect on the acidity benecthetnis materiel has been very lurgely oxidized in the corresponding cro;:ed soils.Thus any increase in negative .cidity vnich hus been brought about b: a removal of buses 7 in th- crorrped soil has been counterbalanced cy the decrease ze fs in rositive aciadity.Tkhe negutive acicity wouls uwiso be some- ts b vnat Ccecreased by the oxidution of the orgunic col: olds,tnus reducins the totul adsorbing surface. ~ 3 In soils 61 ani 65,the orivinsl besic content was ereut enoush to enatle the soil to retuin its basic reaction for u conusiderectle number of yours. In four suniv soils,17,74,50,cena 5o,and ty.o loums, Sl ani 42,there has been cen uctual decrease in the lime reyiiresenteTiis ig no doubt due to the sume causes mentioned 7 in tie prececing discussion but the oxidation of the acids ani “he ors.nic colloids hus be °n more complete.The virsins of Cl and 42 ere very hich in organic matter which hus been udye extensively oxiaizged in the crorred sgolls. In five sund3,1.4.7.52,und 7l,and in one loam,3<, mo line rejuironent is increused.in tre first four soiis men*ivnead vie amount of ucidity lost by cultivation lias been more thun courterbelenced by the increase due to the removal of tuses ana we have an irncresse in ‘i.e lime re.uirenent “Sia result.71 is a sreci.l case and there is a strons rossib ility that the acidity is due in part to un acid fe -rous Sult,as there is a larze umount of iron in tke sgoil.this Cannot be lost - percolution beceuse of tne impervious sub- : --™ 2 IL. 2c 2 a in very poor tiltn ans conse uently there is a Cf © fF c.cnce for the accuzulation of uclac. tous we find th .t contrary to the retner eeneral Cvinion thut crorgin:g increescs acidity,it as often decreases 1+ as i+ does increuse iteTtThis is «et least true in so fur ~ as our present methods ure reliable. It wes hopea th: t sume relation might be vorked out 44. between acidity and crop production,but owing to the large numver of otusr factors tu.t are presert to influence ‘he wields,i%5 is impossible to druw uny conclusions elonz thet 4 a | ineeocme of the soils that rave decreased the most in 4 yrodguctivity save no greater lime resuirement than they had Welen Virvin,es guarzed by the uncronned soiljind some soils weich heve un incre.sed aciiity hve shovun tne least decrease in jproducine c.c.citire 45. SRNL Provably few if any,en*tirely nev facts have been orousht out by tiie work reported in this paper.Certain facts, Kowever huve been emphasized by tne dute presented here, and our knovleds? of Michicun soils nas been greatly increase L.Uffect of chemical com: osition upon productivity. eine Most valuable point brought out by the data on this phuse of the subject is great variation in produc- tivity due avparently to t:ipe. 2eine Sume amount of piunt food willl produce much better results in a seniy soil thun ir loam soile Z4euhe sume umount of plant food in « virgin soil woll produce better results than in «