EFFECTS OF ALKALINITY ON ADSORPTION AND REGENERATION OF PHOSPHORUS IN NATURAL LAKES By PAUL HOWARD BARRETT A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in Partial fulfillment of the requirments for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1952 ProQuest Number: 10008257 All rights reserved INFO RM ATIO N TO A LL USERS The quality o f this reproduction is dependent upon the quality of the copy subm itted. In the unlikely event that the author did not send a com plete m anuscript and there are m issing pages, these will be noted. Also, if m aterial had to be removed, a note will indicate the deletion. uest ProQ uest 10008257 Published by ProQ uest LLC (2016). Copyright of the Dissertation is held by the Author. All rights reserved. This w ork is protected against unauthorized copying under Title 17, United States Code M icroform Edition © ProQ uest LLC. ProQ uest LLC. 789 East Eisenhow er Parkway P.O. Box 1346 Ann Arbor, Ml 4 8 1 0 6 - 1346 Paul H. Barrett ABSTRACT A series of inorganic fertilizer applications were made to four of six experimental, limestone-sink lakes in northern MIchgan in the summers of 1949 and 1950. The methyl-orange alkalinitles of the four lakes ranged from 138 to 192 p.p.m.; alkalinitles of the control lakes were 34 and 74,p.p.m.. .Epilimnial and.hypolimnial water samples, and profundal soil samples, collected at periodic intervals in the summers of 1949, 1950, and 1951, were analyzed for phosphorus contents. Soil samples collected from terrestrial, littoral and sublittoral zones were also analyzed for phosphorus contents. Other routine data obtained included alkalinity of epilimnial waters and organic-matter contents of profundal soils. that: It was shown (1 ) total phosphorus concentrations in epilimnial waters were greater in fertilized lakes than in unferti­ lized lakes during the first year of fertilization; (2 ) control lakes were as high in epilimnial total phosphorus as fertilized lakes during the second year of fertilization, but this was attributed to unusual climatic conditions, and (3) epilimnial total phosphorus concentrations were essentially the same in all lakes the year after ferti­ lization. The rate of disappearance of added phosphorus from epilimnial waters was shown to be statistically 305 96? Paul H. Barrett related to the degree of alkalinity of the lakes, with the greatest decreases occurring in lakes of highest alkalinities. It was concluded that fertilization of lakes of the limestone-sink type will have only temporary effects in increasing the supply of available phosphorus in epilim­ nial waters* No statistically significant increases of available phosphorus in profundal soils occurred following fertili­ zation in the two lakes receiving greatest quantities of phosphorus applications. The two lakes which received the least amounts of phosphorus, and a control lake.,, showed statistically significant increases In available phosphorus between 1949 and 1951. Increases' in hypolim- nial total phosphorus concentrations occurred following fertilization in the fertilized lakes. The concentrations of available phosphorus in the profundal soils were corre­ lated with organic-matter contents and not with hypolimnial total phosphorus contents. The assumption was made that available phosphorus contents of sediments do not consti­ tute a valid indication of regeneration potential from the mud to the water. Greatest phosphorus regeneration occurred in the lakes in which the sediments were intermed­ iate in organic-matter and calcareous-matter contents. was concluded that phosphorus concentrations in lake waters are a function of an adsorption-regeneration It Paul H. Barrett mechanism, the equilibria of which depend upon the quantities of organic matter and marl in the sediments. Evidence obtained from laboratory experiments supported the assumption that precipitation of phosphorus by calcium.may be an adsorption mechanism, rather than simply a molecular formation of tricalcium phosphate. Precipitation occurred at an alkalinity concentration of 116 p.p.m., a value considerably lower than the minimum limit of calcium-phosphate precipitation of natural lakes, as proposed by Naumann (1932). LITERATURE CITED Naumann, E. 1932. Grundzuge der regionalen Limnologie. gewasser, 11: 1-176 Binnen- A CKNOWLE DGMENTS The writer is grateful to Dr. Robert C. Ball for advice and encouragement offered during the course of this study. \ The InvestIgation was financed in part by: All-College Research Fund; Station; (1 ) the (2) the Agricultural Experiment (3) the Department of Biological Science, and (4) The Institute for Fisheries Research of the Michigan Department-of Conservation. The author wishes to express his appreciation to: Dr. Chester A. Lawson, for providing laboratory facil­ ities; Dr. Albert S. Hazzard and Dr. Edwin L. Cooper, for providing housing and other facilities at the Pigeon River Fisheries Experiment Station; Dr. Don W. H a y n e , for suggesting the methods of statistical analysis employed in this study, and Dr. Marvin D. Solomon, Mr. Norman Benson, and Mr. Howard Tanner, for assistance in the field. Paul Howard Barrett candidate for the degree Doctor of Philosophy Final examination, May 5, 1952, 10:00 A. M., Room 404, Natural Science Building Dissertation: Effects of Alkalinity on Adsorption and Regeneration of Phosphorus in Natural Lakes Guidance Committee: Dr. Robert C. Ball, Professor, Department of Fisheries and Wildlife, Chairman. Dr. Peter I. Tack, Professor and Head, Department of Fisheries and Wildlife. Dr. Gerald W. Prescott, Professor, Department of Botany and Plant Pathology. Dr. Kirk Lawton, Associate Professor, Department of Soils Science. Dr. Henrik J. Stafseth, Professor and Head, Department of Bacteriology and Public Health, Graduate Council representative. Outline of Studies Major subject: Zoology Minor subjects: Soils Science, Botany Biographical Items Born: Novenber 7, 1917 Undergraduate Studies: Michigan State College, 1936-1940, B. S., June 1940. Graduate Studies: The University of Connecticut, 19401941, M. S., 1943; The University of Michigan Biological Station, Summer Session, 1941; Michigan State College, 1948-1952. Experience: Fellowship in Zoology, The University of Connecticut, 1940-1941; Fisheries Technician, Michigan Department of Conservation, 1941-1942; Medical Entomol­ ogist, Medical Department, Army U. S., 1943-1946; Research Assistant, Zoology Section, Michigan State College Agricultural Experiment Station, 1946; Instruc­ tor, Biological Science, Michigan State College, 1947Organizations: Sigma X i ; Phycological Society of America; American Society of Limnologists and Oceanographers; American Fisheries Society. TABLE OF CONTENTS PAGE INTRODUCTION ........................................ 1 REV IE V/ OF L I T E R A T U R E ............................... 5 ...................... 6 Alkalinity and phosphorus Alkalinity and p r o d u c t i v i t y ........... Terminology ......... DESCRIPTION OF LAKES *8 . . . . . . . . . . . . . 11 ............................... 15 P R O C E D U R E ........................ 22 LABORATORY M E T H O D S ..................... 27 Total phosphorus of lake, waters Phosphorus in profundal soils ........... .................. Organic matter in profundal soils ........... Inorganic nutrients In water-shed soils RESULTS OF FIELD INVESTIGATIONS . 27 28 29 . . . . 29 . . . . . 30 ......... 30 .................. 31 pH of laboratory experiments Volumes of lake-bottom soils • . . . . . Phosphorus concentrations in surface' waters • • Phosphorus concentrations in hypolimnial waters 33 39 Changes in concentration of total phosphorus in surface waters ................................. 42 Changes in concentrations of total phosphorus in hypolimnial waters .................... • • • • 46 . . . . 49 Total phosphorus in profundal s o i l s ........... 59 Available phosphorus in profundal soils PAGE Available phosphorus In littoral and sublittoral zones * ................................. Per cent organic matter in profundal soils 64 ... 66 Chemical characteristics of drainage-area soils 67 Duration of phosphorus retention in lake waters 72 Alkalinity of lakes. 79 ...................... The relationship of alkalinity to the retention of added p h o s p h o r u s .......................... 81 RESULTS OF LABORATORY E X P E R I M E N T S ................ 88 D I S C U S S I O N .......................................... 95 S U M M A R Y ............................................ 116 LITERATURE C I T E D ................................... 124 LIST OP TABLES TABLE I. PAGE Comparison of the Alkalinity Ranges for Soft Waters and Hard Waters as Published by Various A u t h o r s ................... II. Morphometry and Physico-chemical Characteristics of Experimental Lakes III. 13 ... 17 Amounts, Types, and Dates of Fertilizer A p p l i c a t i o n s ........... IV. _ 24 Theoretical Concentration of Phosphorus Resulting from Application of Fertilizer V. . ............ . . . . . . . . . 32 Concentration of Epilimnial Total P h o s p h o r u s ........... VI. VII. Alkalinity of Epilimnial Waters XI. 36 40 Changes in Concentration of Epilimnial • • • • ............ 43 Changes in Concentration of Hypolimnial Total Phosphorus X. • . • • • • • • • • • • • • • • . Total Phosphorus IX. ............ Concentretion of Hypolimnial Total Phosphorus VIII. 34 ............. Available Phosphorus In Profundal Soils . . . 47 50 Average Concentration of Available Phosphorus per Unit Area In Profundal S o i l s .......................... . 57 TABLE XII. XIII. PAGE Total Phosphorus in Profundal Soils ......... Average Concentration of Total Phosphorus per Unit Area in Profundal S o i l s ......... XIV. 61 Rate of Pho.sphorus Application per Unit Area XV. 60 ............................... 62 Concentration of Available Phosphorus per Unit Weight of Soil In the Littoral, Sublittoral, and Profundal Soils, during August, 1 9 5 1 ............................... XVI. 65 Concentration of Available Phosphorus per Unit Area in Littoral, Sublittoral, and Profundal Soils during August, 1951 XVII. XVIII. . . . . Per Cent Organic Matter in Profundal Soils 68 Chemical Characteristics of Watershed Soils Surrounding Each L a k e ................ XIX. 66 70 Per Cent of Added Phosphorus Present in Epilimnion and Kypolimnion Three to Pour Bays after Fertilizer Applications XX. . . . . 74 Per Cent of Added Phosphorus Present in Epilimnion Several Weeks or Months Following Fertilization ........... XXI. . . . . 75 Analysis of Variance of Per Cent Residual Phosphorus Present Several Weeks or Months Following F ertilization■ 79 . TABLE XXII. PAGE Regression Analysis of Per Cent Residual Phosphorus and Average Alkalinity Values XXIII. 83 Epilimnial Total Phosphorus Concentrations during Summer Periods Least Affected by Fertilizer Additions XXIV. ...................... 85 Analysis of Variance of Epilimnial Total Phosphorus Concentrations during Summer Periods Least Affected by Fertilizer Additions XXV. . '................. 87 The Effect of Alkalinity on Removal of Phosphorus from Calcium Carbonate Solutions in Laboratory Experiments. Not Agitated XXVI. Solutions ...................... 89 The Effect of Alkalinity on Removal of Phosphorus from Calcium Carbonate Solutions in Laboratory Experiments. Solutions Agitated Continuously ...................... XXVII. 92 Rate of Phosphorus Application per Unit Dry-Weight of Profundal Soils . . . . . . . 106 LIST OF FIGURES FIGURE 1. PAGE Seasonal Variation of Available Phosphorus in Profundal S o i l s ......... . ............... 2. The 1949, 1950, and 1951 Mean Average Variation of 3. 52 Available Phosphorus in Profundal Soils 53 Comparison of Concentrations of Hypolimnial Total Phosphorus and of Available Phosphorus in Profundal S o i l s ........................... 107 INTRODUCTION One of the primary problems with which the fisheries biologists are faced in their efforts to improve fishing con­ ditions is that of increasing the supply of natural fishfood organisms in aquatic habitats. Fundamental to the food resources of all organisms are certain essential inorganic nutrients which plants require, and without which, therefore, basic food-organisms could not exist. In the science of agronomy, knowledge of (1) nutrient requirements of plants, and (2 ) available nutrients in soils, has led to the development of major concepts and principles of plant production. Since aquatic plants and terrestrial plants have similar nutritional requirements it has seemed logical to the fisheries biologist that certain principles of agronomy might be applicable to the problem of increasing plant and fish production in lakes and ponds. As a result, a number of Investigators have added commercial inorganic fertilizers to natural bodies of water in an effort to stimulate greater plant growth and thus provide a greater amount of basic food for fish. In general this procedure has been successful enough to warrant continued studies. At present it Is known that in regions where winter climatic conditions are not too prolonged or too severe, the close adherence to the recommen­ dations of experienced fisheries biologists will yield profitable returns from a properly constructed and managed farm pond. There has not been the same degree of success in the management of natural lakes. Here the aquiculturallst Is hampered by not knowing either quantitative or qualitat­ ive aspects of the nutritional requirements of aquatic plants Moreover he cannot yet adequately measure the inorganic nutrient resources in aquatic environments. He does not know what factors control the availability of these nutrients No sound theoretical concept has yet been proposed which attempts to explain biological productivity in terms of basic inorganic nutrients. The present investigation has been a part of a series of experiments initiated In 1946 (Ball, 1950) the object of which was to determine whether the addition of fertilizers to small northern Michigan trout lakes could be justified on the basis of increased fish production compared to expenditure of time and materials. considerable In the initial phase of this work biological response in the form of increased plankton production resulted from the addition of inorganic fertilizers to one of the trout lakes. The decomposition of the decaying plankton caused a depletion of oxygen during the winter following the second summer of fertilization, and because a severe winterkill of fish occurred, it was decided that further experiments should be conducted to find the optimum levels at which lakes of this type could be fertil- 3 ized without producing conditions detrimental to fish life, and yet prove to be economically feasible in terms of Increased fish production. To determine the economic feasibility of such a program It was considered advisable to find out more specif­ ically the effects of added fertilizer on plankton, bottom inhabiting invertebrates, growth of trout, and chemical and thermal features of the lakes. This phase of the work was conducted by and is being reported by Tanner (ms.). Desirable also was information on (1) how long after appli­ cation added nutrients such as phosphorus would remain in the epilimnia of fertilized lakes and (2 ) the effects of alkalinity on retention of phosphorus in epilimnial waters. The present report deals with the investigation of the last two pro blems. Six lakes, four of which received different amounts of a complete commercial fertilizer, were selected because of their similar morphometric characteristics and because they varied in hardness from 34 to 192 p.p.m. methyl-orange alkalinity. Fertilizer was applied during 1949 and 1950. The amount of fertilizer added to the lakes was based upon the degree of hardness of the lakes at the beginning of the experiments. With an increase In hardness, the amount of fertilizer added was also increased. Periodic analyses of the phosphorus content of epilimnial waters, hypolimnial waters and profundal soils were made throughout the summers of 1949, 1950 and 1951, Other data collected included alkalinity, temperature, Secchi-disk readings, percentages of organic matter in p r o ­ fundal soils, soil types of surrounding area, and results of laboratory experiments on the precipitation of phosphorus in the presence of different concentrations of calcium carbon- REVIEW OF LITERATURE That phosphorus is important in the metabolism of all living cells is a well established fact. In fresh water lakes the quantity of this element is often deficient. Many investigations have shown that when phosphorus is added to lakes or ponds, a response typified by increased plant growth is the usual result. in Europe The fertilization of fish ponds (Schaeperclaus, 1933) and Asia has been a routine and successful procedure for many years and this method of increasing food production in these areas is now well estab­ lished. In the United States, however, fertilization of lakes and ponds for food production is still In the experi­ mental phase. Ball (1948a) discussed the fertilization of lakes in Michigan and concluded that harmful effects may offset resulting benefits and that until more Information became available the general use of fertilizer should be avoided. Hayes and Coffin (1951) demonstrated a rapid disappearance of added phosphorus in an acid bog lake in Canada, and therefore could not justify fertilization of this type of lake as a profitable procedure. Einsele (1941) was able to demonstrate higher phosphorus levels in a southern Germany eutrophic lake the year following fertilization, than were present before the 6 phosphorus was added. The rate of loss of the added phosphorus was shown to be very rapid, and even though some increase in phosphorus content was observed the year fo llow­ ing fertilization, only about ten per cent of the total amount applied appeared in the surface waters. No detect­ able difference over pre-fertilization levels was evident by the end of the summer of the second year of fertilization. Alkalinity and phosphorus. The disappearance of phosphorus from lake waters is probably caused by both biological and physical processes. Conclusive evidence of physical factors responsible for its removal is lacking. Biological factors, however, are known to exist. Hayes and Coffin (1951) in a review of their fertilization experi­ ments with radio-active phosphorus conclusively demonstrated uptake by both plants and animals. They also found an increase of phosphorus in lake-bottom deposits after fertil­ ization and postulated a physical-chemical exchange m e c h ­ anism as an explanation of these results. Naumann (1932) proposed a classification of lakes based upon chemical characteristics, and suggested that lakes which contain excessive calcium (alkalitrophic) are also poor in phosphorus due to the immobilization of this element by the calcium. He found a correlation between high calcium concentrations and low phytoplankton production 7 and believed the low productivity to be due to the defic­ iency of phosphorus* Substantiation of Naumann*s concept was given by Gessner (1939) who, after a series of field analyses and laboratory experiments, concluded that in the presence of calcium carbonate, phosphorus was precipitated in natural lakes as tricalcium phosphate. Juday and Birge (1931), on the other hand, after surveying 494 Wisconsin lakes, wrote, "The results ... do not show any correlation between the quantity of soluble phosphorus and the amount of fixed carbon dioxide in the water". These lakes did not fall within the calcium ranges proposed by Naumann. According to him, phosphorus defic­ iency due to calcium precipitation should occur only when the CaO concentration reached 100 milligrams per liter. This would be equivalent to at least 179 p.p.m. methylorange alkalinity. The lakes surveyed by Juday and Birge contained far less alkalinity, having no more than 22 milligrams per liter of fixed G O g > or a methyl-orange alkalinity of slightly less than 60 p.p.m. A report by Wimmer (1929) on an investigation of two limestone quarry ponds in Wisconsin indicated that phosphorus may exist in waters which have alkalinities up to 94 milligrams per liter of fixed COg (equivalent to 213 p.p.m. methyl-orange alkalinity). Unfortunately, the published data of Wimmer do not include phosphorus and alkalinity 8 determinations made on the same samples* Nevertheless, the presence of high phosphorus values in these highly calcar­ eous waters may be significant. The problem of phosphorus precipitation by calcium also has been studied by Hubault (1943) in Europe. In one Alpine lake, which contained 76 milligrams of CaO per liter (equivalent to approximately 135 p.p.m. methyl-orange alkalinity), he found a complete disappearance of phosphorus. On the basis of his observations he suggested that N aumann1s minimum limit of alkalitrophy (100 milligrams of CaO per liter) be lowered. Presumptive approval of H u b au l t*s decision has come from the work of Deevey (1940). In a study Involving statistical analysis of a number of Connecticut lakes, this author found that alkalinity values of 122 to 148 milligrams of HCO^ per liter (equivalent to approximately 100 to 121 p.p.m. methyl-orange alkalinity) were related to decreases in chlorophyll concentration. He suggested that low chloro­ phyll production In these waters was due to the removal of phosphate Ions from solution by calcium as proposed by Naumann (1932) and supported by experimental work of Gessner (1939). Alkalinity and productivity. Limnologists have long speculated about the effects of alkalinity upon lake p r o ­ ductivity. Some advance the hypothesis that there is a 9 direct relationship, whereas others are inclined to be skeptical about any such relationship. The relationship between plant productivity and the character of lake-basin soils has been studied by several investigators* Wilson (1935) reported an almost complete absence of higher aquatic vegetation in the marl bottoms of hard-water lakes of Wisconsin. Softer lakes, with 37*8 milligrams of bound GO2 per liter (86 p.p.m. methyl-orange alkalinity), exhibited best growth of higher aquatics, whereas lakes with 20 milligrams of bound COg per liter (45 p.p.m. methyl-orange alkalinity) were very poor by comparison. Similar evidence of the inhibitory effects of calcium were presented by Roelofs (1944). In a study of four northern Michigan lakes he found lower "nutrient contents"-*- in homogeneous marl than in a marl and organic matter mixture. Recently Wohlschlag (1950) has demonstrated less growth of both higher aquatic plants and invertebrate organisms in an Indiana marl-bottom lake than was present in other types of lakes nearby. This lake had a methyl-orange alkalinity range of 175 to 202 p.p.m. A summary of data on the average standing fish crop of 18 Michigan lakes, as determined by the rotenone ■^■Roelofs used the term "nutrient contents" to desig­ nate a relative concentration the value of which depended on the ratio of the concentration of available nutrients in one soil to the maximum concentration of available nutrients found In the lake soils investigated. 10 poisoning method, was published by Ball (1948b). He compared the quantity of fish recovered with the alkalinity of the lakes, and concluded that, " . . . there appears to be a positive correlation between alkalinity and fish production. The lakes listed as hard-water lakes (above 150 p.p.m. methyl-orange alkalinity) have an average poundage per acre that is higher than for either of the other groups (soft and intermediate)". Relationships of phytoplankton abundance to alkalin­ ity were reviewed in part by Prescott (1939) and discussed in greater detail by the same author more recently (1951). In the earlier paper he classified lakes as being oligotrophic for algae when the fixed COg content was at 2.0 milligrams per liter (equivalent to approximately 5 p.p.m. methyl-orange alkalinity). Eutrophic lakes had a fixed COg content of 20 to 25 milligrams or more per liter (45 to 57 p.p.m. methyl-orange alkalinity). In his later paper he discussed the role of calcium and magnesium as agents primarily responsible in controlling the amount of available COg in water, and consequently the abundance of algae. An increase in carbonates bound by these cations, at the expense of the quantity of half-bound carbonates, was cor­ related with a decrease In phytoplankton. Burr (1941) related photosynthetic rates of algae in experimental laboratory situations to the rate of change of the CO2 11 tension, and concluded that, " carbon dioxide is generally a major factor in photosynthesis by submerged plants, a factor which has been neglected by most workers". Birge and Juday (1934) also were inclined to believe such a relationship to exist, because they found greater plankton production in the southern, hard-water lakes of Wisconsin than was found in the northern, soft-water lakes. Conclusions similar to those of the authors mentioned above were made by Swingle (1947). In pond fertilization experiments he obtained decreased fish production when limestone was added to the organic fertilizer which was applied. He attributed the decline in production to a lower COg tension caused by the combination of CO2 with the limestone. A lower phytoplankton production resulted and was followed by decreased fish growth. (1940) The work of Deevey on Connecticut, lakes indicated no statistical correlation between alkalinity and amount of chlorophyll present in those lakes which contained less than 122 milli­ grams H CO 3 per liter (equivalent to 100 p.p.m. methyl-orange alkalinity). Deevey therefore proposed that, "bicarbonate alone probably has no significant effect upon plankton production, except In hard water lakes." Te rminology. various The wide variability in terms used by authors to express alkalinity Is evident in the references already listed. Some consider CO2 as the most 12 important property of alkalinity and may express its concen­ tration as bound or fixed COg in either cubic centimeters or in milligrams per liter. Others prefer to list its concen­ tration as the bicarbonate or HCO^. Still others follow the procedure of "Standard Methods for the Examination of Water and Sewage" (1946) and list the alkalinity as p.p.m. of CaCOg as determined by methyl-orange titration. Also there is a wide variation in the use of the terms soft-water and hard-water. Table I summarizes the terms, and alkalinity ranges, for soft and hard water as published by various authors. For ease of comparison, the expression of each author has been converted to values equivalent to p.p.m. of CaCOg as determined by titration with methyl orange as the indicator. Each of the authors included in this table has select­ ed a particular range of values to express alkalinity as soft or hard because of the specific nature of the problems involved in each case. Some were concerned with the ecology of either higher aquatic plants or algae; others dealt with the distribution of fish, or perhaps with problems related to water sanitation. ents to alkalinity. and Prescott Some wished to relate dissolved nutri­ Birge and Juday (1911), Pearsall (1930) (1951) were interested in phytoplankton; Veatch (1931) and Wilson (1935) related bottom soil types to higher aquatic plants; Black (1929), Naumann (1932) and Ohle (1934) 13 CO CO COMPARISON OF THE ALKALINITY RANGES FOR SOFT WATERS AND HARD WATERS AS PUBLISHED BT VARIOUS AUTHORS CN in • CM CM 0\ d> vO VO CO vO 8 &vO 8 V in (v - % 9 CA _=J SX CM vO \-P v CM CM » ft SX f* d > in 1oA S' o 0 b) 0 § g CD l iD C o T3 S' o 5 cm 2 O CM CM in c— S S' o & o 43 § 2 S' o •o R 2o c o rH rH §3 I 8 rH vO CM t s C D S' o g e x (X o 43 I in c 2 O O P< O m © b£) CD CO 3 a ■d & vO 3 CM O rH 8 d> in 6 * cP O CD TO & O s a P* 1A c— 43 vO 1A CM CM o 43 t (X o in 3 o >O tu bO S in CM a Oh tx 2 g on 3 CM 0 ) t C© I 1A O ( W0> \ 2 CO 6 o •H 0 2 Pu uo £ p «H O 2 0 ) t ct) 1A• r— CM s w6 ® 2 c — w P K« a O *H S’ o & p 2 s o 43 oo o\ 1A cr\ Ov ft I 4 T3 **•*% CM 4 rH m Ov CM Ov m Ov H Ov rH -= t rv PJ a! i 45 a 0) > 1— 1 © 1 in u 3 ss --N TJ c nJ © 3 oo 00 © A •H U M r l 4 3 © O O Q* O *H bO a) © p o £ P O P JS CCJ CO rH °ISS a> •» T3 TO P .2 3 d i a+> w E w 3 (I) ob0 f fi h £ m c ffl ftOH6 © TG3 3§ O JH O c f l H « H CD O if B 2 CM co s O P JH o a a a, o in © rn OV rH v—’ G o ft e •H O S' t3 cq 43 tQ p -p o o CO *r) -O in CO PJ Ov <0 a. CO P W CD CO © jgftg rH CM r v d studied chemical conditions in lakes; Tressler and Bere (1935) and Ball (1943b) were concerned with fish, while Theroux et_ a l s . (1934) were sanitary engineer^ and dealt with related p r o b le m s• DESCRIPTION OF LAKES The lakes which were investigated In this study are located In the Pigeon River Fisheries Experiment Station area in Otsego and Cheboygan Counties, Michigan. Most of the region Is covered with well-developed second-growth jack pine and aspen* Portions, however, support a second-growth hardwood forest. Five of the six lakes included in this study are located in Otsego County and are within the pineaspen region. These lakes are South Twin, North Twin, West Lost, Lost, and Section Four. The sixth lake, Hemlock, which is In Cheboygan County, is In the hardwood area. The soil of the pine-aspen region, in which are located most of the lakes, Is composed almost entirely of sand. The substratum is a brownish-yellow sand with occasional gravel pockets. rapid. Seepage through the soil is Hemlock lake is located in a region which has sandy to fine sandy loam. The substratum Is reddish clay to sandy loam, with occasional gravel pockets, and is retentive of mo isture• Each of the lakes under investigation, with the possible exception of Hemlock, Is believed to have origin­ ated as a limestone ’'sink". These depressions, often nearly circular in outline, are thought to have resulted from the dissolution of enough of the limestone bedrock to allow the overlying regolith to collapse, or settle, creating the 16 so-called ,fsinkft. There are several such depressions within the Pigeon River area, some of which contain little, or no water. Each depression is characterized by being almost circular and having high, steep, sandy banks. The lakes range in maximum depth from 35 feet for the shallowest, to 72 feet for the deepest. In area, the smallest is 2.6 acres, and the largest is 5.9 acres. The deeper lakes have precipitous drop-offs beyond the littoral zone. The methyl-orange alkalinity varied from 34 p.p.m. for the softest lake to 192 p.p.m. for the hardest lake in 1948. The per cent organic matter in the profundal soils ranged from 24 to 59. None of the lakes has a surficial outlet or inlet, although Hemlock may have an intermittent outlet during periods of high water. General morphometry and physico-chemical character­ istics are given in Table II. The lakes were given consideration in 1931 and 1932 as possible trout lakes by Eschmeyer (1938). He discussed the physical characteristics existing at that time and proposed a management program based on stocking and creel census reports. It is noteworthy to observe several facts reported by Eschmeyer. Secchi readings ranged from approx­ imately 13 feet in Hemlock lake to 30 feet in Lost lake, with an average of 20 feet for the six lakes. obtained on or near August 1, 1931 and 1932. Data were If Secchi 17 *4 g f=G G O *H P vO co CM o CM O ON I pr UN ON UN CM G\ vO MD ON on o CO -p CO ,3 o o 0) UN co on o U\ un -G ~=J ON un VO On CM O ON "UN ON CM I UN «H C^- vO on 0> rH rH on r— CO f—I CM O co C M _=t On "LA vO on on On C E ! M UN C M I C M cm CO UN _G n on rH vO « vO ON "UN ON on CM CO oo ON -zr O CM un On CM co o i C M CO f C M -3 1 vO UN co O CM n UN 9 £"< O VO X P G C M vO CO C M CO UN CO -G CO i —! UN pt ON co C M UN 7? O S3 • P P • O P o <0 o *ri to p a o to © p o O j G cd XI o p P • P P C •H G *rH • n © g o xi cd p © T3 • cd p © ia 3 P © X • © > < •H rH rH G •rH • I—1 o > rH H P O t-i P o i—t d • UO © X • P 0 © P w G O •rH rH rH *H 0 G •H G O •H •H P. • © P P P o • G • a r~i o > g »H * rH O > rH -3 o P p g H © •H o p © G £ G •iH © G *H rH O O a G 5 © S s p o G O Vi vH o E rH • P P 0 o p p o p • Jxf -G P P © X) • P > C M • cd © U) G cd G O t X P © 0 • 0 P P • © t> <4 o i—I UN On G »rH ON C M • P P G © P P p © £ c •H § O o to UD G cd © G o o © C O • p <4 O *H G cdoy LCrH G *H O O w P G rH © © OX G G B P P O G • P © Q) i> o rH 'svl. rH O to E? X 5ON G rH •r-t *rH cd o to • <§ CO _G On rH C M UN On 18 readings are arbitrarily selected as the criterion of plankton abundance, the lakes would have ranged from lowest to highest production in the following order; Lost, South Twin, Section Pour, West Lost, North Twin and Hemlock. The data of 1948 (Tanner ms.), indicate Secchi readings for comparable dates of 9 to 27 feet, with an average of 16 feet. In order, from maximum light transmit- tancy to minimum light transmittancy, they were; Section Pour, West Lost, South Twin, Lost, Hemlock and North Twin. It Is evident that variation is considerable, and that conclusions drawn from limited data of this sort may not be justified, except those of a speculative nature. In the spring of 1950 direct evidence was obtained which supports the concept that these lakes are limestone ’’sinks11. A portion of the basin (approximately 30 by 60 feet in area) of Section Pour lake settled, and much of the former bank dissappeared completely beneath the w a t e r 1s surface. The entire bottom of the lake, within the pro- fundal zone, which had been a calcareous marl, pulpy peat soil, became overlain to such an extent with a clay-marlsand mixture that the former material could not be collected with an Ekman dredge. This event may mean that this lake is the youngest of the group, and in fact Is still in the process of ’’birth’*. Other factors which support this hypothesis are that it (1) 19 is the most calcareous, (2 ) has less organic matter in the bottom (changed from 41 to 6 per cent between 1949 and 1950), (5) is the deepest, (4) has the smallest area, (5) had the deepest average Secchi reading in 1948, and (6 ) had the greatest weight-per-volume relationship in the profundal soils. If these features are used as general criteria for determining ages of all the lakes, the lakes group themselves, according to the data of Table II, from oldest to youngest as follows; (1) North Twin, Hemlock, (5) Lost, and (6 ) Section Four. (4) West Lost, (2) South Twin, (3) North and South Twin lakes are located adjacent to each other with a ridge about 200 feet wide separating them. In general appearance the lakes are very similar, both having narrow sandy shores and a black, sand-peat bottom In the littoral zone. South Twin has supported luxuriant growths of potamogeton:, in the littoral and sublittoral zones every year since the application of fertilizer in 1947. North Twin had only a few scattered plants, mostly Fotamogeton n a ta n s . South Twin had an average methyl-orange alkalinity in 1948 of 74 p.p.m.; North Twin averaged 34 p.p.m. the same year. North Twin is the deeper of the two, being 44 feet, compared to the 35 feet of South Twin. West Lost and Lost lakes are approximately one-half mile from each other. West Lost contained few or no higher aquatics before fertilization. In 1951, the year following 20 final applications of fertilizer, abundant growths of Chara appeared over most of the littoral and sublittoral zones. Similar changes occurred In Lost lake, with Chara also developing In 1951. A few water lilies (Nuphar a dv e n a ) were scattered around the edge of the lake. The water of West Lost was the softer of the two, having a methyl-orange alkalinity of 138 p.p.m. in 1948 compared to 177 p.p.m. for Lost. The soil in the littoral zone of West Lost contains sand, gravel, undecomposed plant debris and calcareous particles. The littoral zone of Lost lake consists mostly of marl with the amount of organic matter increasing with depth. Lost lake is about 55 feet deep, and West Lost about 45 feet deep. The deepest lakes of the group, Hemlock and Section Four are 64 and 72 feet deep, respectively. Section Four is highly calcareous, with a very light-colored marl bottom extending from the w a t e r 1s edge into the profundal regions. Hemlock is also a marl lake, with a high percentage of organic matter in the sublittoral and profundal soil. Higher aquatic plants were scarce in both these lakes, although during 1951 Chara beds appeared over much of the littoral and sublittoral soils of Hemlock. Section Four never exhibited more than a light growth of C h a r a . Hemlock had an average methyl-orange hardness of 163 p.p.m. in 1948 and Section Four, the hardest of all the lakes, an average 21 hardness of 192 p.p.m. The presence of dissolved oxygen in the lower hypolimnion of the lakes was related to the depth of the lakes and to the fertilization program initiated in 1949. The two shallowest lakes, North and South Twin, did not receive fertilizer, and dissolved oxygen was usually detectable in their lower waters. The three deep lakes, Hemlock, Lost, and Section Four, received fertilizer applications in 1949 and 1950 and almost without exception exhibited anaerobic conditions in the lower hypolimnial waters In 1949, 1950, and 1951. West Lost, the lake which was Intermediate in depth, and which received the least amount of fertilizer of the four lakes which were fertilized, had hypolimnial oxygen the year before fertilization began, but in 1949, 1950, and 1951, had little or no detectable oxygen present. PROCEDURE Applications of commercial inorganic fertilizers containing phosphates were made to four of the six experi­ mental trout lakes in the Pigeon River area* The two principal factors that were selected as the basis for determining the type, time, and amount of fertilizer to be applied were (1) previous experiments conducted on North and South Twin lakes (Ball, 1950) and (2) the difference in alkalinities of the experimental lakes. South Twin had received equivalent to 2 p.p.m. per week of a 10-6-4 commercial fertilizer^. Although this quantity failed to produce a plankton bloom, a winterkill of fish and an oxygen deficiency occurred in the lake. Because of this, it was decided to reduce the amount applied to the lakes below that which had been put in South Twin. The quantities which were arbitrarily selected for addition to Section Pour, Lost, Hemlock, and West Lost lakes were 80, 60, 40, and 20 per cent respectively, of that applied per application to South Twin. The lake with the highest alkalinity, Section Pour, was given the greatest amount of phosphorus because it was assumed that greater amounts of phosphorus would be needed in the hard lakes to offset the possible effects of calcium ■^A 10-6-4 ratio indicates 10 per cent total nitrogen, 6 per cent phosphoric anhydride ( ¥ 2 ^ 5)9 and 4 per cent water-soluble potassium monoxide (K2O). 23 in precipitating available phosphates. ing levels of hardness Lakes with decreas­ (Lost, Hemlock, and West Lost) therefore, received correspondingly lower quantities. Fertilizer applications were made by hand from a boat to the littoral zone at approximately three week intervals beginning In mid-June 1949, and continuing until the last of July, or until three separate applications had been made. The same procedure was repeated in 1950. A summary of fertilizer applications is presented in Table III. Chemical analyses for total phosphorus contents in the lakes were begun in May 1949, nearly six weeks before the first fertilizer applications, and were continued at approximately three week intervals during the next three summers. Samples were collected at the center of the lakes from (1 ) the surface, (2 ) the lower hypolimnion within three feet of the bottom, and (3) the profundal soils. The water was collected with a modified Kemmerer water sampler and the soils with an Ekman dredge. An effort was made to obtain samples 24 hours previous to, and within three to four days following each fertilization. The samples were placed In wide-mouthed, glass-topped, mason jars, and shipped express to East Lansing for analyses. Chloroform was added to the water samples as a preservative, according to the recommendations of ZoBell and Brown (1944). Until G w 3 1 1 o p to 0H 1 1 SC o rH a o «H P 03 O T3 0 o CO 0 o i o rH 1 CD 0 0 f O rH 0 1 O rH O P rH CD CD CD CO -<0 1 1 1 1 1 o 1 o in m in in m in CM CM CM CM CM 0 0 0 0 CO PH rH CO «1 PH 1 O H Eh O M 1-0 CH Ph <; M M M 9 CO < EH rtf Cri R! P ►0 P Eh cs Pd P«H o CO r=rq Eh Pi P SC 0 1 i 1 o CD o 1 t—i i—1 i 1 O rH CD CD 0 0 1 1 o o rH p 1 1 CD CD in m CM "0 in CM m CM in LTD CM CM 0 0 -0 P 03 o 03 p *o a o c &4 !> CO 0 0 0 1 O P 1 CD g p o +rr* <1 •t o CM © ^W G d tado 0 0 CO 0 G 03 o 1—1 at bO to d X5 p d G P P o x} a, do on p p P to P Xi n © 0 X} p p P X 0 O > d 0 o S <+-« . o S p 0 p bO 0 0 0 P 0 dp 0 o o a< G 0 O o 0 0 0 1 1 t o o O 1 —I p 1 1 CD CD 0 1 O i —1 P 1 1 CD CD S.p 0 0 1 0 O rH 0 O P O P 0 p d 1 0 Oi £ 0 p © bO 0 O P d 0 0 bO O 0 G P © £ Oh © O TJ G d E * Xi CM w 1 P PH i 03 1 CD o sc 1 O p rH p 0 TO CO 0 g O d g P 0 o © O, o &5 0 o CM Ph 0 0 0 1 1 1 1 i o o o O o P •—i rH rH p 1 CD O 1 CD 1 CD 1 CD i CO O o O o O o o I —1 i—1 t—I p O o rH at 0 p 0 a CO rH 1 CD «—1 0 G P *-* rH rH at ao CO § £> CM s t> CM rH S>> rH 0 P P P at CO P 1 0 p 0 G P P P P at P at CO 1 o- p p p a, at © —' ,d m Xi © 0 > P 0 O 0 T3 0 P o & CM O PH P 0 G G 0 & d 03 03 03 o o o "0 0 0 in in in 03 03 03 03 03 03 rH O G CQ 0 cd at CO CM t © G T3 © P p 0 p CuOCM G P CM 0,0 0 to t» 0 CD P P O P ,G o (X 0 O G 0 O, O 0 XI 0 h 25 determinations were completed, the samples were stored under refrigeration at 40 degrees Fahrenheit. Delays in the transportation of collecting materials and breakage during transportation accounted for the loss of a few samples. The 1951 water temperatures were obtained with a Foxboro resistance thermometer. The 20 centimeter Secchi disk was used for light penetration determinations. Alkalinity was determined according to f,Standard Methods for Examination of Water and Sewage” (1946), and is reported as p.p.m. methyl-orange alkalinity. This is a total alkalinity in terms of milligrams of CaC03 per liter of solution, and is equivalent to the total milli-equivalents of acid used in titration with methyl orange as the indicator. Samples of bottom soils were taken from the littoral, sublittoral and profundal zones. The littoral zone included the portion of the bottom from the water's edge to the 12 foot depth contour; the sublittoral zone included the p o r ­ tion from the 12 to the 30 foot contour, and the profundal zone from 30 feet to the maximum depth. These zones were selected because they represented average summer depths of epilimnion, thermocline, and hypolimnion for all the lakes during the years of fertilization (1949 and 1950). Soil samples collected from the watershed area surrounding each lake were analyzed for certain available and reserve nutrients. The soil samples were composite 26 collections taken from sites located (1 ) approximately half way up the bank from the water's edge to the top of the bank, and (2 ) at regular intervals around the periphery of the lake . Laboratory experiments were conducted to determine approximate levels of CaCO^ alkalinity at which phosphorus is precipitated. Calcium bicarbonate solutions were pre^ pared by dissolving CaC03 in a carboy containing carbonic acid. The acid was made by adding pieces of dry ice to distilled water. In one group of experiments, five different concentrat ions of alkalinity, ranging from 24 p.p.m. to 212 p.p.m. were prepared and KH2FO4 was added to each. 0.1 p.p.m. phosphorus as The solutions were placed in Fernbach culture flasks, kept in the dark, and examined at periodic intervals for methyl-orange alkalinity, phenolphthalein alkalinity, pH, and phosphorus content. experiments were concluded after 95 days. The In another group of experiments, calcium carbonate solutions which contained different concentrations of phosphorus were agitated in a laboratory shaker for a period of five days, or until phosphorus precipitation occurred. Alkalinities, pH, and phosphorus contents were measured at periodic intervals. LABORATORY METHODS Total phosphorus of lake w a t e r s . The method used for the determination of total phosphorus in lake waters was that of Ellis, e_t aJL. (1946). A 100 milliliter sample, or an aliquot, was digested with sulphuric, nitric, and hydrochloric acids. Following digestion the phosphorus concentration was determined colorimetrically with a KlettSummerson photoelectric colorimeter. Color development was accomplished by adding a molybdate salt which forms a phospho-molybdate complex. Upon the addition of stannous ions, this complex is reduced and a blue color develops. An absorption filter which transmits light waves in the 640 to 700 millimicron spectral range was used in the color­ imeter. molybdate Absorption of these rays by the blue phosphocomplex follows Beer's Law. The amount of light absorbed therefore Is proportional to the concentration of the complex, which in turn depends upon the concentration of phosphate ions present in the solution. The Klett-Summerson photoelectric colorimeter is calibrated in units which correspond directly to optical density units, thus enabling the operator to determine with ease the concentration of the unknown substance. The phospho-molybdate concentration is directly proportional to the units which represent the optical density of the solution and therefore concentrations 23 may be read from the machine without the need of making a standard curve of the per cent of light transmitted for known concentrations. Phosphorus of profundal soils. Soil samples taken from profundal regions were analyzed for both available and total phosphorus. The soils were prepared for analysis by drying In a water bath at 60 degrees centigrade. They were next ground with a mortar and pestle until reduced to a size small enough to pass through a number 35 sieve (0.50 millimeter size opening). Available phosphorus was determined by a photo­ elect ric -colorimetric method according to Bray and Kurtz (1945). This method, as suggested by the authors, is superior to the "quick-test techniques" which do not involve precise measurements of either soil or color development. They found this method to be a quantitatively acceptable procedure for the determination of adsorbed and acid-soluble phosphates. Phosphates were extracted with an ammonium flouride, hydrochloric acid solution. A phospho-molybdate complex was then developed by the addition of an ammonium molybdate reagent. Color was produced upon the addition of stannous chloride and concentrations were then determined by comparing optical densities of unknown solutions with optical densities of known solutions. The Klett-Summerson color­ imeter was used for this analysis. 29 Total phosphorus content of the lake profundal soils was determined by (1 ) using the perchloric acid digestion technique as recommended by Piper (1950, p. 272 and p. 294) and (2 ) developing a phospho-molybdate color according to a modification of the technique used by the Soil Testing Laboratory of the Soil Science Department, Michigan State College. This method utilizes the organic reduction reagent, amino-naphthol-sulfonic acid. The Lumetron Photoelectric Colorimeter was used in this method. It was found advisable to modify the procedure by neutralizing the acid digest with a sodium hydroxide solution. phenol was used to indicate neutrality. then Para-nitro- This solution was analyzed directly for phosphate content. Organic matter of profundal soils. The per cent organic matter In the profundal soils was determined by the wet combustion procedure as described by Peech, e_t_ a l . (1947). This method involves the oxidation of acidified organic matter upon the addition of a measured quantity of a dichromate solution. Titration with ferrous sulphate to a diphenylamine end-point completes the reduction of the dichromate. The milli-equivalents of dichromate reduced by the organic matter may be calculated and an estimate of per cent organic matter obtained. Inorganic nutrients in watershed soils. The "rapid test" techniques of Spurway and Lawton (1949) were used to 30 determine what they refer to as "active" and "reserve" nutrients. These techniques are useful when estimates of available and reserve nutrients are desired. They are not designed for precise quantitative work, and results there­ fore must be judged accordingly. The "active" tests utilize an acetic acid extraction solution. The "reserve" tests are made with a hydrochloric acid extracting solution. In addition to the determination of active calcium and magnesium contents according to the published techniques of Spurway and Lawton, estimated reserve quantities of calcium and magnesium contents were also obtained by neutralizing the hydrochloric acid soil extract with sodium hydroxide after the color-development reagents had been added. Concentrat ions of these two cations were then determined by ordinary visual colorimetric comparisons. pH of laboratory experiments. All pH determinations in the laboratory were performed with a Beckman Model H 2 , Glass Electrode pH Meter. Volumes of lake -bottom soils. The wet volumes of the soil collected from the littoral, sublittoral, and profundal zones were measured by placing them in a graduate cylinder and allowing the soil and the supernatant fluid to separate. RESULTS OP FIELD INVESTIGATIONS The determination of the amount of fertilizer which was to be put into each of the experimental lakes was based on fertilization studies which had been previously conducted on South and North Twin lakes* (See PROCEDURE) The theoretical concentrations of phosphorus which should have been present as a result of the addition of the fertilizers to West Lost, Hemlock, Lost, and Section Four lakes are listed In Table IV. In calculating these concentrations it was necessary to make the assumption that all added phosphorus would go into solution and would become distributed uniformly throughout the lake. The lake receiving the least amount of fertilizer, West Lost, should have had an increase of 24.2 micrograms of phosphorus per liter; Hemlock, Lost, and Section Four should have had increases of 40.6, 80.8, and 94.5 micrograms of phosphorus per liter respectively. With one exception, these are the amounts for each application, (The first fertilization in 1949 was slightly higher than the others). For comparative purposes it is interesting to note average amounts of total phosphorus present in surface waters of different lakes as published by various authors. In terms of micrograms per liter the following concentra­ tions have been reported: 23.0 in Wisconsin lakes and Birge, 1931); 14.9 in Connecticut lakes (Juday (Deevey, 1940); TABLE IV THEORETICAL CONCENTRATION OP PHOSPHORUS RESULTING PROM APPLICATION OP PERTILI TER1 Units in micrograms per liter Period of analysis We s t Lost Hemlock Lost Se ctIon Four 1949 June 16-18 29.0 43.5 87.7 113.5 July 24.2 40.6 80.8 94.5 July 27-28 24.2 40.6 80.8 94.5 Total 77.4 124.7 249.3 302.5 1950 June 14-16 24.2 40.6 80.8 94.5 July 24.2 40.6 80.8 94.5 July 27-28 24.2 40.6 80.8 94.5 Total 72.6 121.8 242.4 283,5 150.0 246.5 491.7 586.0 7-8 7-8 Grand Total 1-South Twin lake received 93.99 m l c r o g r a m s o f phosphorus per liter per application during 1946 and 1947. Five applications were made each year for an annual total of 469.9 micrograms per liter, or a grand total of 939.9 micrograins • 33 21.0 in Linsley Pond, Connecticut (Hutchinson, 1941); and 20.5 in the Schleinslee, Germany (Einsele, 1941). Thus, on a relative basis the quantities of phosphorus added to the lakes in the present study are seen to be at most five times as great as the levels which are ordinarily encountered in many fresh-water lakes. Actually, the amount added to West Lost lake was approximately the same as that which normally exists in many natural lakes. Phosphorus concentrations in surface w a t e r s . Total phosphorus concentrations which were found in surface waters of the lakes during 1949, 1950, and 1951 are presented in Table V. No 1949 data are available for South Twin except for one sample collected late in August. A comparison of the data of the two unfertilized lakes with the data of the four fertilized lakes indicates (1 ) that there were higher total phosphorus averages in the fertilized lakes than in the unfertilized lakes during the first year of fertilization, (2 ) that the total phosphorus averages In the unfertilized lakes were as high as the fertilized lakes in 1950, and (3) both fertilized and unfertilized lakes contained relatively low phosphorus con­ centrations during 1951, the year following the last fertilizer applications. The abnormally high values which appeared in the unfertilized lakes during 1950 are correlated with o Pn d o 0 t> CO 0 0 O • rH LO CO I s- -sf 05 to l> O to H H CD ^ O CM 05 lO CXI O O lD O O O rH K )H O O O ^ CO CO O W O H O CO W 0 0 b H IT) H CD CO 03 ^ O LO O O O O ^ • • # • • • • • • • 0 1 0CO LO 0 tO rH C" LO rH to rH • • 1 0 O £> LG O O © © H ^ 10 CD CD CO CM CM 40.9 fu d •r-t -p o © P CO lO £> ^ CO 03 03 CD CO O ►4 ^ H J> LG O O CO Cs3 CO lO CD CO LO o 1 0 CO iH 03 47.0 co CO o O K P- CO d © rH CD H MMiOOOO CD O o- CM CD LO t" o> CO LO CM CO CO «-H CO o CM ^ 05 CO CO CD H O CM CO CM LO 05 CO fH CO CO H CO E " o - O CD CM ^ 05 CO IH o LO CO O O •vt* ^ CD CO O ) LO CM CO ^ H LO O CM CO CM CD O C j H H H 05 CM E> 05 O O CM LO to O LO O O CO LO O O o O' O LO LO O LG CM 'M1 £> £> LG D- E> © ro rH H W 1—I rH rH rH CO rH rH p •n; E -* O Eh PI . i—1 u © Q4 p © o PI p CO TO E © cd O O LO o O LO O O O - iO O r~ H H CM H ' 5f H O O 05 O CO ^ £- 29.4 o t d P -, LO O •O * • • • CM LO • O to CM tLO o I—I E 83.1 P cel Ph E d •h £ Eh d ,d o o •H CO £ w o So o o • O O O ■ • LQ P p CO •O d o CO •H d P £ (H • «••*• LG • 05 E> £> CD P 27.7 P-< PH o fzr, m •H CO >> rH © d ©

rH 1 rH 1 — 1rH CM CM 1 CO rH CD CD 1 L" CM O CO 1 05 CM CO CM 1 CM CM * © © d d rH rH bO © d d d d d £3 p p *“3 P> < © to © p © > CM 1 CO CM LO 05 »— 1r— 1£> 1 1 1 'V1 CO LD rH 1 —t 1 — 105 rH 1 O rH CM 1 CO CM • • © © r*9 > > P d d rH rH t o jd d d d d d **=H p © to © P © > ■— 1CO 1 1 CO CM CD rH • p © a •rH p 05 o LO LO © PH 05 05 05 rH CO 1 O CO 1— 1 CM 1 O CM K^' • © d rH rH t o d d d d P) p >P) © Average *< EH to © P © > Grand PO 51.6 P b£3 o a 34 35 exceptional climatic conditions of this year. heavy throughout the spring and summer. Rainfall was Furthermore, the soil was probably especially permeable to water because of unusual weather conditions which prevailed the previous winter. An early autumn freeze, and a sudden spring thaw kept the melting snow from soaking into the ground as normally occurs. Consequently the entire w i n t e r ’s accumula­ tion of water ran off into nearby streams and left the soil in a comparatively dry porous state. assume It seems logical to that the three inches of rain, for example, which fell during the first eight days of July might well have percolated rapidly through the soil and in so doing leached considerable quantities of nutrients from the soil and Into the lakes. Evidence supporting this view is the general increase in plankton which occurred in both North and South Twin lakes In 1950 (Tanner ms.). In 1951 lower phosphorus contents were correlated with lower plankton concentrations. Gradual increases in lake water levels, and continued decreases in alkalinity (Table VI) are further evidences of important changes occurring in the lakes. The amounts of phosphorus which appeared in surface waters of West Lost lake In 1949 averaged about three times as great as those of North Twin. The average for the summer In West Lost was 20.0 micrograms per liter. In 1950 and 1951, this lake was very similar to North Twin, both 3C & o o> o lO f— 1 rH t- 00 H4 03 «H 1— 1 LQ rH H CO P O CO 1 — 1 P 00 > -P •H c P > W 3 CQ Eh CO rH C tJ p dj I'* rH C fi <=£ Hi JS Hi M PM o r’H Eh Hi Hi < w Hi < E J CO CO t> i— 1 p 1 — t p 03 CO ' M o o rH E C D tn C D to £j «5 co p p t> IQ P 03 lO P CD P P LO CO r-H CO CD 03 P rH P to CO 03 o I— 1 P tH P i— 1 CO CO to 03 P CD CO P to CO o p o C\3 i— I 03 LO CO CO 1 « — 1 i— lO I— 1 o 1 — 1 p 03 03 r— 1 03 LO o 03 1 — 1 P t> o p CO o p 03 P r 1 rH I> r> CD CD CD LO CO CO CD ID CD CO CD CD E "~ E> CD 03 CD 00 E> CD 03 LO xjH LO 03 CD to P CO P to p CO 03 co CO O CO CO £> 03 P to O CO 03 CD 03 03 CO 03 ■st4 03 r— t P 03 o rH p 03 o Hi u o 1 1 — I t>> o: P 03 03 S: p — p 8 C 3 •H £ Eh • E • ox » Cu P 2 O CO c: p 00 P •H C! to £3 P & E Eh X P O is — r 1 CD 03 1 p 1 03 r— 1 to to do P l CO Ol t>> p 2 Hi 03 rH to 1 LO p • bp p 03 p H O lO 03 P P i to 03 Hs !>s P P P H P H 03 03 1 C O p • to p GO 1 CD I Q; 03 t—1 03 1 O 03 P i— i # bO P C D bfl cd f-r © > *=d 03 P 37 Increasing to about 60 micrograms per liter in 1950 and decreasing to about 7 micrograms per liter in 1951* The averages of the means for the entire three years were 27.7 and 29.4 for North Twin and West Lost respectively. Thus, although West Lost received about 24 micrograms of phosphorus per liter per application for three applications during 1949 and 1950, and North Twin received none, the overall three year averages for the two lakes were essentially the same. In 1949 the addition of phosphorus to those lakes receiving fertilizers was evident by the greater quantities of phosphorus found In the epilimnial waters of these lakes. This situation was only partially true in the second year of fertilization and was not particularly evident during the year following phosphorus applications. Hemlock lake had the highest averages for the two years of fertilization, even though it received less fertilizer than Lost and Section Four. Hemlock is high in organic matter in the sublittoral and profundal soils, and the water possesses a distinct brownish-color. The organic-matter content of this lake is correlated with the high average total phosphorus for the three years* This average was 83.1 micrograms per liter as compared with 47.0 for Lost lake. These levels are interesting when compared to the amounts of phosphorus which were put into the lakes. Hemlock lake was fertilized at the rate of 40.6 micrograms per liter, and Lost lake at the rate 38 of 80*8 micrograms per liter. It is apparent that there must have been fundamentally different physical and/or biological conditions existing in the two lakes. In Hemlock lake the three year phosphorus average was about twice as high as the average amount of phosphorus added per applica­ tion. The other lake (Lost) only contained about half as much on the average, as was put into the lake per application. Section Four, the lake which received the greatest amount of fertilizer, had the greatest epilimnial phosphorus concentrations of the six lakes on only three occasions, two of which were during the first year of fertilization immediately following fertilizer applications. time was in 1951, tion. The third just 12 months after the last fertiliza­ The averages for this lake were 51.5, 59.7, and 11.6 micrograms per liter for the years 1949, 1950, and 1951 respectively. The overall average for all samples examined during the three year period was 40.9 micrograms per liter. Six applications of phosphorus, each of which averaged about 95 micrograms per liter, were added to this lake in the two years of fertilization. Thus, the three year average was only slightly less than half the amount added per application, and the average for 1951 was only 11.6 micrograms per liter. Although no averages for the years preceeding fertilization are available, It is probable that little or none of the phosphorus present in 1951 can be due to the phosphorus 39 added by fertilization. This conclusion is based on the fact that the lakes receiving no fertilizer, North and South Twin, averaged 7.7 micrograms of total phosphorus per liter, or essentially the same as Section Four. Phosphorus concentrat ions in hypolimnial w a t e r s . Samples of water collected from the bottom of the hypolimnion were analyzed for total phosphorus. collected The samples were from within three feet of the profundal soils from a position approximately at the center of the lake. The data for these samples are summarized in Table VII. As demonstrated by Juday and Lirge (1931) and by Hutchinson (1941), as well as many other investigators, low­ er waters of stratified lakes usually contain higher concen­ trations of phosphorus than are present in surface waters. In the first year of fertilization, phosphorus concentrations in the lower hypolimnial waters, with one exception, averaged about the same in all the Pigeon River area lakes. The exception was Section Four. This lake averaged 21 micrograms of total phosphorus per liter. North Twin, West Lost, Hemlock, and Lost averaged slightly more than twice this amount. Section Four 3nd Hemlock averaged higher in phosphorus concentrations in the epilimnion than in the hypolimnion. and Lost. The reverse was true in North Twin, West Lost, These three lakes thus exhibited conditions normally found in lakes, as mentioned above. The high Fh CD Q-, m g od U W to . LO LO • LO O O . . . . . CM • t> o o to to to CM to . CM O to CO 10 to o o o o o C~~ CM O CD O to CM CO O H to to to H O O O O O l O to C V lc o O O O EH co to to to ^ O) CM to ^ C- O o o o o o o o to • CM lO CM CO O LQ LQ O rH O O t> CO CM GO O G i H Z> O O O O O to CO 05 05 O to O O CMLQ to ^ C O O C O H t> CM CM CMCM rH C- E> O LQ O M 1 H LO LO tO CM ^ tO ^ CM O C M O O O O ^ 05 O tO O O LO tO £> CM C O C- tO lO tO tO C OCM to CO to O to rH CM tO C“~- tO ^ to to O t o O O O LO to O rH O CM CO ^ to H H a J r H tO O O i^ ^ H rH E> to 00 to to O 05 -P to o -p CO a> lO LO O LO o . • . * . O E> LO CM LO rH LQ O O' LO . to LO 69.3 F-t CD -P •H rH o o r— i E to to lO E> LO to O O CO O lO o H CJi t o ^ o o •H £ C! •H m -p «rt C! •H & Eh A -p P* o CO . • • * LO LO . * • • * . • • • • CM CM i— 1 i— 1 • » . . . . . • • ^ •• . ^ rH • to CM 46.5 u a •H £ Eh lO O LO O LO . . . • * O rH CM LO O tO LO LO tO H E> * O LO O O O O O O LO LO C- to 05 'COCO t - t> ^ LO O M 1^ to LO CM CM rH O CM CM CM O CV1 CO o 40.3 P3 Cm m O l>» t>> • CJ C! H rH bo Pi 2 PS P PJ h b h b c a> 05 <— 1 M*LO O) rH 05 rH rH C M I I I t> I I CM a> to aJ Fh 0 t> •ci* 00 I O CO CM rH rH tO rH CM to ^ U aJ s o l>> • I CM tO rH i— t tO CM 00 t 1 to o to to CM *"3 LO LO 05 05 O CM i—1 i— 1 CM 181.6 Fh P* O Ex, 63.6 40 41 concentration in the epilimnion of Section Four and Hemlock was presumably due directly to the addition of the commer­ cial fertilizer. North Twin showed no unusual deviations from the expected epilimnial-hypolimnial relationships, and of course, was not fertilized. West Lost, although receiving some fertilizer, probably didn't get enough to alter normal relationships. Lost lake, which received 80.8 micrograms per liter of phosphorus per application, did not show, at least during the first year of fertilization, an increase proportional to rates of applications in either surface or profundal waters. This leads to the speculation that the fate of the 250 micrograms per liter of phosphorus, which was the total added in 1949, was adsorption by the bottom mud s , Phosphorus concentrat ions in the hypolimnia of all lakes averaged higher the second year of the investigation than during the first. Furthermore, there existed a general increase, with one exception, from lake to lake correspond­ ing to the amount of fertilizer applied. Section Four received the greatest amount of phosphorus, but ranked about the same as West Lost, the lake which received the least. The highest level of total phosphorus recorded during 1950 was 640.0 micrograms per liter. This occurred in Lost lake in early July. Even greater hypolimnial differences In phosphorus 42 contents between the lakes appeared in the summer of 1951, North Twin was exceptionally low and Lost lake was excep­ tionally high* A maximum of 21*0 micrograms per liter was recorded for the former, and a maximum of 1,180.0 micrograms per liter was recorded for the latter* Two of the fertil­ ized lakes averaged higher than the previous year, while two were lower. Hemlock and Lost showed considerable increases while West Lost and Section Four decreased, but only sl ightly. Changes in concentration of total phosphorus in surface w a t e r s . Table VIII has been compiled to facilitate the interpretation of absolute changes in total phosphorus which occurred in surface waters of the lakes during certain period s . If the data of 1950 are excluded from consideration for the moment, and if periods which included fertilization are also not considered, the degree of variation in total phosphorus levels due to normal causes may be analyzed. North Twin exhibited an increase of 11.0 micrograms of phosphorus per liter in a three-week period in 1949. In 1951 the greatest increase in a three-week period occurred in Lost lake, where 20.2 micrograms of phosphorus entered the epilimnion between the middle of June and the first week in July. The normal rate, and degree, of decrease of total phosphorus In the surface waters was just as great as the ca HD co o d O o • • • • • • o •H d - d a - CO XA XA rH P d t vO rH XA XA - d O o 1 + 1 I + 0) f t co co XA o OJ CO o XA CA c a | • • • • • •1 • • • • CA _d* O ' o 1 XA CA CA O' W te S« M »d HH ft ft a PQ < EH ft o P o w Eh C OS Eh 1*5 M O 13 O o 3 M CO ft e> (3 O • • • XA XA \A • • • rH OJ OJ OJ S + cA O ' a> + -d" w + • • • XA XA O O • • • • OJ XA CA XA i—1 CA i—1 CA 1 1 rH + H* O vO TA XA XA XA HD • • • • • • • o o vO _d* XA O CA 1 + rH OJ C" CA rH 1 + rH 1 1 • I fl 0 •H 4—I d 9 rH d p • • • O XA CA rH rH H- 1 O CA • • P + rH 1 1 • O o rH 1 CA - d • • O' _ d I CA 1 rH XA CA rH 0 d d P ^9 fc*"9 • jO rH rH C d hs d P * * CA -d* XA • 60 O) 1 CA OJ • rH 1 * CA rH 1 OO rH d 0 d d d 0 d -d is P • CA + • O + rH + rH XA CA rH i—1 XA CA i—1 rH XA CA rH d 9 se 0 d d P + rH XA CA rH •» »—1 OJ 1 1 O O 1 CA OJ I rH CA y l>9 >9 I—1 rH d • CA d •y t*)0 dfl P «*s rH XA CA rH rH XA CA rH * * *• i—1 CA •• rH OJ CQ 1 * i 1 1 1 sO O CO CA OJ -o rH OJ rH •k «—1 o o O o XA XA XA XA CA CA CA CA O rH rH rH rH r—1 CA -d * d • • • O rH 1 * * «* CA OJ t T OO * 1 1 1 CO CA OJ vO OJ rH 0 C 3 P rH vO XA O ' OJ * rH rH 1. O rH * OJ • • I •• •* a XA CA rH rH O1 1 1 CA - d oo OJ rH rH O XA CA rH o OJ OJ 1 + O XA CA rH OJ 1 1 CA OJ OJ OJ d vO -d - ao O XA CA rH CA rH o o XA CA i—( O XA CA rH « * ho d d d 9 d d rH rH d rH rH d d d d 0 d d d p p p 1 O o rH 0 ® ca • • • • • • • o - d + UN • OJ ft 1 • « • • • • X A X A CO ao CA - d o o • • • • • • • • O J C^- c o f t - d * - d _ d A OJ ft 1 f t O J (— C O C A 1 i + + 1 + + • • • X A 'LA • • o -d ft ft + + o o o o co 1 • CA ft + XA XA • • A - OJ OJ + + o • OJ OJ 1 co CO o • • • XA _ d _ d ft C A C O 1 + + XA • OJ ft + • u • • • • • • CA _ d O \D XA - d C O OJ c o + CO 3 i 1 ft + + i O • ft ft ft CHANGES IN CONCENTRATION OP HYPOLIMNIAL 03 d Fh O Jd • • * CO ft X A X A • • 9 * f t G O 1— t C A OJ v Q ft - d 1 + + + XA « vO ft 1 'LA X A X A « * • ft OJ A ft _ d + i + XA O o o • • • • ft CO vO vO co _ d vO + i + + c a CA _d CA ft • • • o o • CO CA 1 co co C A C A CO • • • • • • c— X A C O f t C A X A 1 CO CO CO CO CO • 1 I 1 + o ,d ft ft x i d o co aS < Fh bO O Fh O •rH e a ft Fh * O Eh *25 - d c a ft CO -p ft d 'd d o CO o ID CD 03 CO Ai rH CO rH >-» CO ft o 03 xJ o ft Fh (D -P f t 03 Fh ft ft * •* OJ C O OJ OJ 1 1 C O c— ft OJ CA CA - d -d CA CA ft ft m o no 1 CA OJ •V CO OJ 1 C\J 0J vO oo - d o vO o • * * • • • • OJ CO o ft C O c o o 1 ft C O + C O OJ ft 1 1 + i 1 o o XA CA ft o XA CA ft o o XA XA CA CA ft ft o XA CA ft ft ft o XA XA XA CA CA CA ft ft ft •k d OJ I co OJ «k XA ft 1 - d ft •t •k CA f t t— 1 1 CO vO ft •k rH ft 1 O ft •k •k «k CA - d ft C O OJ •k i 1 1 C O C O OJ v O rH OJ a> d d ft Fo ft d F”s • ft hQ d d ft • Fh a) 35 CA CA - d - d CA CA ft ft CA CA -d - d CA CA ft ft CA - d CA ft o o o XA XA XA CA CA CA ft ft ft •k co OJ 1 0J OJ •k -d OJ i CO 0J •s XA ft 1 _ d ft •k •k CA ft A 1 1 co v O ft • bO d < • Fh CO IE 03 d d ft 03 d d ft a? d •“0 •> C" ft 1 ft ft «s OJ OJ 1 CO ft •t OO OJ 1 AOJ > A * ft ft d d ft ft •k o co 1 CA OJ > Fo ft ft d d ft ft o XA CA ft • bQ d • Fh CO s ft ft d d ft ft ft O o XA XA XA CA CA CA ft ft ft ft XA CA ft •k ft ft 1 O ft •k •k - d ft C O 1 t OJ v O ft «k CA 0J •k 1 ao CO 0J • F o F*» • ft f t b O F h C0 d d d ISE ft ft C3 d d ft •k ft OJ 1 o OJ • bO d <*J ft ft XA XA CA CA ft ft ft d ft •k ft co 1 o CO ft d ft 47 48 lakes. In the "unfertilized lakes, during the more normal years of 1949 and 1951, the maximum increase in a three-week period was 39.9 micrograms per liter. in South Twin lake. This increase occurred In 1950, North Twin gained 66.0 micrograms per liter in only five days. The Increase is associated with other unusual data already attributed to exceptional climatic conditions. The maximum increase in the fertilized lakes In 1949 was 40.0 micrograms per liter in West Lost following the application of a theoretical concentration of 29.0 micrograms per liter. maximum increase was in Lost lake. following the first In 1950, the In a three-week period, phosphorus applications, 318.0 mi c ro ­ grams per liter entered the waters of the hypolimnion. Hemlock lake had the greatest rate of phosphorus increase in the summer of 1951, with 233.0 micrograms per liter more appearing in early July than was present three weeks earlier. The degree of phosphorus fluctuation In the hypolimnia of the lakes, as shown by data of Table IX, is seen to be of a very considerable magnitude. may accoifnt for these fluctuations are Factors which (1) failure to duplicate precise vertical positions In lakes when collect­ ing consecutive samples, (2) changes In redox potential which affected solubility of inorganic phosphate colloids, (3) turbulence of waters adjacent to profundal soils caused by lake seiches, (4) vertical migration of plankton, 49 (5) depositions of organic and inorganic matter from upper layers, (6) influx of subterranean water, and (7) deposition of a silt-load carried by run-off water following heavy rains. Available phosphorus in profundal soils. Samples of profundal soils collected from near the centers of the lakes were analyzed for available phosphorus. To increase the quantitative precision of the analysis, the moisture content of the soils was removed from the samples by drying in a water bath in the laboratory. Weights accurate to ten milligrams were thus obtainable. Data for these analyses are given in Table X. One of the outstanding features of these data is that there exists among them considerable quantitative variation. There were great differences not only between one lake and another, in the same lake during different years, but also during a single season in one lake. account for these differences. Many factors may A statistical analysis of the variability seems adviseable before discussing these factors. Statistical comparisons were made by utilizing a graphical method developed originally by Dice and Lerass (1936) and outlined more recently by Simpson and Hoe (1939). Standard errors for these comparisons were computed on the basis of the number of samples, and the standard deviations derived by an analysis of variance technique. One analysis Four 50 -H OO Section _d-oj-d*A-ooo H OJ H ^ O I A O COO 0 _ d ^ rH OJ OO Lost West Lost • OJ rH _ d " i—I i —I +1 *0 0 * 0 0 *_d- • OJ rH rH ( —I • i —I • i —I • +1 if CO C O A - C A rH OJ CO X A 0 0 \ A d - H A A d - sO A- CO vO OO • OJ i 1 HH XA CA • HH O CO OJ +1 OJ rH C A O OO O C A C A OJ rH rH CA CA C A OO OJ o O '-O - d 'C O 'L A X A A - CO r H _ d - r H O CO o j oj oj o j o j _ d - r o o j +1 • • ~j~ * • XA O CO • v O OJ A - •-d-UMA •t— oo • COCA +1 OO AOJ Twin vO XA rH OJ A - O O '-O OJ OOOJ OO oo_d*oooo OO CO OJ oo rH HH OJ XA HH XO XA oo CA O CA -sO HH rH sD OJ OJ C A O J C A c O A - A— — »—1 • • CO • OJ A- • oo HH XA v£> m o fH Jh 0 T) nO North CA • rH +1 A- OJ rH ♦ XA XA -d" OO OJ South Twin +1 0s * C — 0— rH _d " UNPOCA-d-OJ OU\vDvOoo H-l if o OJ rH CO OOnO X A OJ CO t O OJ OJ Hemlock of air-drysoil^ IP— — —— CA rH rH OO rH -d - kilogram CO OJ OJ ca rH +1 _d"CA_d*oj co_d“iA a- o _ d " r H r O _ d -'-d '^ -A _ d 'C 7 N rH rH H H rH rH rH • A— CA OJ OJ * CA O O O O O — •O J * OJ rH OJ X A ♦O J * OJ OJ rH •vO +1 vO CO rH OJ A - v O rH A OO r H X A ^ O OJ OJ OJ rH +1 CO rH OJ rH • OO rH HH rH A- fn TJ fl 0 0 0 -P 0 d rH O Period of Analyses i|. ■■■— «. — — n i HH CO (— o O rH C— CO OJ vO in milligrams per Units ■»»» ^I,||,IW| 111 11 ■ — — SOILS OJ CO • OO AVAILABLE +1 -d " vO TABLE X PHOSPHORUS IS PROFUNDAL oo XA vO XA a -oj rH OJ II o oo OO OJ II IH CO A - C A O J X A C A r H iH O J O J C M O J O J C V j 0 bO 0 0 bO 0 © 0 ^>s '•♦ • • C C rH rH bO bO t O P , 0 t> d d d d d 0 td-3(d-3 h3 »“3 «J* csj; <: CO CA _d~ CA rH OJ H 1 t— 1 C O X A CO 1 O OJ rH rH v 0 rH • (D G G 0 d ^ , b O XA CA rH rH rO 1 OJ O 1 OO rH CA OJ 1 CO OJ © (>) t>» * • C H H H bo bO d d d d d d h h > b b < < 0 bO 0 U 0 > < _d“ rH 1 GO OJ 1 rH v O rH rH OO OJ 1 1 OO CA OJ rH © Aa C H H d d d b b b rH XA CA i—1 ♦ bO d < u © bO 0 IU 0 > < C 0 0 to 0 > 0 tj > 0 c 0 u C5 H 51 of variance was computed for the results of each lake over the three-year period. An analysis of variance of the combined data of all the lakes together was not made because of the great differences in variability between the data of one lake and that of another. The standard deviations were computed from the variations existing "within years” for each lake, and serve as the basis for the standard errors used in the graphical presentation. Each vertical rectangle in the graphs (Figs. 1 and 2) represents twice the standard error on each side of the mean. The center of the rectangle is located at the position of the mean of these data, and the range of the data is indicated by the length of the vertical line^ passing through the rectangle. ence A statistically significant differ­ (at the one per cent level) may be said to exist between two groups ,of data when the vertical extremities of the rectangles do not overlap in the horizontal direction. If the vertical ends do overlap, the averages of the data represented by each rectangle probably are not significantly different at the one per cent level. Thus the reliability of the differences existing within the data of each year for each lake which is listed in Table X may be judged by reference to Fig. 1. To facilitate comparison of overall average phosphor­ us levels in different lakes over a three-year period, a Fig. 1. Seasonal variation of available phosphorus in profundal soils. dry soil. Units in milligrams per kilogram of air- 52 mg/kg in o> 400 O in cd 300 in cd O in 5 ± SEC. FOUR Fig. 2. The 1949, 1950, and 1951 mean average and variation of available phosphorus in profundal soils. in milligrams per kilogram of air-dry soil. Units 53 mg/kg 400 hco _i 300 cn LU X I” QC 200 o UJ X li. o cn UJ 54 second graph similar to Pig. 1 has been prepared. In Fig. 2 the mean of each lake is the average of the means for each of the three years of data. An examination of Pig. 2 indicates that differences in the three year averages of available phosphorus in the profundal soils between most lakes was significant. West Lost, the lake which was given the least quantity of fertilizer, averaged 252 milligrams per kilogram of soil, the highest of all the lakes, whereas Lost and Section Pour, the lakes receiving greatest quantities of fertilizer averaged only 7 milligrams and 15 milligrams per kilogram, respectively. North Twin, although not fertilized, contained more avail­ able phosphorus than any of the fertilized lakes except West Lost. South Twin, an unfertilized lake during this period, contained essentially the same amount as Hemlock, a lake which was given considerable quantities of p hosphates. The statistical analyses of the averages of each lake for each of the three years have been presented in Fig. 1. (1949, 1950, and 1951) The variation which occurred in the control lake, North Twin, betv/een 1949 and 1951 was great enough to be statistically significant at the one per cent level. The average for 1950 was intermediate and not statistically different from either 1949 or 1951. West Lost lake was the only lake which gained a significant amount of available phosphorus between 1949 and 1950, others, except Section Pour and Lost, had increases, All but these increases were not demonstrated to be statistically significant at the one per cent level by the methods of analysis employed in this study* Lost lake remained essentially the same over the three year period, and Section Four decreased by a significant amount between 1949 and 1950, due undoubtedly, to the alteration in the physical nature of the bottom which occurred in the spring of 1950. Hemlock increased significantly from 1949 to 1951. There is little evidence that the addition of fertilizer to the lakes resulted in an immediate increase of available phosphorus to the soils of the profundal regions. Of fifteen samples collected following fertiliz­ ation only eight showed an increase over levels present just before fertilization. The other seven were either lower or did not change. Factors responsible for the variation in the data of Table X may be, (1) differences in available phosphorus from place to place in the same lake, (2) inadequacy of the Ekman dredge for the collecting of upper layers of profundal soil samples, the chemical contents of which are to be quantitatively analyzed, (3) varying degrees of decomposition of organic matter of soils between time of collection and time of preparation for analysis, (4) 56 biological activities of soils in situ which resulted in release or fixation of available phosphorus, (5) influx of phosphorus, or phosphate exchangeable Ions with subterran­ ean water flow, and (6) deposition of phosphate-containing organic and inorganic matter from hypolimnial waters. The concentration of available phosphorus expressed in units of micrograms per kilogram of air-dry soil represents a condition which exists after the soil has been removed from the lake and subjected to dessication in the laboratory. To gain a more representative concept of the quantity of phosphorus actually available in any given unit of surface area of the profundal soil, the yearly averages of Table X have been converted to units expressed in terms of milligrams of available phosphorus per square meter of wet soil one centimeter deep. This calculation was made by multiplying the number of micrograms of phosphorus per kilogram of air-dry soil by the number of grams of air-dry soil per 10,000 cubic centimeters of wet soil. One square meter, one centimeter deep was consid­ ered equivalent to 10,000 cubic centimeters of wet soil. The results are listed in Table XI. Converted to units of available phosphorus per unit volume of profundal soil, the lakes remained essentially in the same relationship to each other as when compared on the basis of available phosphorus per unit weight of air- dry soil. West Lost contained more available phosphorus per unit area than others. North Twin, although not fertilized, ranked second, and Hemlock gained enough by 1951 to be probably statistically equal to North Twin. Lost and Section Pour lakes reversed positions. Section Pour had more available phosphorus per unit area than Lost due to greater compactness of clay and sand per unit volume than the organic matter per unit volume of Lost Lake. The differences, however, between the two lakes should not be statistically different if the variation exhibited in Fig. 2 is a criterion. TABLE XI AVERAGE CONCENTRATION OF AVAILABLE PHOSPHORUS PER UNIT AREA IN PROFUNDAL SOILS Year 1949 Units in milligrams per square meter of wet soil one centimeter deep West North South Lost Twin Lost Twin Hemlock 12.5 54.7 17.6 6.8 30.0 Sect ion Four ••• 1950 50.7 19.1 117.0 28.8 5.3 14.7 1951 59 .5 22.4 153.0 55.9 4.5 7.0 For purposes of comparison, the quantity of available phosphorus in the profundal soils may be converted to concentrations of total phosphorus per liter of lake water which would occur if the entire amount available in the soil diffused into the lake water. ten milligrams of phosphorus were to diffuse into one If 58 cubic meter of overlying water, the concentrat ion in the water would be 10 micrograms per liter. West Lost lake in the summer of 1951 theoretically could have released 153 micrograms of phosphorus from one square meter of p r o ­ fundal soil to the stratum of water immediately overlying the profundal soil. The lake is approximately 15 meters deep, and a dilution throughout the lake would reduce the phosphorus concentration to approximately 10 micrograms per liter, a figure on the same order of magnitude as the value which actually existed in the epilimnial waters of the lake in 1951. Hypolimnial waters, however, had a maximum concentration of about 65 micrograms per liter, a value considerably higher than that of the epilimnion. Lost lake contained only about five milligrams of available phosphorus per square meter of profundal soil. This would have produced a concentration of five micro­ grams per liter in the lower one meter stratum if all diffused into the water. Compared to the actual hypolim­ nial values, which exceeded 1,000 micrograms per liter in 1951, five is an exceedingly small amount. Therefore, it would seem inadvisable, under certain conditions, to base estimates of potential lake-water-nutrient concentrations upon measurements of the "available" phosphates in lake soils • 59 Total phosphorus in profundal soils. A number of soil samples collected from profundal zones of each lake were analyzed for total phosphorus content. of these analyses are presented in Table XII. The results In general the data indicate that total phosphorus in soils high in organic matter is higher than in soils low in organic matter. North Twin, the lake which did not receive phosphorus applications, had an average of 1.76 milligrams per gram compared with 0.65 milligrams and 0.58 milligrams per gram of soil for Lost and Section Pour lakes, respectively. Lost and Section Pour, the two lakes which received the greatest amount of fertilizer were the lowest in total phosphorus content. Section Pour had lower concentrations In the summer of 1950 than were present in 1949 and in the early spring of.1950* This decrease is correlated with the change in bottom type which occurred as a result of the subterranean collapse mentioned previously. A slight increase, although perhaps not statistically significant, occurred in North Twin lake between 1949 and 1950. The other lakes remained essentially the same during the two years for which data are available. On a wet-soil basis there was a higher percentage of total phosphorus in the available form in West Lost lake than in the other lakes. In 1950, 21 per cent of the 1.00 Section Four rH E> • O O os CO ao . 0 0 Oi 0 0 0 . CD . . CD O• O O * O * O . • O 00 to . rH CD to * O O O CD CD CD O O O • CO 00 . • 00 to . • • O 0 0 »— 1 CM 00 00 to m 0 0 . * cd O tO as G oTii o O O (D 0 o as ... O rH O co CX5 • 00 0 0 • CM • rH CD rH . rH CD os . O CM CM CM 1 —l m as as • • . . 1 —1 •— 1 0 0 0 0 . rH O O • rH TO © TJ G rH O Mg CM l7l6 lest Lost Hemlock Lost G © CO as « 0 CM • rH CO as « O CD rH • 1 —l • CO • O • . rH • . • OS O • r-H tD 00 00 1—1 CM CM CM CM . • . • • rH 1 —1 rH rH rH CM in . rH 00 CM . rH CM to . rH OS CM • rH OS rH . rH •P X © ■P © © O o TV. South Twin CO o o 00 to CM O O • • f-1 rH to Os « O • CM O • rH to . . • >—1 rH rH . rH Os O © bo . rH cd G © > aS i;~90 ID O E- t- tD tD tO rl O) C^ O rH tO LO r~t 1— I O '—* O O* O « * CM CM 00 O CD 00 rH * . »— 1 rH . ^ 1— 1 to • CM O . . • . CM CM 1— 1 *0 © -P ■P •H O © © rH CU S 11-17 bO CM CM I 1949 June of analyses North Twin S o Period of air-dry per gram of phosphorus in milligrams Units IN PROFUNDAL PHOSPHORUS TOTAL TABLE XII SOILS soil______________ 60 © G G O to to 1 CM 1 CO t> 1— t CM as CM CM CM CO CM rH G *“3 m CM 1 t>» * rH bO G G 1 CM as CM CM . • • LQ hO G < to a © G *< CO rH 1 bO CO CM rH SU © > . G © © G a *~3 © as o to OS 3 rH CO aJ OS to CM to CM I I CO I I O o I 00 © 1—I CD > • rH rH bo G G G 3 4 g bO aS • G to ? © > h tj c as G Cj ai tn X a G aS « H • -p X © -P © © CO 61 total phosphorus was available, while only 0.2 per cent of the total phosphorus in Section Four was available. Lost lake averaged about one per cent and the remaining lakes had between six and nine per cent of the total phosphorus in the available form. The average concentration of total phosphorus per unit area for the six lakes are given in Table XIII. TABLE XIII AVERAGE CONCENTRATION OF TOTAL PHOSPHORUS PER UNIT AREA IN PROFUNDAL SOILS Units in milligrams per square meter ______________ of wet soil one centimeter deep_______________ West North South Section Lost Year Lost Four Twin Twin Hemlock 470 324 494 410 280 1949 •*• 550 393 485 545 324 4230 1950 The number of pounds of phosphorus applied to the lakes has been converted to units expressed in terms of weight per unit area. The results are listed in Table XIV. These data are useful for comparing the quantities of phosphorus which were added by fertilization to the quantities which were present as determined by analysis. Over a period of two years West Lost received a total of 824 milligrams of phosphorus per square meter by fertilization. In 1950, 550 milligrams were present in one square meter one centimeter deep in the profundal zone# Lost lake received 3,711 milligrams of phosphorus 62 TABLE XIV RATE OF PHOSPHORUS APPLICATION PER UNIT AREA Units in milligrams per square meter Period of application West Lost Hemlock Lost Section Four 1949 June 16-13 159 364 736 July 7-8 July 27-23 133 133 425 339 339 1042 595 595 1926 133 595 595 595 797 797 797 Total 956 797 797 ' 2550 1950 June 14-16 July 7-3 July 27-23 133 133 339 339 339 Total 399 1017 1735 2391 Grand Total 824 2059 3711 4941 63 per square meter in the two years and had an average of 435 milligrams per square meter one centimeter deep in the summer of 1950. Apparently much of the phosphorus added to the lakes was not present in the top centimeter of the profundal soils. On the basis of these data, the fact that other lakes did not show increases in total phosphorus of the profundal soils proportional to the amounts of phosphorus added, and because the two unfertilized lakes contained as much total phosphorus as the fertilized lakes, the inference may be drawn that fertilization did not materially add to the quantity of total phosphorus normally present in the top centimeter of the profundal soils. Perhaps the quantities of added phosphorus, which have not been accounted for, remained in the littoral zone at the site of application, or perhaps they actually were present at the mud-water interface of the profundal soils, but were not detected in the analysis of these soils because the methods used to collect the samples were not quantitatively precise enough to prevent a mixing of the phosphorus at the mud-water interface with the remainder of the mud sample. Such a mixture would cause a dilution of the high phosphate concentrations at the interface. Also, there is evidence to indicate that those quantities of added phosphorus, which are not accounted for in the 64 lake sediments, accumulated in the lower hypolimnial waters. This possibility is given consideratIon in a following section entitled DISCUSSION. Available phosphorus in littoral and sublittoral z on e s . Soil samples were also collected from the littoral and sublittoral zones. The terms littoral and sublittoral as used in this investigation refer to (l) the bottom included In the area delimited by the w a t e r 1s surface and a depth of 12 feet (littoral), and (2) the area of the bottom extending from a depth of 12 feet to a depth of 30 feet (sublittoral). This classification corresponds closely to that of Eggleton (1931) for Third Sister Lake, Michigan, a lake which is morphometrically lakes of this study. similar to the The data concerning available phosphorus in a series of samples collected in August 1951 from the Pigeon River experimental lakes are listed in Table XV. Included in this table are also data of available phosphorus of profundal soils for the same period. The greatest phosphorus concentrations In littoral zones were in the two unfertilized lakes, North and South Twin. South Twin had 75.0 milligrams of phosphorus per kilogram of air-dry soil and North Twin had 46.4 milli­ grams of phosphorus per kilogram of air-dry soil. Lost was only slightly lower than North Twin. 7/est These three 65 lakes are characterized by having high percentages of organic matter in the soils of their littoral and sub­ littoral zones, as well as their profundal zones* The three marl-bottomed lakes, Hemlock, Lost, and Section Pour, contained much less available phosphorus on a dryweight basis than the first three mentioned, none of them having more than 7 milligrams per kilogram of soil in the littoral zone* TABLE XV CONCENTRATION OP AVAILABLE PHOSPHORUS PER UNIT WEIGHT OP' SOIL IN THE LITTO RA L ,.SUBLITTORAL, AND PROFUNDAL SOILS DURING AUGUST, 1951 Units in milligrams per kilogram of dry soil West North South Sect ion Zone Twin Lost Twin Pour Hemlock Lost 46.4 44 *3 Sublittoral 167.0 Profundal Littoral 75.9 59.8 72.0 40.7 276.0 6.7 103.0 336.0 89.6 1.0 11.5 3.7 0.7 2.1 1.3 Hemlock, Lost, and Section Pour lakes contained more available phosphorus in their sublittoral zones than in either the littoral or profundal zone. This was not true in the three soft-water lakes, North and South Twin and l e s t Lost* Roelofs (1944) also recorded differences ‘in quantities of available phosphorus between the littoral and sublittoral zones in certain lakes* On the basis of available phosphorus per unit volume of wet soil a somewhat different aspect appears* 66 The data of Table XVI have been computed to indicate the milligrams of available phosphorus per 10,000 cubic centimeters of wet mud. This is the volume of soil equiv­ alent to an area of one square meter by one centimeter deep. On this basis, the three soft-water lakes contained higher concentrations in the littoral zone than in the profundal z o n e , whereas the three hard-water lakes remained in essentialy the same relationship as when available phosphorus was measured on a dry-weight basis. TABLE XVI CONCENTRATION OF AVAILABLE PHOSPHORUS PER UNIT AREA IN L I T T O R A L , ,SUBLITTORAL, AND PROFUNDAL SOILS DURING- AUGUST, 1951 Units in milligrams per square meter of wet soil one centimeter deep North Twin Zone South Twin We st Lost Hemlock Lost Section Four Littoral Sublittoral 63 15 239 25 217 213 6 34 2 12 2 5 Profundal 45 22 145 32 3 9 These data suggest that when analytical studies of available nutrients in lakes are performed, there be Included the determination of the nutrients on the basis of unit area as well as units of dry-weight material, because nutrient contents of soils which are high in organic matter may be more favorably compared with those of mineral soils high In inorganic constituents. 67 Pep cent organic matter in profundal soils. As a part of the chemical analyses of this study, a number of organic-matter determinations were made on certain profundal soil samples. XVII. The results are listed in Table These data show a gradual decrease in per cent organic matter from the soft-water lakes to the hard-water lakes. North and South Twin lakes averaged highest, having a 57 to 60 per cent organic-matter content. Lost and Section Four lakes, the two highest in marl deposits, were lowest in per cent organic matter, having values which ranged from 46 to 6 per cent respectively during the summers of 1949 and 1950. Section Four changed from 59 to 2 per cent between March and June of 1950. This abrupt decrease was due to the raarl-clay-sand layer deposited on the bottom as a result of the subterranean disturbances already discussed. The March sample was not included in the determinations of the 1950 average for this reason. Chemical characteristics of drainage-area soils. An investigation of the inorganic chemical nutrients in lake waters and lake-bottom soils would not be complete without a study of the chemical nature of the soils surrounding the lakes. It is through these soils that much of the water which is destined to enter the lakes must flow. Rawson (1939) endeavored to explain lake productivity on the basis of the hypothesis that Ttwhile 68 TABLE XVII PER CENT ORGANIC NATTER IN PROFUNDAL SOILS Period of analyses 1949 June 15-23 North Twin 60 July 27-29 56 Aug. 22-29 56 Average South Twin • • West Lost Hemlock 46 43 Lost 30 Section Four 40 44 57 60 53 54 54 50 44 42 57 60 51 47 46 42 62 60 62 58 49 44 58 54 44 51 39 2 6-10 1-7 55 59 47 49 53 58 50 48 61 55 61 58 32 26 Aug. 28-29 Average 57 56 57 58 13 3 7 Grand Avera ge 58 59 50 53 43 1950 Mar. 23-24 June 13-19 July Aug. * • ^■March sample omitted from average. 42 39 See text. 6^ 24 69 the edaphic factors determine the kinds and amounts of primary nutritive materials, the morphology of the basin and the climate may to a -large extent determine the utilization of these materials.11 To this proposal, Deevey (1940) added the comment, ’'Morphometric features are irrelevant in regional comparisons and represent a compli­ cating agency to be eliminated . . The shed soils results of the chemical analyses of the water­ are listed in Table XVIII. table are presented in two parts. The data of this The first part lists relative quantities of the easily available or "active" nutrients; the second part indicates the relative quantities of "reserve1* nutrients, those made available only by intensive leaching. Of particular interest to the present study are the results of the calcium, magnesium, and phpsphorus analyses. North Twin, which had 24 p.p.m. methyl-orange alkalinity in the lake waters, contained an average of 48 p.p.m. of active calcium in the surrounding soils. When this calcium content is compared to the 10 p.p.m. for South Twin, a lake which had an alkalinity of 74 p.p.m., it seems relatively high for a lake which contains relatively soft water. Furthermore, Lost lake, with a methyl-orange alkalinity of well in excess of 100 p.p.m., contained approximately the same active calcium content in the terrestrial soils as did North Twin. In this regard 70 d P O d o rH * O O « 03 to o ♦. • CD O o to • • o CO to LO rH +3 O © CO B CO «=C t-i O p 03 O Pi Ed o O • • d o o p O rH CO o o * LQ "TO rH t O o • 083 • rH 118 o CO to • rH • O to • rH o • to • "TO CO • to • "TO 03 • o o w d 5d M Q P—> d to CO M M M > X M CO vP M O CO Q pq Ed 00 *4 cd PQ •TO Eh XI Eh o o rH -p o cd d p H pL, o 1— 1 Eh CO M Dd W Eh o •TO cd < td o id o o • • • GO "T O o o CO to o • LO LO • CO CD LO • o £ •TO CD LO CO 00 • o • p. • o O P o o • • o o (d d rH • O • d P LO « O GO P d O LO CO CO • • d rH -P O * (X LO cd to to • • o CO LO o • O LO • CD LO • o d © •d to to o d LO 'O t>> Ed o (X d TO P o co d •H £ Eh to to o d O o d o o d > «H e d 00 o © • * o o Eh rH CD CD xi -P TO p TO P d P 03 -P *H d E=> © Eh rH •H Eh d *tH • to •rH TO *h CD d CO X CD O CO • o © -TO rS LO 268 o cd LO CO CO • • d O P "TO LO CO LO o o LO -p GO 03 a CO cu Oh to CO 52? M 23 d M Ed o .d TO -P o ■H p 03 TO TO •rH d © P O TO d TO & O TO © P d O TO rg d< d © P p !s o jd Ph P P ♦H TO TO TO e n p •rH 03 a d •rH P P CLh O C p TO © to TO *■=3 d o d M © TO © d TO to d TO *2=1 TO TO P 6 o •H d ,d to (X 03 03 OS e *pH 03 0) o d d -p I—I to o o o TO 0} £ Pu pH o S3 d a CD 03 CD d aJ to d TO w CL. 71 there seems to be, little correlation between edaphic features and the chemical nature of the lake waters. The relative differences in magnesium contents of the soils of the different lakes are similar to the calcium data. highest active calcium test Section Four soils. The (118 p.p.m.) was from the This is correlated with high concen­ trations of marl and a high alkalinity of the lake. The reserve calcium and magnesium tests were essentially in agreement with relative differences in active tests between lakes and do not contribute toward a change in conclusions drawn on the basis of the active-test data. Quantities of easily available phosphorus in the drainage-area soils were so low for each lake, that differences between lakes are probably of slight signif­ icance. samples. Positive tests were obtained on only a few Hemlock and Section Four lakes had averages of 0.1 p.p.m., the highest values for any of the lakes. The reserve phosphorus tests, although exceedingly low, were high enough to measure with greater accuracy than the active tests. On the basis of reserve phosphorus, North Twin ranked lowest of all lakes; it averaged 0.25 p.p.m., while the other lakes all averaged 0.5 p.p.m. or slightly higher. The agronomist would consider these levels far below the minimum necessary for good crop growth. Adequate supplies for ordinary agricultural purposes would be 72 10 p.p.m. according to Spurway and Lawton (1949). A direct correlation existed between the calcium content of the soils and the pH. Soils with greatest acidity were those around North Twin and Lost lakes, the lakes which were lowest in calcium content. Hemlock lake is located in a region of sandy to fine sandy loam soil, in contrast to the almost pure sand around each of the other lakes. Higher nutrient co n te n ts ,would be expected in the loam, but the differences which were detected between the soils of Hemlock lake and those of the lakes in the pure sand regions were slight. One fundamental cause of the minor variations which did exist from lake to lake may be that the lakes are of different ages. This factor would account for greater leaching of the surface soils, with a consequent increase in their acidity and a decrease in available nutrients. If the two lakes which represent the extremes in age, as determined by the alkalinity of the water, are selected for comparison on the basis of the alkalinity of their drainage-area soils, a correlation is seen to exist. North Tv/in was lowest of the six, not only in the alkalinity of its water, but also in the reserve calcium content of its terrestrial soils. Section Four, a lake which probably is still in the process of origination, was highest in the alkalinity of both its water and surrounding terrestrial soils. 73 Durat ion of phosphorus retention in lake w a t e r s . The fisheries biologist planning fertilization experiments is interested in knowing how much of the phosphorus which he puts into the lakes can be expected to be utilized by organisms, and how long it can be expected to remain avail­ able to organisms. The present study reveals some useful information pertaining to this problem. The per cent of the applied phosphorus which was actually present in the epilimnial and hypolimnial waters three to four days following fertilization has been determined and is presented in Table XIX. These data were derived by dividing the theoretical phosphorus increase due to fertilization into the actual increase which occurred, and multiplying the quotient by 100. The percentages of phosphorus remaining in the epilimnial waters several weeks or months after applica­ tion have also been determined. These data were derived by dividing the units which express the quantity of total phosphorus present at the time of analysis by the units representing the entire amount of phosphorus applied either during the concurrent season or during the preceeding season or seasons, and multiplying the quotient by 100. results are listed in Table XX. The data of Table XIX indicate that in 1949, between 60.4 and 103.0 per cent of the quantity of p hos­ phorus put into West Lost lake was present in epilimnial The 74 8 P«H a o P -P O © CO ss; +3 o M in O £5 S P »-3 O P P to to • • CD CD to 03 CO • m 00 CO LO 02 CO o • • CO o 03 03 LO o • • o o CO O o • • o o o * CO o 03 o • CD 03 03 o o • • o o o O * • 05 CO o p o o • • o CD CD o • CO rH i —1 o o • • 02 p 03 CD LO CO o o • • o o w £3 cd O o •rH P £3 £ M •rH P rH rH O P P P t>> w w £3 £3 O o •rH £ 0 g P P rH o P p P P M C» 05 CD O • • CO co CD rH . >H CO dd fz; o P M {2d Eh h o » P P S h 0 © cd P 0 a < m O O P -p •0 0 O N P P £• P © *H Ph -P £, © cp O •H £3 •rl rH O P P P P Cu fx | • CO • • * o CO 85 >* o * CO o 1 —1 3 0 £3 o •rH *s £ p 1—t o *o p d E p p 0 W p p >? w w CD 05 t—I r —1 05 05 rH P O LO CD P o to CD P o to CD p o to 05 p •\ 03 03 1 CO rH 03 03 1 GO rH O o CO to l 05 02 03 05 rl 1 CO i —1 CD P 1 CO p f —1 1 —t 1 o p p p t o 1 —1 0 £3 0 £3 rl P 0 »—1 P 1 1 rH rl rH P •t 0 0 £3 £3 2 p P 3 a CO 03 I o 03 *• CO 03 1 £> 03 t>> rH i —1 0 P 2 P P P 0 p •s to P 1 0 0 P t>> 0 p 0 A D~I 1 CO to c- p 0 p t>» p £J p 75 waters three to four days following fertilization. In the two hard-water lakes, Lost and Section Pour, the percentages of phosphorus present in the surface waters were between 26,8 and 6 8 ,2 , During 1950, following the fertilizer applications in mi d- J un e , phosphorus increases ranged from 67,5 to 527,0 per cent of the amount applied to the various lakes. The soft-water lakes, West Lost and Hemlock, again exhibited greatest increases, while Lost and Section Pour, the predominately marl lakes, exhibited lower increases. Ho increases were recorded in the surface waters for the period following the early ^Tuly phosphorus applications. TABLE XX PER CENT OP ADDED PHOSPHORUS PRESENT IN EPILIMNION SEVERAL WEEKS OR MONTHS FOLLOWING- FERTILIZATION Period of analyses West Lost Hemlock Lost Section Four Aug. 22-25, 1949 June 14-15, 1950 23 30 56 18 20 11 18 7 Aug. June Max. Aug. 24 7 7 5 32 15 8 2 8 2 5 1 8 1 4 0 28-29, 12-14, Value, 20-21, 1950 1951 19511 19511 ISee t e x t , The data of Table XX demonstrate the theoretical degree of phosphorus retention in the lakes several weeks or months following application. These data represent a calculated approximation of the per cent residual 76 phosphorus in the epilimnial waters. possible If it had been to determine by empirical methods the quantity of phosphorus that would have been present on a particular day of analysis had there been no artificial phosphorus additions, it would have been possible to calculate the absolute percentages present, but not having data of conditions unaffected by fertilization for any one day of analysis, it was impossible to obtain anything but a calcu­ lated approximation. The data of the control lakes for 1949 and 1950 indicate that, as time went on, the calcu­ lated approximation probably became closer and closer to the absolute levels which would have existed had there been no fertilization. Moreover, the calculated percentages of residual phosphorus represent maximum values and it can be assumed that at least no more than this percentage of the amount added artificially was present. Thus the percentage retention values would have been less, if the quantities representing normal phosphorus concentrations had been removed from the calculation. By the end of the first year of fertilization, no more than 56 per cent of the total phosphorus applied remained in any of the lakes. Hemlock was highest, with 56 per cent, and the other lakes ranged down to Section Four, the lowest, with no more than 18 per cent still present. At the beginning of the next season, In June 1950, there 77 was about 20 micrograms of total phosphorus per liter in Section Pour lake* This represented seven per cent of the total phosphorus applied in 1949. If one were to consider the pre-fertilization levels of May and June 1949 as being normal phosphorus concentrations, only 12 micrograms per liter of the 20 present in June 1950 would be due to fertilizer added. This would only be a four per cent retention. By the end of the second year of fertilization the greatest amount remaining in any of the lakes was 32 per cent. This was in Hemlock, the same lake which retained the greatest amount the preceeding year. At the beginning of the next summer, 1951, Hemlock continued to have greater quantities of total epilimnial phosphorus than the other lakes, although the concentrations continued to decrease, and there was present only 17.6 micrograms of phosphorus per liter, or 15 per cent of the quantity that had been added the previous year. Only 1.6 micrograms per liter was detected in June 1951 in Section Pour. This represents less than one per cent of the quantity put into the lake the previous year. Actually there is no reason to believe that an appreciable amount of the fertilized phosphorus was included in the quantity present at that, time, for 1.6 micrograms per liter is such an extremely low concentration for a total phosphor- 7S us content, that any contribution of the fertilizer must have been inconsequential. Table XX also includes determinations of the per cent residual phosphorus remaining from the entire fertil­ ization program over the two year period from 1949 to 1950. The next-to-last horizontal row of figures indicates the maximum per cent of the total phosphorus added In 1949 and 1950 which appeared in the lakes during the summer of 1951. The last horizontal row indicates the per cent of the total phosphorus added in 1949 and 1950 which was present at the end of the 1951 summer* At most only eight per cent of all that had been added during the previous two years appeared in the lakes the year following cessation of phosphorus applications. Hemlock lake had 18.6 micrograms per liter In July 1951, a value equivalent to approximately eight per cent of the total 246.5 micrograms per liter that had been put into the lake In 1949 and 1950. If the quantities present in Hemlock lake In early 1949, before fertilization was started, are taken as representing normal phosphorus levels, the theoretical eight per cent residual phosphorus is reduced to zero because there were already nearly 30 micrograms of total phosphorus per liter in the epilimnion before the lake was fertilized. A statistical analysis of the significance of the degree of variation between the per cent residual pho s­ phorus present several weeks or months following fertili­ zation (the data of Table XX) has been computed and results of this analysis are recorded in Table XXI, The analysis of variance method described by Snedecor (1946) was employed in these calculations. An F value of 6.8 was obtained for the variation existing between periods. This value indicates that the variation between periods was highly significant and therefore this degree of variation could be expected to happen by chance only one per cent of the time. The assumption that time is an important factor responsible for losses of phosphorus added through fertil­ ization is thus given statistical support. TABLE XXI ANALYSIS OP VARIANCE OP PER CENT RESIDUAL PHOSPHORUS PRESENT SEVERAL WEEKS OR MONTHS FOLLOWING FERTILIZATION Source Degrees of freedom Sum of squares Mean square P 23 3913 Lakes 3 936 312 5.1* Periods 5 2064 413 6 .8 ** 15 913 61 Totals Error -"-Significant at the 5 per cent level. **Very significant at the 1 per cent level. Alkalinlty of l a k e s . Methyl-orange alkalinity deter­ minations were made on samples of water collected during 80 periods when samples were collected for phosphorus analysis. The results of the alkalinity determinations were previously listed in Table VI (page 36). Samples collected during 1948 (Tanner ms.), the year preceeding fertilization, had average alkalinities of 34, 74, 138, 163, 177, and 192 for North Twin, South Twin, West Lost, Hemlock, Lost, and Section Pour, respectively. Marked changes in alkalinities of each lake occurred over the four-year period under investigation. With one exception gradual decreases in both fertilized and unfertilized lakes o c c u r r e d .from 1948 through 1951. Section Pour Increased between 1950 and 1951, while all other lakes continued to decrease. A discussion of the biological factors related to the changes in alkalinities during the years 1948, 1949, and 1950 Is included In a report by Tanner (ms.). It may be noticed from the data of Table VI that variations in the methyl-orange alkalinity were consider.able not only within one lake over the four-year period, but also In the relative differences from lake to lake during the four years. Thus Hemlock lake in 1948 and 1949 had higher alkalinities than West Lost, but by 1950, and in 1951 it averaged lower than West Lost. The changes that occurred, especially in Hemlock and Lost lakes, were of such magnitude as to render difficult the categorical classification of these lakes as hard-water or soft-water 81 lakes, as is often done in routine limnological surveys. The relationship of alkalinity to the retention of added p ho s ph o r u s . The data of Table XXI (page 79) includes not only a statistical analysis of the variation of the per cent of total phosphorus remaining from time to time, but also a statistical analysis of the differences existing between lakes. An examination of the data presented previously In Table VI (page 36) and XIX (page 74) seems to indicate that with increases in alkalinity in the lakes general decreases occurred in amount of applied phosphorus remaining. Thus the average (Table XIX) for all the surface samples collected from West Lost and Hemlock lakes was almost two and one-half times as great as the average for all the surface samples collected from the two hardwater lakes (Lost and Section Four). The same trend appears to exist in the data of Table XX (page 75). The average per cent remaining in West Lost and Hemlock lakes was also approximately two and one-half times as great as the average per cent remaining in Lost and Section Four lakes. The analysis of variance (Table XXI, page 79) of the per cent of phosphorus remaining several weeks or months after fertilization indicated that the degree of variation between the lakes was significant at the five per cent level. In other words the chance occurrance of this variability was five per cent. 82 To further test the probability of alkalinity as a factor affecting difference In percentage of added phos ­ phorus remaining in different lakes, a statistical regres­ sion analysis (Snedecor 1946) was computed. This analysis was a determination of the correlation between average alkalinities and the percentage of residual phosphorus (Table XX, page 75). The alkalinity values selected for the regression analysis were those existing during periods of phosphorus applications and during periods of analysis of per cent residual phosphorus remaining. The regression analysis was computed by relating the average of these alkalinities values to the phosphorus values. The results of this analysis are presented in Table XXII. A regression analysis h^d also been previously computed by comparing, not the average of the alkalinity values, but the actual, alkalinity values which existed during each period. No statistical significance was found to exist when the per cent of phosphorus remaining was related to actua1 alkalinity values. However there was significance at the five per cent level when average alkalinity values were related to residual phosphorus. ant for two reasons. These findings are import­ First, the probability is high that certain concentrations of alkalinity are responsible for the loss of added phosphorus from epilimnia of lakes. Second, when the differences In alkalinity values which 83 occur during a particular season in a lake are quite v a r i a b l e , the average of these values may be substituted for individual alkalinity determinations in certain statis­ tical analyses. TABLE XXII REGRESSION ANALYSIS OF PER CENT RESIDUAL PHOSPHORUS AND AVERAGE ALKALINITY VALUES Degrees of Sum of freedom squares Source Total Between lakes 2064 3 936 312. 1 476 476 2 460 230 5 1215 15 913 243 61 • Between periods Error 23 F CD Regression on alkalinity Deviation from regress ion Mean square -^Significant at the 5 per cent level. Alkalinity is only one among several factors that may have been responsible for the relative differences in amount of phosphorus remaining in the lakes. Other causal factors could have been variations between lakes In (1) the ratio of the epilimnial volume to the hypolimnial volume, cent organic matter in lake-soils, {3} (2 ) per degree of phosphorus absorption by organisms and rate of deposition, (4) rate of regeneration through decomposition, and (5) physical turbulence of epilimnion as affected by wind and thermal conditions. Each of these items certainly enters into the 84 phosphorus cycle of lakes. It has not been the purpose of the present study to thoroughly investigate each item, however, included in the following section (DISCUSSION) is a general resume of the relationship of some of them to the phosphorus cycle. The work of Deevey (1940) supported the concept of Naumann (1932) that phosphorus contents would be low in highly calcareous lakes because of the formation of insol­ uble calcium phosphate compounds. Gessner (1939) believed tri-calcium phosphate to be the compound. To obtain further information on relationships between natural phosphorus levels and alkalinity concentrations in highly calcareous lake, a number of the data presented in Table V (page 34) have been selected and are listed in Table XXIII. These data were selected because they represented the most alkaline lakes and because they represented periods least likely to be Influenced by artificially added phosphorus. Thus the May and early June 1949 samples were collected before fertilization began, and the June 1950 and all of the 1951 samples were collected ten months or more follow­ ing a previous fertilization period. An inspection of Table XXIII indicates that in general the variability in the data within any one lake is approximately as great as that between lakes. Differences of from 20 to 25 micrograms of total phosphorus per liter 85 TABLE XXIII EPILIMNIAL TOTAL PHOSPHORUS CONCENTRATIONS DURING SUMMER PERIODS LEAST AFFECTED BY FERTILIZER ADDITIONS ______________ U n i t3 in micrograma per liter Period of analysis 1949 May 1 West Lost Hemlock Lost Section Four 10 28 10 10 June 11-17 10 30 14 6 1950 June 14-15 23 22 28 20 1951 June 12-14 5 10 18 12 4 24 2 3 8 19 5 4 3 23 3 10 19 12 10 July 6-8 July 30-31 Aug. 20-21 Average 6 86 were present within all lakes, and the greatest difference between any two lakes was 28 micrograms per liter* All lakes ranged from less than six micrograms per liter to over 20, but none exceeded 30 micrograms per liter. Three of the four lakes averaged between 10 and 12 micrograms per liter, while Hemlock averaged slightly over 19 micrograms per liter. The analysis of variance (Snedecor 1946) of these data has been computed and Is presented in Table XXIV. No statistically significant difference was found to exist betweer; the data of the different lakes, although the variability within different periods was significant at the five per cent level. Deevey (1940) considered an average total phosphorus content of approximately 15 micrograms per liter, as found in lakes in Connecticut (and Wisconsin, Juday and B i r g e , 1931), to be minimal In concentration. He used the term "oligotyplc” to express conditions typified by such low total phosphorus concentrations. He did not indicate total phosphorus ranges for his next two terms, mesotypic and polytypic, which referred to proportionally greater concentrations of phosphorus. It may be concluded, on the basis of D e e v e y 1s classification that in general, with the exception of Hemlock lake, the Pigeon River area lakes which were included in this investigation are oligotypic for phosphorus. There is no direct evidence in the present report 37 that indicates a precise relationship between alkalinity and normal concentrations of total phosphorus in lakes which range from about 24 p.p.m. to 170 p.p.m. methylorange alkalinity. however. Such a relationship may actually exist, A series of samples collected at more frequent intervals should provide evidence upon which more valid conclusions could, be made. TABLE XXIV ANALYSIS OP VARIANCE OP EPILIMNIAL TOTAL PHOSPHORUS CONCENTRATIONS DURING- SUMMER PERIODS LEAST AFFECTED BY FERTILIZER ADDITIONS Mean square Degrees of freedom Sum of squares 27 2130 Lakes 3 398 Periods 6 850 133 142 18 882 49 Source Totals Error -"-Significant at the 5 per cent level• F 2.72 2.9-::- RESULTS OP LABORATORY EXPERIMENTS A series of laboratory experiments were conducted to obtain further information concerning relationships between alkalinity and the precipitation_of phosphorus. In the first group of experiments, five Fernbach culture flasks containing two liters each of a 0a(HC03 )2 and KH2PO4 solu­ tion were placed in the dark at room temperature for a period of 95 days. The initial phosphorus concentration in all flasks was. approximately 100 micrograms per liter. The initial methyl-orange alkalinities in.the five flasks were 28, 72, 116, 168, and 216 p.p.m. respectively. The methyl- orange alkalinity, phenolphthalein alkalinity, pH and soluble phosphorus were determined at certain intervals for all solutions. The results are given In Table XXV. A marked precipitation of phosphorus occurred only in the flask which initially contained 216 p.p.m. methylorange alkalinity. The phosphorus concentration decreased from 103 to 2 micrograms per liter. There was also a decided decrease in alkalinity to a final concentration of 60 p.p.m. In none of the other solutions was there much change in either alkalinity or phosphorus. Bound carbonates (as shown by phenolphthalein tests) appeared in flasks of 72 to 216 p.p.m. alkalinity, but did not appear in the one containing the lowest initial1 alkalinity (28 p.p.m.). The highest phenolphthalein alkalinity which occurred was 20 G9 to • CO rH 0 lO OS t> rp O C O •n 4©3 C O 43 •rH to OS 43 o 03 •rl C O I>> pH C O C 3 rH as 02 • ^ O COC O 02 C O OS to • O rH rH rH CO 02 03 02 t£» 0 0 0 03 E> rH rH ^ O 02 m • t 'H 03 02 • rH O C Oto 03 o- ^ 02 o• 0- 03 C O to • 02 02 CO lQ 03 t> C O • ^ O t> 02 02 03 to • C O tO C O 03 to to • 02 C OCO 03 03 rH rH C O0 to rH 1— 1 C O • ^ O O 03 02 C O to • C O02 C O1—1 03 to to • Oto O C OC 03 rH f— 1 • to ^ C OH* 03 to rH rH 0 C m O 02 ■ 00 02 C O • • H* C OC O03 03 rH rH 10 • to C OC O1 0 03 1 — 1 1 — 1 to • 00 02 co m 03 to rH rH 0 0 <0 O • co to to 0 0 ^ • CO to ^ to • COC O 02 0 • O 02 co in 03 rH rH C O • C O0 0 C O 0 < — I 1 — 1 rH 0 k“4 as O 10 r— 1 02 • C O^ 1 — 1 02 ^ to 1— 1 d 03 to d o 43 2 I—I o •r4 C O to • ^ SF C O02 tO rH O 0 rH 1 — 1 rH to m C O O £> 02 02 03 • tO 02 CO0 to O rH * CO0 02 rH 1— 1 to • O03 02 O C 03 rH 02 02 CO 1— 1 as •rH 43 •rH LO • O- 02 O 1— 1 C O0 02 d 10 • 02 O {> 0 I> 03 C O • to 0 0 - 0 03 I— 1 rH C O • t> 02 to rH O 1— 1 02 C O0 to 1— • 03 • 0 t> to 0 rH M C O •H C O K's rH as a 43 •H >> d 4 3 t4 •H 1— 1 d as co as M rH • S d as as d as as O d 43 XI rH PH • as X O ^ PH C O >•» 43 •H t>s d 43 -H •H 1— 1 d B •H as d as as *4 1—1 1— t as to 0 i—1 d as J W H PH p «;s • jd 43 O 43 0 W -H • rH d 43 »iH •rH pH bo as ,d O 43 •rH •H X rH 1— 1 B •H ^sj 1— 11— 1 as < 1 rH \ PH O ^ • 43 *H 1 — 1 \ >> d PH 43 *H •rt 1— I G Q d as •H ^ s j B rH rH as d to 0 rH d as 0 ^ K PH as as H • x 43 • 43 1— 1 •rH PH 43 *H •H rH \ d 0 2 d d as *— < •H ^ as 1—1 f— t d as as to O 1— t • C O 1 —1 p4 M M M ► > hH M M as d to O as .d d 0 43 0 O X ! W -H m ph p « ra hH 03 B d 0 j d t d *h S PH « — M CO • rH \ PH r> 90 p.p.m. This occurred in the. flask initially containing 216 p.p.m. methyl-orange alkalinity. recorded for all solutions. Increases in pH were The lowest alkalinity solution (28 p.p.m.) never exceeded pH 8 while all the remaining solutions reached at least pH 8.3, The maximum was 8 .8 , which occurred In the 168 p.p.m. solution. Several hypotheses concerning the precipitation of phosphorus in calcium carbonate solutions are suggested by these data. Because no relatively significant changes occurred in the phosphorus concentration of the flask which contained 168 p.p.m. alkalinity and in which the pH increased to and remained above 8.5 for 89 days, it may be concluded that a pH measurement alone may not be a safe criterion on which to base a prediction of the formation of insoluble calcium phosphate compounds in lake waters. The fact that alkalinity decreased from 216 to 60 p.p.m. in the flask which contained an initial concentration of 216 p.p.m., and no decreases occurred in the flasks which contained between 60 and 168 p.p;m., indicates that a calcium phosphate compound may have developed through adsorption processes. If the laws of solubility product constants were operating, the reductions in soluble calcium carbonate (in the bicarbonate form) should, have reached an equilibrium at a concentration no less than the highest which actually remained in solution. The loss of calcium 91 carbonate from solution to a point below the equilibria established, in the 72, 116, and 168 p.p.m. solutions may have been due to the adsorption of calcium, carbonate, and phosphate ions to a micro-nucleus of aragonite or perhaps tri-calcium phosphate. This hypothesis is not in agreement with the conclusions of Schlosing (1900) and G-essner (1939) who believed that phosphorus combined with calcium to form tri-calcium phosphate and precipitated in this form. If it is true that colloidal adsorption of phosphate ions by calcium carbonate particles occurs, It follows that the laws of mass action and solubility product constants do not explain the precipitation of the phosphate from solution as Schlosdng . and Gessner believed. Further evidence supporting the possibility that a colloidal adsorption mechanism was operating was provided by the results of a second series of experiments the data of which are given In Table XXVI. These data were obtained by repeating, with two modlfications, the first group of experiments. The Fernbach flasks were placed in a lfshaker,! and agitated at the rate of approximately 100 oscillations per minute. A second difference was that d if f e r e n t 'KH 2P04 concentrations were used. Methyl-orange alkalinity, phenolphthalein alkalinity, pH and soluble phosphate phosphorus were measured each day for five days or until phosphorus In one of a pair of flasks of different 92 O rH • in • ^ 03 O co O O C O b CO O CO CO O 03 CO rH 03 • i—I m O * co .n O CO • o rH CO rH to LO 03 o in CO rH CO rH • ^ 03 CO O O r-4 • 03 rH ... O to O 03 to • m rH • • ObO to to O O 03 o • O O GO CO • • CO o n co co o o o - o O rH • 03 O CO ^ rH in rH • • 03 CO to I—I CO n • lO 03 * • • rH n rH co «n LO to O• ^« ^HTOLO 03 rH CO 03 0 0 0-0 CO to in * • n co rH 03 •H CO 03 2 O d d •H t>* rH CO O 03 d co 03 43 d o o T) to CO O O C O ^ CO O 03 00 CO n l>* CO Pi 43 aj 43 bO o• o• to o CO o in o co ^ m 03 i— I rH * 03 O CO CO CO i —I flj "OTT) O'SO • • m oi cO o•^• O 03 • O CO n O rH CO rH CO 03 CO 03 n to DQ a o •H 43 2 co r-4 o CO •r4 d O rH o m o• o• • • * • tO o co m f O O b CO to o t> CO in CO 03 I—I 1» 43 •rH • 43 ■ rH •rl tn d r— 1 PH 43 *H •rH i— 1 d •H •H rH d CO •rH & rH rH 03 CO 4= 03 •H CO *» rH CO d «J3 to co i—i m B co d bO i— 1 • O CO £1 d 43 o o m in •H PH P 4 S d co •H .id r— 1 i—1 co co rH • co ,d PH >2 43 •H CD s CO d hO X O rH aj d 43 o O ^ W H g d d rH .H i— 1 co JSJ i— 1 CO • O t> 03 a a) 43 *H •rH |— 1 d co *H X i---1 1-< CO CO X to i —t 4^ 03 S CO d hO PQ 1 < 1 CQ 1 M M M IH >> « 43 —l •H 1 !>» d PH d hfl • 1— i * o 0 d d co .d 43 43 O O O C W H 0 d tn irf H ft S PH 03 CO to O O t- )— I >* d PH O t> • 4 * • 43 •H • 1---1 >s d 43 •H d •H rH CO o o CD O J> 03 CO n 43 «H •H 1— 1 d co •rH 1— 1 1— 1 CO CO ^£j rH • CO d 43 g >> a PH 03 B CO d bO O d 0 W H d d g 1 l—1 M • rH 43 «H •H i— 1 d as •H Jsj — 1 rH 1 S 05 rH * co d 43 PH 03 a CO d to 0 d 0 S PH PQ 1 M M 93 phosphorus concentrations had decreased to zero or near zero* Methyl-orange alkalinities were adjusted to 116, 136, and 160 p.p.m. in separate flasks. A pair of flasks was prepared for each of the three alkalinity levels. Into separate *flasks of a pair were put different concentrations of K H 2PO4 . An effort was made to adjust the phosphorus concentration to values within ranges normally found in lakes. Each pair contained a low (8 to 25 micrograms per liter) and a medium (32 to 53 micrograms per liter) concentration of phosphorus. An examination of the results of these experiments (Table XXVI) reveals that phosphorus precipitated more rapidly and to a relatively greater degree in the flasks initially containing the lower phosphorus concentrations. This was true at methyl-orange alkalinities of 116, 136, and 160 p.p.m. A noticeable decrease in methyl-orange alkalinity also occurred in two of the three flasks which decreased in phosphorus concentrations. None of the solutions containing 25 or more micrograms of phosphorus per liter showed marked decreases in alkalinity. The reason for the lack of phosphorus and calcium carbonate precipitation in those flasks in.which there were medium levels of phosphate may have been that a colloidal particle did not form because of the stabilizing Influence of potassium ions. (This assumes the calcium phosphate 94 precipitation in the flasks of low phosphorus to be due to colloidal coagulation. Solutions which had been given greater quantities of phosphorus were also given greater quantities of potassium because the phosphate compound used in these experiments was KHgPO^.. Marshall (1949) has discussed the coagulating and peptizing properties of potassium and calcium ions in conjunction with colloidal properties of clays, and has demonstrated slower coagula­ tion in solutions containing potassium ions than in solutions containing calcium ions. DISCUSSION The role of phosphorus in the metabolism of a lake is a complex one. Phosphorus ions enter into many physico­ chemical and biochemical reactions. The utilization of phosphorus by living cells in lakes may be so great as to reduce the available supplies to such low concentrations that their detection by laboratory techniques is difficult. Also, the chemical properties of phosphorus make it highly sensitive to chemical and physical changes. Because of this relative instability, phosphorus reflects the dynamic ■biological and physical processes of lakes by undergoing conversion from one component of the lake to another. Ohle (1938) described several different fractions or components of lake waters in which phosphorus may be present. His classification includes organic and Inorganic fractions of both dissolved and suspended matter. The suspended fractions are particulate and may be separated from soluble forms by centrifugatIon or by filtration. The phosphate ion, according to Ohle, may exist (1) in the available soluble form, (2 ) as a part of an organic m ol e ­ cule or adsorbed to organic colloids, (3) in a suspended, particle as an iron or calcium phosphate complex, or adsorbed to these complexes, and (4) as a part of partic­ ulate organic matter either living or dead, or adsorbed 96 to this organic material* A number of investigators, among them Juday and Birge (1951), Hutchinson (1941), Einsele (1941), and Hayes and Coffin (1951), have demonstrated that vertical strati­ fication of phosphorus occurs in thermally stratified lakes. Thus phosphorus may vary in lake waters from fraction to fraction within the inorganic-organic, soluble-particulate phase, and it may also vary from one thermal layer to another thermal layer within the lake as a whole. There are similar fractions, or components, of lake sediments in which phosphorus may be present. Ohle (1937) described a phosphorus exchange system composed of differ­ ent colloidal gels. distinct fractions: This system is composed of three (1 ) an insoluble phase; interphase, and (3) a soluble phase. (2 ) an exchange On the basis of laboratory experiments which Ohle performed, he proposed that the acidity of lake sediments to a large extent would affect phosphorus adsorption in the exchange system. He further proposed that acidity would be determined by the concentrations of organic matter and marl in the sediments. Very few analytical investigations have been made on the phosphorus content of lake sediments. WoHack Hutchinson and (1940) examined a 43 foot core sample of the sedi­ ments of Linsley Pond, Connecticut and found that the total phosphorus content varied within the core from 0.055 97 milligrams to 0,354 milligrams of phosphorus per cubic centimeter of wet mud. The most recent deposits increased from 0,063 milligrams per cubic centimeter at the three foot depth to 0,140 milligrams per cubic centimeter at the mud-water interface. On a dry-weight basis the P2O5 content varied from 0.055 per cent to 0,254 per cent within the 43 foot core, with the higher value being at the surface of the mud. Converted to units of phosphorus per unit weight of dry soil the range was from 0.24 milligrams to 1.13 milligrams of phosphorus per gram of soil. No correlation between the amount of organic matter and the phosphorus content was evident in the sedimentary material of this lake. Black (1929), after analyzing bottom deposits of several Wisconsin lakes, concluded that, irthere is no sig­ nificant difference between hard and soft water lakes in the relative amounts of phosphorus in lake deposits.” data ranged from 0.24 per cent His for a soft-water lake high in organic matter to 1.39 per cent Pg05 for a lake with less organic matter and with a considerably higher alkalinity. The percentages were determined on a dry- weight basis and would approximate a range of 1.04 to 6.07 milligrams of phosphorus per gram of dry soil. Fatchichina (1939) measured FgO^ contents of the deposits of several Russian lakes and reported a range of 98 0.29 per cent to 1.65 per cent of PgO^. When these values are converted to units of phosphorus per weight of dry soil they are 1.27 milligrams to 7.20 milligrams per gram of soil. The sediments, in the lakes in which the lowest phosphorus contents were found, were autochthonous, and the sediments of the lake in which the highest phosphorus co n­ tents were found, were allochthonous. Also, the alloch- thonous sediments were higher in organic matter than the autochthonous sediments. The results of the authors cited above indicate that the variability of total phosphorus contents in the sedi­ ments of a single lake, and also in the sediments of different lakes, may be considerable. Similar results were obtained in the study of the lakes included in the present investigation. In a comparison of the total phosphorus in the profundal muds of the Pigeon River area lakes it was found that there was variation between the lakes ranging from 0.58 milligrams to 1.76 milligrams per gram of dry soil. The values obtained by Hutchinson and Wollack (1940) for Linsley Pond, Connecticut, were on a comparable order of magnitude, being 0.24 milligrams to 1.11 milligrams of phosphorus per gram of soil. Although phosphorus concentrations in the bottom muds of the Pigeon River lakes and of Linsley Pond were not as high, on a dry-weight basis, as those of the Wisconsin 99 lakes studied by Black (1929), nor those of the Russian lakes studied by Patchichina (1939), there was, nevertheless, the same trend in both the Pigeon River lakes and the Russian lakes in the correlation between the phosphorus content and the organic-matter content of their soils. In both groups of lakes, high phosphorus contents were correlated with high organic-matter contents. This relationship is to be expected when it is considered that: (1) unweathered limstone, as shown by Twenhofel (1939), contains only 0.13 milligrams of phosphorus per gram of rock (0.03 per cent P2O 5 ); (2) phosphorus contents of aquatic plants may range from 3.0 milligrams per gram of dry material for Na.jas flexilis (Schuette and Alder, 1929) to 5.98 milligrams per gram of dry material for the alga Volvox sp. (Juday and B i r ge , 1931). terrestrial soils there generally Furthermore, in o c c u r s a direct correla­ tion between phosphorus content and organic-matter content. According to Lyon and Buckman (1943) a representative peat soil has 0.20 per cent comPared with 0.10 per cent P2O5 for a representative mineral soil. Converted to units of atomic phosphorus these concentrations are equivalent to 0.875 milligrams phosphorus per gram of dry peat, and to 0.437 milligrams of phosphorus per gram of mineral soil. Th© very high phosphorus contents found by Black and Patchichina may be due to the allochthonicity and associated 100 high organic-matter contents of the lake sediments from which the samples containing these high phosphorus concen­ trations were collected. The Pigeon River lake sediments are probably also primarily allochthonous, but the lakes are not located in a region characterized by high phosphate contents in the terrestrial soils. The conclusion of Rawson (1939), which was that the productivity of lakes is primarily a function of environmental edaphic conditions, seems to be applicable here and may account for the differences in the total phosphorus contents of the sedi­ ments of lakes. The presence of phosphorus in t l ake soil does not guarantee its availability to living plants. In lake soils, as in lake waters, phosphate ions may exist as a part of different fractional entities. soluble form, They may be (1) in the (2 ) adsorbed to both organic and inorganic colloidal complexes, and (3) in molecular combination with the inorganic and organic molecules which make up the particulate matter of the bottom. Investigators who have studied the nature of these phosphorus fractions include Ohle (1937), Einsele (1938), and Mortimer (1938), Waksman, Stokes and Butler (1941). Einsele (1938) performed a number of laboratory experiments designed to measure the effect of ferrous and ferric ions on the adsorption of phosphate ions. His data indicated that an increase in 101 dissolved oxygen led to a conversion of ferrous iron to ferric Iron with a consequent formation of a colloidal complex upon which phosphate ions were adsorbed and were thereby removed from solution, Mortimer (1941) demon-- strated an increase in soluble phosphate ions in the lower hypolimnial waters of a lake in England as oxygen concen­ trations diminished. Correlated with the increase in phosphorus and decrease in oxygen in the hypolimnion were increases in iron concentration and a reduction in redox potent i a l . Other investigators have demonstrated that there is a high degree of regeneration of phosphorus Ions from organic particulate fractions to soluble fractions as a result of decomposition. Waksman, Stokes and Butler (1938) measured the rate of release of phosphorus which occurred as a result of the decomposition of diatoms. They found in a period of 420 hours a liberation of phosphorus in amounts ranging from 1.18 micrograms to 1.14 micrograms per milligram of dry diatoms. This represented nearly 70 per cent of the total phosphorus originally present in the diatoms. Barrett (ms.) has demonstrated phosphorus release from lake-bottom sediments in laboratory experiments. In these experiments 30 micrograms were released from one gram of soil in a seven-day period. The soils used in these experiments had been collected from the profundal region of 102 one of the Pigeon River lakes (North Twin). If, in nature, phosphorus is released at the same rate as in the experiments, then it can be calculated that in a lake such as North Twin, there would be released in one week, 8.1 milligrams of available phosphorus from a section of the’ bottom one square meter in area one centi­ meter in depth. A continuation of this rate of regeneration over an extended period of time could account for the quantities of available phosphorus actually found in the Pigeon River lake profundal soils. The average values of the different lake sediments as determined by laboratory analysis ranged from 1 to 355 micrograms of phosphorus per gram of dry soil. In terms of wet soil there were from 4.5 to 153.0 milligrams of available phosphorus per square meter one centimeter deep. It is apparent from these data, that phosphates are converted from insoluble organic molecules to soluble Inorganic Ions at a rate sufficiently great to produce the concent rations of phosphorus found not only in available form in lake soils, but also within the adjacent hypolimnial waters. The changes which occurred In the lower hypolimnial waters of the Pigeon River lakes may have been produced by these mobile phosphate ions moving up into the water from the muds. This hypothesis explains the increase of 66 micro- 103 grams per liter of phosphorus which occurred within a five hay period in the lower hypolimnial waters of North Twin lake in 1950. The mud of this lake at tlqat time contained 50.7 milligrams of available phosphorus per square meter one centimeter deep, a fact which further increases the probability of this process having occurred. Prior to studies of recent investigators it was thought that stagnant conditions existed throughout hypolimnia of thermally stratified lakes, and hence this lower stratum was not subject to turbulence. Hutchinson (1938), however, obtained evidence which suggested that horizontal streaming does occur in this portion of stratified lakes. He ’demonstrated changes in hypolimnial alkalinities which were of such magnitude that he was led to conclude that molecular diffusion alone could not have accounted for the observed rate of change, and therefore hypolimnial turbu­ lences must have occurred. Mortimer (1941) obtained further evidence which supported this hypothesis and he concluded that the, fl. . . movement of water masses in the hypolimnion is mainly horizontal and this is sufficient to maintain an eddy diffusion coefficient and a rate of spread of dissolved substances roughly two thousand times as great as would occur if molecular diffusion alone, were operat- ft i•v e ." If hypolimnial turbulences actually occur it can be 104 assumed that they constitute an important factor in the distribution of phosphate ions from the mud-water interface to the overlying hypolimnion. The adsorption capacity of lake-bottom soils for phosphate ions has been quantitatively studied by Patchichina (1939). 71 milligrams of This author demonstrated an uptake of P er ^ O grams of lake mud from a solution containing an initial concentration- of 5.5 m i ll i ­ grams of P2O 5 per liter. The adsorption occurred within a five minute period of constant stirring. This adsorption is equivalent to 310 milligrams of phosphorus per kilogram of dry soil. In other experiments of longer duration he obtained a maximum uptake of 991.01 milligrams of P2O 5 pe^ 100 grams of dry mud. Converted to units in terms of weight of phosphorus per kilogram of soil, this is equivalent to 4,320 milligrams of phosphorus per kilogram of dry soil. Because phosphorus adsorption may occur at an expo­ nential rate, as demonstrated by Hayes and Coffin (1951), and, because the quantitative degree of phosphorus adsorp­ tion may be on the order of 4,000 milligrams per kilogram of soil, (Fatchichina, 1939), it seems reasonable to assume that a large amount of the phosphorus added artificially to natural lakes will be rapidly adsorbed by bottom soils. The nearly complete disappearance of the phosphorus added to the epilimnia of the Figeon River lakes probably was due 105 in part to a direct, or indirect, adsorption mechanism at the mud-water Interface* If the Pigeon River lakes had as high an adsorption capacity as the Russian lakes studied by Fatchichina, this degree of adsorption by the profundal soils would be, with certain exceptions, entirely possible. Table XXVII lists the rate of phosphorus application on the basis of milligrams of phosphorus applied for every 1,000 grams of dry profundal soil located within the top centi­ meter of the bottom mud. This value is comparable to the term p.p.m. used by agronomists in the fertilization of terrestrial soils. In agronomy the terra p.p.m. is, by convention, arbitrarily accepted as meaning the ratio of the number of parts by weight of a constituent to the number of parts by weight of the soil located within the upper six Inches of the surface soil. Thus an application of 10 p.p.m. of phosphorus Is equivalent to adding 10 pounds of phosphorus to 1 ,000,000 pounds of soils within the top 6 inches. This rate of application may be converted to pounds per acre by using the generally accepted figure of 2,000,000 pounds per acre-six inches. In lakes the term milligrams of phosphorus per kilogram of soil (in the top centimeter) may be converted to milligrams per square meter-centimeter by multiplying the number of kilograms of dry soil in one square meter-centimeter (10,000 cubic centi­ meters) by the number of milligrams of phosphorus in one 106 kilogram of dry soil. TABLE XXVII RATE OF PHOSPHORUS APPLICATION PER UNIT DRY-WEIGHT OF PROFUNDAL SOILS' Units in milligrams per kilogram of dry soil one centimeter deep West Lost 1949 988 2890 2540 350 1950 930 2830 2350 327 1938 5720 4890 677 Total Lost Section Four Year Hemlock Hemlock and Lost lakes were the only two lakes which received more phosphorus than the 4,000 milligrams per kilogram capacity demonstrated by Fatchichina. The total amount put into Hemlock lake was 5,700 milligrams per kilo­ gram of soil, and the total amount put Into Lost lake was 4,890 milligrams per kilogram of soil. These two lakes showed the greatest hypolimnial phosphorus increases in the waters adjacent to profundal soils (Fig. 5). This suggests that the maximum adsorption capacity of the profundal soils had been exceeded. The general theory of exchange equilib­ ria as the mechanism which primarily determines phosphorus concentrations in lake waters is thus given excellent s upport. The data available for Section Four lake are also of particular significance. This lake received 677 milligrams of phosphorus per kilogram of soil. Fig. 3 demonstrates that, on a comparative basis, no large Increases in phos- Fig. 3. Comparisons of concentrations of hypolimnial total phosphorus and of available phosphorus in profundal soils. Upper portion of graph shows hypolimnial phosphorus. Units In micrograms per liter. profundal phosphorus. air-dry soil. Lower portion of graph shows Units in milligrams per kilogram of 107 pg/i IT- 800 A 1949 1950 I 1951 600 400 200 V. Vs 11_Q 'A mg/kg a . 4 00 3 00 200 00 NORTH TWIN SOUTH TWI N £ HEMLOCK LO ST SEC. FOUR 103 phorus concentrations occurred in either the hypolimnion or in the adsorbed and acid-soluble complex within the p r o ­ fundal soils. These results are explained by the hypothesis that phosphate ions become effectively fixed in the non­ exchangeable fraction when added to lakes in which there are high concentrations of homogeneous marl, as were present in Section Pour. The importance of organic matter on phosphorus fixa*tion by profundal soils becomes apparent when the data for Lost lake are compared with those for Section Pour. Al­ though the amount of phosphorus in the exchangeable phase was approximately the same in each lake, the quantities in the hypolimnial phase were significantly different. Lost lake contained 43 per cent organic matter compared to 6 per cent for Section Four, and correlated with the higher organic matter content in Lost lake sediments were greater hypolimnial concentrations of phosphorus. These data are further support for the general hypothesis that one, of .the mechanisms which control adsorption and regeneration of phosphorus in lakes is that there exists an equilibrium between the fixed-exchangeable-soluble phases of an exchange system, and that each phase within the system is character­ ized by having a capacity for phosphate ions the maximum of which depends upon (1 ) the concentration of phosphate, or of phosphate-exchangeable ions within each phase, and (2 ) the 109 ratio of the per cent organic matter to per cent marl in the particulate and soluble phases of the system* This hypothesis not only explains the data already cited but also other data concerned with adsorption and regeneration of phosphorus. For example, Lost lake did not increase in the amount of phosphorus in the adsorbed fraction of the profundal soils, but did increase in the hypolimnial waters. Thus, apparently, maximum quantities of phosphate ions had been fixed in the insoluble form within the particulate phase, and. the adsorptive capacity of the exchange fraction (exchange-threshold) had also been exceeded, thereby causing the excess phosphate ions to move out into solution. The situation in Hemlock lake may also be explained on the basis of the exchange-mechanism hypothesis. In this lake an increase in both the exchange-interphase component and the hypolimnial component occurred following addition of phosphate fertilizers. Apparently, the adsorption capacity of the fixed complex was satisfied either previous to or as a result of fertilization, because not only the interphase complex increased in concentration, but the hypolimnial phase also increased. This conclusion suggests that a certain amount of excess phosphorus had been applied with the added fertilizers, and that the excess had moved out into the hypolimnial complex. Because Lost lake and Hemlock lake were fertilized at comparable rates of 4,890 milligrams 110 and 5,720 milligrams per gram of soil respectively, and because Hemlock lake increased markedly in the adsorbed phosphorus fraction, while Lost lake did not, it can be assumed that a higher interphase threshold was present in the profundal soil of Hemlock lake, and that the greater threshold capacity was due to greater concentrations of organic matter. Had the organic matter content of Hemlock lake been less, a lower exchange capacity probably would have existed, and the same quantity of added phosphorus would theoretically have saturated both the fixed fraction capacity and the exchange fraction capacity, and resulted in hypolimnial phosphorus increases on the same order of magnitude as those which occurred in Lost lake. An extension of the exchange hypothesis to the West Lost lake problem is also indicated, since the concentration of phosphorus in the adsorbed fraction increased, but no hypolimnial increases occurred following application of fertilizers. The total amount of applied phosphorus was only 1,920 milligrams per gram of soil, which was apparently not enough to satisfy the combined adsorption capacities of the fixed and the interphase complexes. The effect of calcium on the exchange of phosphorus across iron and humus colloidal interfaces was suggested as early as 1937 by O h l e . O h l e 1s laboratory experiments indicated that increases in Ca(HC0 3 )2 concentrations caused Ill increases in soluble phosphate ions in solutions containing iron-phosphate colloidal complexes. These results led him to speculate that marl would be an important factor affecting phosphorus adsorption and regeneration in natural waters rich in iron and humus colloids. The results of the present report are thus seen to support the hypothesis of Ohle. If the general concept of the exchange properties of lake soils as outlined above is true, then it may be nec­ essary to modify (1 ) general theories concerning product­ ivity of lakes and (2 ) certain aspects of some of the methods now followed in lake and pond fertilization programs. For instance, the possibility of reducing the phosphorusfixation capacity of marl, soils in lakes need more experimental investigation. It may prove to be profitable to apply organic matter in conjunction with inorganic phos ­ phorus when fertilizing lakes which are high In homogeneous marl. Calcium is probably not only important as a regulator of the iron-humus-colloid-exchange mechanism, but also as an agent responsible for the formation of insoluble phosphate compounds. Agronomists know that in agricultural soils phosphorus unites with calcium to form insoluble compounds as the pH increases above 7.6. The experimental work of the present study demonstrates that pH alone cannot always 112 be used as a criterion for determining phosphate precipi­ tation in calcium carbonate solutions. Furthermore, the concept that there exists a minimum concentration of dissolved calcium below which phosphorus is not precipitated, as suggested by Naumann (1932), may be true; but there are probably several factors which affect the formation of insoluble calcium phosphate compounds, and until these factors are known, Naumann* s concept of a specific minimum limitation of the "alkalitrophic11 condition seems to be an oversimplification of a complex phenomenon. The present study has demonstrated phosphorus precip­ itation in a solution containing an initial concentratIon of 216 p.p.m. methyl-orange alkalinity. without This reaction occurred the solution being subjected to agitation. The 216 p.p.m. concentration alkalinity is equivalent to approxi­ mately 118 p.p.m. of CaO, a value only slightly higher than the 100 p.p.m. CaO suggested by Naumann as the lower limit of phosphorus precipitation. Other solutions, having initial methyl-orange alkalinities of 168 p.p.m., produced precipitates when agitated, but did not form precipitates when not agitated. 94 p.p.m. CaO. A value of 163 p.p.m. is equivalent to Phosphorus precipitation even occurred at methyl-orange alkalinities of 116 p.p.m. (a value equiva­ lent to 65 p.p.m. CaO) when the solutions were continuously agitated, and when the phosphorus and potassium concentra- 113 tions were low* Because non-agitated calcium bicarbonate solutions decreased in methyl-orange alkalinity from an initial con­ centration of 216 p.p.m. to 60 p.p.m., and because other solutions containing 72, 116, and 168 p.p.m. did not decrease, but remained essentially constant, the assumption was made that the principles of solubility product con­ stants may not be entirely responsible for marl formation, or for phosphorus precipitation by calcium. Were this the case, identical CaCHCO^)^ concentrations should have developed in each flask. An adsorption mechanism could account for the differences which did occur. Furthermore, if the principles of solubility product constants were operating, an increase in a common ion should accelerate precipitation, but this did not occur. quantities of phosphorus, When greater in the form of KH^PC^ were added to calcium bicarbonate solutions, both calcium and ph os­ phorus precipitation were inhibited. mechanism concept The adsorption explains this behavior, for it is known that potassium ions possess a lower coagulating influence on colloidal particles than calcium ions. Until the mechanisms of phosphorus precipitation have become definitely established, and until the factors affecting these mechanisms are also understood, it is doubtful that a shprp line of demarkation can arbitrarily be established as a 114 criterion differentiating an alkaline-induced phosphate oligotype from a phosphate mesotype or polytype, suggested by Naumann (1932) and Hubalt as (1943). In general, it appears that phosphate fertilizers \ which are; added to stratified lakes of the limestone-sink type cannot be expected to remain in the epilimnia of these lakes for more than a few months, at least in concentrations comparable to those existing during periods immediately following fertilizer applications. Possibly a small percentage remains in the epilimnia the following year, but there is no evidence that this is significantly different from amounts' present before fertilization. The fate of added phosphorus appears to be regulated not only by the nature of the bottom sediment, but also by the alkalinity of the lake waters. Soils consisting of homogeneous marl are capable of adsorbing and fixing in a non-exchangeable form most of the added phosphorus. The amount of phosphorus present in any of the three phases of the insoluble-exchangeable-soluble equilibrium system depends upon the capacity of each phase, and also upon the concentration of phosphate ions within each phase. Any change In phosphate concentration on either side of the mud-water interface is followed by a re-establishment of a state of equilibrium. There appears to be a theoretical organic matter-marl ratio which favors a transference of 115 phosphate Ions from the exchangeable fraction to the sol­ uble fraction. If the organic matter-marl ratio exceeds this point the adsorptive capacity of the exchangeable fraction at the mud-water interface increases. If the amount of phosphorus added to the lakes is not sufficient to exceed the adsorptive capacity not remain in solution. of the bottom muds it will If the amount added does exceed the adsorptive capacity of the sediments, and if the lakes are stratified, the excess phosphorus will eventually appear In the hypolimnial waters, but apparently will not appear in discernable amounts in the epilimnial waters. With certain exceptions, the greater the alkalinity of epilimnial waters the greater the chances for the formation of insoluble calcium phosphate compounds. Alkalinity values above 200 p.p.m. can be expected to produce insoluble phosphate compounds at a relatively rapid rate. Values of 116 p.p.m. also cause phosphorus precipitation, but probably only If certain restrictive factors are absent. summary 1. Periodic applications of an inorganic fertilizer containing phosphorus were made to four of six small exper­ imental, limestone-sink lakes In northern Michigan in the summers of 1949 and 1950 in order to find out (1) how long added phosphorus would remain in epilimnial waters of trout lakes after application, and (2) the effects of alkalinity on retention of added phosphorus in epilimnial waters. The alkalinities of the lakes ranged from 34 to 192 p.p.m. 2. Selection of amounts of fertilizer to be added to the lakes was based on (1) previous fertilization experi­ ments on similar lakes, and (2) the differences in alkalin­ ities of the lakes. With an increase in alkalinity progress ively greater quantities of fertilizer were applied. 3. Chemical analyses were performed for total phos­ phorus contents on samples of water collected just previous to, three days following, and three weeks following ferti­ lizer applications In 1949 and 1950, and at three-week intervals in 1951, the year following fertilization. Chemical analyses for phosphorus were also made of the littoral, sublittoral, and profundal soils. 4. The water samples were collected from the top of \ 117 the epilimnion and from the bottom of the hypolimnion near the center of the lakes. 5. Data on methyl-orange alkalinity, lake tempera­ tures, organic-matter content of profundal soils, and chemical characteristics of surrounding terrestrial soils were also collected. 6 . Several laboratory experiments were performed to determine concentrations of cald.ium carbonate necessary to cause precipitation of phosphorus. 7. During the .first year of the addition of ferti­ lizer the average total phosphorus concentrations in epilimnial waters of lakes receiving phosphate fertilizers was approximately one-half of the theoretical concentration which would have been present if all the phosphorus which was added had gone into solution and had remained only until the next application three weeks later. 8 . During the first year of fertilization the lakes receiving fertilizer contained epilimnial concentrations of total phosphorus which ranged from 3 to 17 times as much as the control lakes. 9. During the second year of fertilization the control lake epilimnial total phosphorus concentration averages were as high as those of the fertilized lakes, but 118 I this was attributed to excessive rainfall and intensive leaching of surrounding terrestrial soils. 10. The average total phosphorus concentrations in epilimnial waters of both fertilized and unfertilized lakes during the year after cessation of fertilization applications were all very low and there was no reason to believe that a significant amount of residual phosphorus of the previous .two years applications was present. 11. The fluctuations in epilimnial total phosphorus concentrations ranged to increases of 137 micrograms per liter during a three-week period in an unfertilized lake (caused apparently by heavy rainfall) and to decreases of 155 micrograms per liter in a five-day period in one of the 'fertilized lakes. The maximum increase in a three-week period in a fertilized lake was 473 micrograms per liter which occurred following an application of 80.8 micrograms per liter to the lake. 12. Total phosphorus concentrations in hypolimnial waters were about the same in the fertilized and unferti0 lized lakes during the first year of fertilization. During the sec.ond year of fertilization, and the year after, the fertilized lakes had considerably higher concentrations •than the unfertilized lakes. 13. No statistically significant increases in avail- 119 able phosphorus in profundal soils occurred following fertilization in the. two lakes receiving greatest quanti­ ties of phosphorus applications. The lakes which received least amounts of phosphorus and a control lake demonstrated statistically significant increases between 1949 and 1951. 14. The three-year averages of available phosphorus in profundal soils were highest in the unfertilized lakes and in the lake which- received the least quantity of fertilizer. These high concentrations are correlated with high organic-matter contents. 15. Quantities of available phosphorus in profundal soils did not seem to be correlated directly with quantities or total phosphorus in adjacent hypolimnial waters and it would seem inadviseable to base estimates of potential lake-water nutrient phosphorus supplies upon measurements of available phosphorus in profundal soils. 16. Total phosphorus contents of profundal soils were correlated with organic-matter contents, but not with amounts of phosphorus added to the lakes. 17. There was no indication that quantities of total phosphorus in profundal soils increased as a result of the application of fertilizers. If increases actually occurred they were not demonstrated by the methods employed In this study. 120 18. Concentrations of available phosphorus in littoral and sublittoral soils were correlated more directly with organic-matter contents than with amounts of ferti­ lizer applied. 19. Considerable changes in relative differences of available phosphorus in littoral and sublittoral soils between different lakes may occur when results are converted from a dry-weight basis to a wet-volume basis. It is therefore suggested that studies of available nutrients of lake' soils Include these conversion data. 20. No correlations were shown to exist between edaphic phosphorus and lake-water or lake-sediment phos­ phorus contents. This is probably due to extremely low nutrient levels in terrestrial soils of this region, and to the characteristically low mobility of phosphorus in soil. 21. The maximum per cent residual phosphorus of that applied to any of the lakes was 56 per cent at the end of the first season; 32 per cent at the end of the second season, and 8 per cent at the end of the third season. 22. The differences in per cent residual phosphorus between the different lakes was shown t o ‘be statistically significant, and to be statistically related to average 121 alkalinity concentrations when tested by a regression a na lysis• 23. No statistically significant differences were found to exist between !,normal,r phosphorus levels of four of the six lakes investigated. These four lakes ranged from 104 p.p.m. to 170 p.p.m. methyl-orange alkalinity. Thus it could not be concluded that within these ranges of alkalin­ ity "normal” epilimnial phosphorus levels are affected by alkalinity. 24. In laboratory experiments It was shown that at 168 p.p.m. methyl-orange alkalinity, and at pH of 8.5,. phosphorus will not precipitate when present at a concen­ tration of 100 micrograms per liter if the solution is not agitated. It was concluded that pH may not be a safe criterion on which to base predictions of phosphorus precip­ itation in combination with calcium, as is done by agronomists for terrestrial soils. 25. Phosphorus precipitation occurred in calcium carbonate solutions which had 116 p.p.m. methyl-orange alkalinity. This happened when the solutions were contin­ uously agitated. The lower limit of the alkalitrophic condition at which phosphorus precipitates, which Naumann (1932) believed to be 100 p.p.m. of CaO, thus apparently may be as low as 65 p.p.m. CaO. 122 26. There was evidence to support the assumption that the precipitation of phosphorus by calcium carbonate may be an adsorption mechanism rather than the molecular formation of tri-calcium phosphate as proposed by Schlosing (1900) . 27. The concentrations of total phosphorus found' In the profundal Sediments of the Pigeon River lakes were on the same order of magnitude as those found in Linsley Pond, Connecticut, but were not as high as those in the lake bottom .sediments of some Wisconsin and some Russian lakes. 28. Following fertilizer applications the distri­ bution of phosphate ions in (l) the lower hypolimnial waters; (2) the available exchange complex of the sediments, and (3) the fixed complex of the sediments, supported the general assumption that' an exchange mechanism operates between the water and the mud. 29. The capacity of each phase of the exchange system for phosphorus Is correlated with marl and organicmatter contents of the sediments. hypothesis of Ohle This supports the (1937) that phosphates In lake waters are largely affected by the organic matter-marl ratio. 30. The capacity for phosphate ions of the exchange complex in the equilibrium system of lake sediments is 123 directly related to the concentration of organic matter and Inversely related to the concentration of marl. 31. If the capacity of the exchange complex in lake sediments for phosphorus Is exceeded the excess phosphate ions will either be fixed in the insoluble phase, if high marl concentrations are present, or will appear in the adjacent waters if the organic matter-marl ratio favors such movement. In sediments high In organic matter the adsorptive capacity of the exchange complex is very high. 32. On the basis of the very slight percentages of residual phosphorus which remained in the epilimnia of the fertilized lakes during the year following fertilizer applications, It is concluded that fertilization of lakes of the limestone-sink type will have only temporary effects in increasing the supply of available phosphorus in the epilimnial waters. LITERATURE CITED Ball, R. C. 1948a. Fertilization of lakes, good or bad? Conservation, 17 (9): 7-14. Michigan Ball, R. C. 1948b. A summary of experiments in Michigan lakes on the elimination of fish populations with rotenone, 1934-1942. Trans. Amer. Fish Soc., 75 (1945): 139146. Ball, R. C. 1950. Fertilization of natural lakes In Michigan. Trans. Amer. Fish Soc., 78 (1948): 145-155. Barrett, P. H. 1951. Effects of phosphorus on decomposition, of lacustrine sediments in laboratory experiments. Unpublished manuscript, 1-6. BIrge, E. A., and C. Juday. 1911. The Inland lakes of Wisconsin. The dissolved gasses of the water and their biological significance. Wise. Geol. Nat. Hist. Surv., Bull. (22): 1-259. BIrge, E. A., and C. Juday. 1934. Particulate and dissolved organic matter in Inland lakes. Ecol. Monogr., 4: 440-474. Black, C. S. 1929. Chemical analysis of lake deposits. Trans. Wise. Acad. Sci. Arts Lett., 24: 127-133. Bray, R. H., and L. T. Kurtz. 1945. Determination of total, organic and available forms of phosphorus In soils. Soil Sci., 5 9 r 39-45. Burr, G. 0. 1941. Photosynthesis of algae and other aquatic plants A symposium of hydrobiology: 163-181. Univ. of Wiscon­ sin Press, Madison. Deevey, E. S., Jr. 1940. LImnological studies in Connecticut. V. A contribution to regional limnology. Amer. Jour. Sci., 238: 717-741. 125 Dice, L. R., and H. J. Leraas. 1936. A graphic method for comparing several sets of measurements. Univ. of Michigan, Contr. Lab. Vert. Genetics., 3: 1-3. Egglet o n , F. E. 1931. A limnological study of the profundal bottom fauna of certain fresh-water lakes. Ecol. Monogr., 1: 231-332. Einsele, W. 1938. Uber chemische und kolloidchemische Vorgange in Elsen-Phosphat-Systern unter limnochemischen und limnogeologischen Gesichtspunkten. Arch. fur. Hydrobiol., 33: 361-387. Einsele, W* 1941. Die TJmsetzung von zugefuhrten, anorganische Phosphat im eutrophen See und ihre Ruckwirkungen auf seinen Gesamthaushalt. Zeischr. fur Fish, und deren Hilfswissensch., 39: 407-488. Ellis, M. M., B. A. Westfall, and M. D. Ellis. 1946. Determination of water quality. U. S. Dept. Inter., Fish and Wildlife Serv., Research Rept., (9): 1-122. Eschmeyer, R. W. 1938. Experimental management of a group of small Michigan lakes. Trans. Amer. Fish. Soc., 67 (1937): 120-129. Fatchichlna, 0. A. 1939. The absorption capacity of the lake deposits* (Summary in English.) Proc. of the Kossino Limnol. Sta. of the Hydromet* Serv • of USSR, 22: 5-34. Gessner, F. 1939. Die Phosphorarmut der Gewasser und Ihre Beziehung zum Kalkgehalt. Intern. Rev. der ges. Hydro­ biol. und Hydrogr., 38: 203-211. Hayes, F. R . , and C. C. Coffin. 1951. Radioactive phosphorus and exchange of lake nutrients. Endeavor, 10 (38) : 78-81. Hubalt, E. 1943. Les grands lacs subalpine de Savoi sont-Ils alcalitrophes? Arch. Hydrobiol., 40 (1): 240-249. 126 Hutchinson, G. E. 1938. Chemical stratification and lake morphometry. Proc. Nat. Acad. Sci., 24: 63-69. Hutchinson, G. E. 1941. Limnologlcal studies In Connecticut. IV. Mechan­ ism of intermediary metabolism in stratified lakes. Ecol. Monogr., 11: 21-60. Hutchinson, G. E., and A. Wollack. 1940. Studies on Connecticut lake sediments. II. Chemical analysis of a core from Lins ley Pond, North Branford. Amer. Jour. Sci., 238: 493-517. Juday, C., and E. A. BIrge. 1931. A second report on the phosphorus content of Wisconsin lake waters. Trans. Wise. Acad. Sci. Arts Lett., 26: 353-382. Lyon, T. L * , and H. 0. Buckman. 1943. The nature and property of soils. Co., New York, N. Y. 499 pp. The MacMillan Marshall, C. E. 1949. The colloidal chemistry of the silicate minerals. Academic Press Inc. New York, N. Y. 195 pp. Mortimer, C. H. 1941. The exchange of dissolved substances between mud and water in the lakes. Jour. Ecol., 29: 280-329. Mortimer, C. H. 1942. The exchange of dissolved substances between mud and water In lakes. Jour. Ecol., 30: 147-201. Mortimer, C. H. 1949. Seasonal changes in chemical conditions near the mud surface In two lakes of the English Lake District. Verhandl. der Intern. Vereinig. fur theoret. und angew. LImnol., 10: 353-356. Naumann, E. 1932. Grundzuge der regionalen LImnologie. Blnnengewasser, 11: 1-176, Ohle, W. 1934. Chemische und physikalische Untersuchungen norddeutcher Seen. Arch. Hydrobiol., 26: 386-464, 584-658. 127 Ohle, W . 1937. Kolloidgele als Nahrstoffregulatoren der Gewasser. Die Naturwissensch., 29: 471-474. Ohle, W. 1938. Zur Vervollkommnung der hydrochemische Analyse. III. Die Phosphorbestimmung. Angew. Chemie, 51: 906-911. Pearsall, W. H* 1930. Phytoplankton In English lakes. I. The propor­ tions in the water of some dissolved substances of biological importance. Jour. Ecol., 18: 306-320. Peech, M., L. T. Alexander, L. A. Dean, and J. F. Ford. 1947. Methods of soil analysis for soil-fertllity Investigations."!!. S. Dept. Agr, CIrc., (757): 1-25. Piper, C. S. 1950. Soil and plant analysis. Interscience Publishers Inc. 368 pp. New York, N. Y. Prescott, G. W. 1939. Some relationships of phytoplankton to limnology and aquatic biology. Problems of lake biology. Amer. Assoc. Adv. Sci. Publ., (10): 65-78. Prescott, G. W. 1951* Algae of the western Great Lakes area. Cranbrook Inst, of Sci. Bull., (30). Bloomfield Hills, Mich. 946 pp. Rawson, D. S. 1939. Some physical and chemical factors in the metab­ olism of lakes* Problems of lake biology. Amer, Assoc. Adv. Sci. Publ., (10): 9-26. Roelofs, E. W, 1944. Water soils in relation to lake productivity. Michigan State Coll. Agr. Exp. Sta. Techn. Bull., (190): 1-31. Schaeperclaus, W. 1933. Lehrbuch der Teichwirtschaft. Verlag Paul Parey. Berlin. 289 pp. Schlosing, T. 1900. L facide phosphorique en presence des dissolutions saturees de bicarbonate de chaux. Comptes Rendus de L fAcademie des Science, 131: 211-215. 128 Schuette, H. A., and H. Alder. 1929. Notes on the chemical composition of some of the larger aquatic plants of Lake Mendota. III. Castalla odorata and Naja flexilis. Trans. Wise. Acad"! Sci. Arts Lett., 24: 135-145. Simpson, G. G., and A. Roe. 1939. Quantitative zoology. Numerical concepts and methods in the study of recent and fossil animals. McGraw-Hill Book Co. New York, N. Y. 414 pp. Snedecor, G. W. 1946. Statistical methods. The Iowa State College Press, Ames, Iowa. 485 pp. Spurway, C. H., and K. Lawton. 1949. Soil testing. A practical system of soil fertil­ ity diagnosis. Michigan State Coll. Agr. Exp. Sta., Techn. Bull., (132): 1-41. "Standard Methods for the Examination of Water and Sewage." 1946. Ninth Edition. American Public Health Assoc. 286 pp. Swingle, H. S. 1947. Experiments on pond fertilization. Alabama Exp. Sta., Alabama Poly. Inst. Bull., (264): 1-34. Tanner, H. A. 1952. Experimental fertilization of Michigan trbut lakes. Michigan State Coll., Unpublished thesis. Theroux, F. R . , E* R. Eldridge, and W. L. Mallmann. 1943. Laboratory manual for chemical and bacterial analysis of water and sewage. McGraw-Hill Book'Co. New York, N. Y. 274 pp. Tressler, W. L . , and R. Bere. 1935. Plankton studies in some lakes of the MohawkHudson Watershed. New York State Cons. Dept. Ann. Rept. 24 (Suppl.), Biol. Surv., 9: 250-266. Twenhofel, W. H. 1939. The cost of soil in rock and time. Amer. ^our. Sci., 237: 769-780. Veatch, J. 0. 1931. Classification of water soils is proposed. Michigan State Coll. Agr. E x d . Sta. Quart. Bull., 14 (1): 20-30.