TH! IFFECTS 0' W CONCENTRATDNS OF NITROGEN ON PRIMARY PRODUCTION 1" MI Msbrflnmdlfi MW STAT-I W Robtrt Mitch." SM» 1960 IUIWIILIIIIIIHIIUHIHUI!"HHIHlllllHUIHIIHIHW 310402 4017 ‘ LIBRARY Michigan Stan . . 511E515 ' 1923535 THE EFFECTS OF LIMITIKG CONCELTRATIONS OF NITROGEN ON PRIMARY PRODUCTION IN AN ARTIFICIAL STREAM by Robert Mitchell Stokes AN ABSTRACT Submitted to the College of Agriculture of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIEIGE Department of Fisheries and Wildlife 1960 Approved: Robert Mitchell Stokes ABSTRACT The effects of growth-limiting nitrogen concentrations u upon lotic primary production were studied under the con- trolled environment of an artificial stream. The stream water contained an excess of all major nutrients except nitrogen. Effects of variable water velocity (riffle and pool areas) and light intensity were also examined. Three successive additions of this element (1 mg 1" each) brought about the establishment of an attached algal community followed by increased levels of phytopigment, or- ganic nitrogen, and total dry weight. The concentration of all major nutrients which were measured decreased with growth of periphyton. The reduction of each new supply of nitrogen was inversely preportional to a rise in cellular chloropiyll. After nitrate reduction the chlorOphyll con— tent decreased. Gorrelation coefficients indicated tiat the relation- ship of pigment concentration to total dry weight and pig- 5 , K , , , .. ment to organ10‘were nign. The cellular nitrogen content approximated two percent on a dry weight basis. L. f‘ fhe production of organic matter (measured as total dry weight), phytOpigment, and organic nitrogen was sig- nificantly different between riffle and pool areas. Early in the project pool values were higher than riffle values. Later the converse of this situation occurred. Similar fluctuations occurred in the percent of cellular nitrogen. This difference between areas could have been produced by three factors: variable incident radiation, variable water velocity, and competition. The incident radiation on pool and riffle surfaces approximated daylight. This radiation along with limiting nitrogen concentration accounted for a primary production of 250.2 and 201.h mg dry wt deaydin the riffle and pool respectively. In diminishing light intensities p ytOpig- ment production decreased, but even in this area production increased with nitrocen additions. {ESIS THE EFFECTS OF LIMITING CONCENTRATIONS OF NITROGEN ON PRIMARY PRODUCTION IN AN ARTIFICIAL STREAM by Robert Mitchell Stokes A.THESIS Submitted to the College of Agriculture of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1960 G .434 ~x 'V'x r0 4 \‘3\\ ‘3 ACKNOWLEDGMENTS The author wishes to extend appreciation and gratitude to Dr. Robert C. Ball for his guidance and recommendations throughout this study. He is also indebted to Dr. Phillip J. Clark for advice on statistical procedure; 1r. Phillip Halicki for aid in the identification of algae; Mrs. Mary Wacasey for flame photometric analysis; the graduate stu- dents of the Fisheries and Wildlife Department for many .helpful suggestions and discussions; and to his wife Carole for her constant encouragement and hours Spent typing this thesis. This study was conducted under a graduate research assistantship from the National Institute of Health. ii TABLE OF CONTENTS INTRODUCTION . o . . . . . Nitrogen Fixation . S UI‘H‘IARY o o o o o a METRODS Equiyme nt . . . . . Nutrient Medium . . . Water Chemistry . . . Periphyton Analysis. . Light . . . . . . RESULTS Species Composition. . Water Chemistry . . . Periphyton Analysis. . Light . . . . . . iii 10 13 15 23 25 67 71+ 76 TABLE l. 10. ll. 12. 13. LIST OF TABLES Concentration of Major Salts (g 1”), Major Elements (mg l“), and Micro- elements in Nutrient Medium B Contain— ing Ag Microelements Stock Solution Inclu ing Modifications . . . . . . Water Chemistry. . . . . . . . . Nitrogen and Phosphorus Determinations in Milligrams per Liter . . . . . . Potassium, Sodium, and Silica Ion Con- centrations and Corrected Total Reduc— tion of Each in Milligrams per Liter. . Riffle PhyotOpigment Absorbancy per Unit Area After Two Weeks EXposure . . . . Pool Phytopigment Absorbancy per Unit Area After Two Weeks Exposure . . . . Analysis of Variance of Riffle and Pool Areas for Four Periods of Shingle Exposure . . . . . . . . . . . Mean Phytopigment Absorbancy Units and Milligrams Organic Nitrogen per Period . Milligrams of Organic Nitrogen per Unit Area During a Two Week Exposure Period in tlle P001 Free. 0 o o o o o o o Milligrams of Organic Nitrogen per Unit Area During a Two Week Exposure Period in the Riffle Area. 0 o o o e o o PhytOpigment per Unit Dry Weight per Unit Area After Two Weeks of ExPosure . Percent Cellular Nitrogen per Unit Area After Two Weeks Exposure at the Begin- nings and Ends of Periods 2, 3, and h . Artificial Stream Primary Production. . iv 11 32 36 U}. \C') 62 66 V' 1 e'u .‘ | {L'é‘a TABLE lu. LIST OF TABLES (Cont.) Page Mean Absorbancy Units (AAXlO3) of Phyto- pigment per Unit Area After Exposure to Four Periods of Nitrogen at Decreasing Intensities of Light Energy g-cal cmfmin? 72 meals Figure l. 2. 3. 10. ll. 12. 130 LIST OF FIGURES Photograph of Artificial Stream. . Diagram of Artificial Stream. . . . Diagrans of Substrate Arrangement in Artificial Stream and Dates Sampled by Each Shingle Set. . . . . . . Correction Graph for Adjusting Measured PhytOpigment Absorbancy Values to Units Related to Concentration . . . . . Correction Graph for Converting Expo- sure Meter Readings to Gram-calories per Square Centimeter per Minute . . Alkalinity and pH in Artificial Stream 'C‘J 8.t e r O O O O I O O I O O 0 Specific Conductance of Artificial Stre 8171 L‘I’atcr . Q Q Q g Q Q o 0 Total Phosphorus in Artificial Stream Fla-tor o o o o o o o o o o 0 Potassium, Sodium, and Silica Ion Con- centrations of the Artificial Stream . Total Available Nitrogen and Ammonia Nitrogen in the Artificial Stream . . Mean Phytopigment Absorbancy Units per Unit Area After Two Weeks Exposure. Arrows Indicate Bi-weekly Nitrogen Additions . . . . . . . . . . Mean PlytOpigment Absorbancy Units per PeriOd o o o o o o o o o o 0 Mean Organic Nitrogen per Unit Area After Two Weeks Exposure. Arrows Indi- cate Bi-weekly Nitrogen Additions . . Mean Milligrams of Organic Nitrogen per PeriOd o o o o o o o o o o 0 vi 19 22 27 30 33 37 53 ELL Figure 15. lo. 17. 18. 20. LIST OF FIGURES (Cont.) Regression Lines Expressing the Re- lationship Between PhytOpigment Absorb— ancy Units and Milligrams of Organic Nitrogen for Riffle and Pool Zones . Regression Expressing the Relationship of PhytOpigment Absorbancy and Milli- grams of Total Dry Weight for All Arti- ficial Stream Communities . . . . . Percent of Cellular Nitrogen at the Beginnings and Ends of Periods 2, 3, and 14.. o o o o o o o o o o 0 Primary Production of Riffle Area in Grams per Square Meter After Two HOOKS Exposure to Periods 2, 3, h, and S . . Primary Production of Pool Area in Grams per Square Meter After Two Weeks Expo- sure to Periods 2, 3, A, and S . . . Mean PhyOpigment Absorbancy Units per Unit Area at Decreasing Intensities of Light Energy for Periods 2, 3, L, and 5 vii 57 60 68 69 mu". L" ‘ J“ ‘L'dls .Alilolll'l ll INTRODUCTION In recent years the rapid ex;ansion of the human pOp- ' I)? {JCS o ulation 12s CTGLth many problems which affect our 37 One 0 the major problems rWs ult: from a need for increased sport and food fish production while parad.oxically an in- crease of pollution is unfavorably altering many of the pre :3 ent environment 3 . Many linniolOgists have looked to pr‘nary “”OdLColOD on basic SDPBJH ecologg, for the which has direct bearin answer to this perplexitV. Prinary production is the rate a - 1 .1. ° .. s ,. -,_‘.. -‘ - n,“ .73 ,_ .‘- - , M at whicn or inic M"tbCT 18 formed b” Pgutogthuuwtlc Jud ‘ n "V v KL. '.' I 1" " . ‘9— '5 ‘~ 1' N. A "fl‘, a 1' ‘a ’. r» ',‘\ r. cnenosynche ic SCciVlLO oi producers iroi sleic raw L.— teria ls (Od11 lSS3, Ruttncr 1953). If tne efficiency of 't‘rl . -- V) '1 N. "‘ CI ('11 v x '. 1‘ r1fifirxt‘ n.\d r" ' ,‘v r)“-'s1 ‘? .". 3 "I 1‘} 1- r\ t . \ 1 -. 1.8 I)- O‘w'x/SV )(J .Lil'k/J. C'Cuo‘g 91..“ ‘. LI'O L‘ivii' - .t. 4C1 .LLJ. a Hr,\c lC/a R 1d suc;ws Hf 11 manner, a broader base would be provided for the odu ction of censun,1 etch {2 fish. Vari: ti on in primary production rates and the composition of the F. I 81” ‘73 phItOD-CO~.~.ity may 5180 provide a useful m. thod of detect- ing sub-1;;nal 3nd chronic pollutiO* Q -. ar'ificivl streat FJ Keeping these problems in mind, a1 was constructed in which communities of periphyton could be grown under controlled environmental conditions. The present study included investigations of community growth patterns under limiting concentrations of nitrOgen with 2 variations of light intensity and current. All other physi- cal and chemical conditions were held constant. Although there have been numerous studies on the subjects of primary production, photosynthesis, and nutrient metabolism in both natural waters and laboratory vessel cultures, few have incorporated the use of an artificial stream. Odum and Hoskin (1957) studied the metabolism.of algal communities under artificial stream conditions in a recirculating apparatus which consisted pr'narily of clear plastic tubing. Several other projects employing artificial streams are now being conducted throughout the country which indicates this approach is now receiving consideratior. This study was undertaken to determine the effects of limiting concentrations of nitrogen upon the growth patterns and composition of an artificial stream algal community under variable light intensity and current. £515 ,A I. a... METHODS Equipment The recirculating mechanism employed for this project was designed with the following requirements in mind: sim- ple, sturdy, light-weight construction; ample stream bottom for extensive sampling procedures; elimination of contami- nation; high degree of flexibility; an area for efficient bacterial decomposition; maintainence of desired fluid temperature; source of controlled light; and economy. This apparatus is illustrated in figures 1 and 2. A An aluminum trough 21; feet x 11; inches x 10 inches formed the stream bed. This length was a product of Join- ing six sections each )4. feet in length. These were held together by small c-clamps. Flexibility of section con- struction allowed an 8 inch variation in depth at each Junction. Pools or variations in gradient might then be formed as desired. Contamination from the aluminum was prevented by lining the stream with )4. mil, clear polyethyl- ene sheets. The trough was supported by a % inch steel frame of channeling and bent pieces under each joint. Adjustments for the desired depth were located upon the steel struc- ture. Coarse adjustment of 1 inch or more was made by 3.310515 Figure l. w ' ~11: Vofivfi ho og-apn of Artnficicl Stromi S c“. L}; ‘ i .T"; ab‘ 5:121:35 Figure 2. Diagram of Artificial Stream i _ 32> Em " +ccE+mzP< 22L 2 anti .0 u _ +zus+w=$< mwnuov .Z $220 :82st m n m “If .._ 2:3 u u n . Mr +2: :23 mzzoou x we?) 3sz d 1 W n J. Lioov 6 mood $2.; 0 tozomom .H “ESQ .m 53% .I sat Esta .< I 02354 I fiJJ. . Tl.m|l_ 32> 22m moving the horizontal cross beams supporting the stream joint. Fine adjustments, consisting of two adjacent bolts vertically movable within a 1.5 inch.range, were located on each.cross beam. Metal plates were welded to each.bolt head fer stream.support. ‘Water was pumped into the stream from a reservoir lo- cated beneath.the trough by a Model 259.h81 Hmmart sump pump. The plumbing into the stream consisted of 1.5 inch polyethylene pipe containing a valve Which regulated the flow. A centrifugal pump best suited the requirements fer recirculating a relatively high volume of 50 gallons per minute under loW'pressure and a 5 foot head. Any pump, classified as "noncontaminating” and meeting the require- ments above, would.have been expensive and difficult to obtain. The inexpensive sump pump used.met all requirements except contamination, and this was reduced by using a bronze intake. A constant pump discharge was possible by adjust- ment of the inlet valve and maintainence of a relatively even water depth in the reservoir. Originally a wooden baffle had been constructed to decrease turbulence of inflowing water. Since there might have been a possible release of harmful resins from.the wood, this baffle was replaced by one of plexi-glass con! struction. The stream outlet included a 1.5 inch sink fitting and -I‘ J c 1-15 '4. O .f‘,‘ J3 7 drain connected to 1.75 inch polyethylene tubing. Maximum_ filtering surface was obtained by entering this stream out- let into the bottom of a 10 gallon milk can filter. Upon entrance the solution upwelled through rocks, coarse gravel, and aquarium.sand separated by layers of fiberglass. The liquid then spilled over the can lip into the reservoir ' return. This return consisted of galvanized eaves trough lined and covered with polyethylene. This filter was so designed'because maximum drop from outlet to the reservoir 'was only 9 inches, and a trickle type might provide in- sufficient filtering depth. The first Operation of the filter proved ineffective in that the force of upwelling water was so great that amounts of sand from.the filter were carried into the reservoir. This problem was finally solved by placing two layers of fiber glass screen over the filtering material surface. Stiff wire was used to hold the screen edges solidly against the can sides. The inner surface of the filter was painted with Krylon plastic paint to prevent rust and iron contamination. Black paint was used to insure that all parts had received paint. This inert paint was highly insoluble in water and remained intact during the entire experiment. All filter- ing materials were soaked overnight in one percent hydro- chloric acid solution and rinsed'before use. The reservoir was a 3 feet I 3 feet I 14 inches wooden :31: 15.7515 box lined and covered with clear polyethylene sheeting. It had a capacity of 298 liters which was in excess of the circulating fluid. Before each Operation all polyethylene pipes were cleaned of all mineral and algal deposits by circulating a dilute formalin - HCl solution for a period of 2A hours. To control the corrosive action of HCl upon the brass pump intake the pH of the cleaning solution was kept within a range of 6 to 7. The polyethylene sheeting seems to be the most versa- tile component of the stream. It was used to line the stream.bed, reservoir, and return. Only the brass pump intake and valve could contribute heavy metal contamination. Also the lined parts need not be watertight. When an ex- periment has terminated, the polyethylene can be easily removed, discarded, and the system.relined. During preliminary operations of the stream.it was noted.that large amounts of water were being lost via evap- oration. Since plans had.been made to circulate distilled water, this loss, amounting to 38 liters per day, was im- portant because the demand exceeded the water supply. The polyethylene covering of exposed surfaces reduced the loss to h liters per day. Excessive amounts of air contamination also were eliminated by this covering. A cooler, containing a.movable cooling unit of stainé less steel coils, was designed and constructed by the college for wide range temperature control. The cooling unit was immersed into the reservoir and only used in this project to hold temperatures constant at 70°': 20F. The up variability was due to the fact that the temperature control of the cooler Operated within a 2.5°F:range. Res- ervoir temperatures were recorded upon a Taylor Recording Thermometer. Illumination for the stream was provided by a rack of nine 100 watt incandescent bulbs, each with.a 1h inch.shade reflector suspended 1h inches above the stream.bottom. Al- though.a great deal of heat was created by the incandescent 'bulbs, it was considered unimportant in this system.pro- vided with circulation and a cooler. To reduce the number of variables constant illumination was maintained. The artificial stream.was set up for experimentation in the following manner. It was adjusted to contain an 8 feet X,lh inches pool with a maximum.dcpth of 6 inches. This pool was preceded by a riffle area 12 feet X.lu inches .X one inch. The pump was valved to deliver a flow of 25 gallons per minute. This produced a velocity of approxi- :mately 1 foot per second in the riffle. The velocity in the pool was not measured with acceptable accuracy. Veloc- ity was measured.by a Micro Gurly Current Meter. The total stream.fall from origin to outfall was set at 1 inch per 2h feet. This is the equivalent of a stream with a gradi- ent of 18 feet per mile. This represents a fast-flowing 10 stream. Nutrient Medium A nutrient medium B with A5 microelements was selected from several media employed.by Kratz and Meyer (1955) in the culturing of blue-green algae (Table 1). Such a medi- um (plus initial elimination of nitrogen) with.various minor modifications would insure the presence of excessive amounts of all major elements necessary for culture of algae except nitrogen. This view must be taken although excessive quantities might themselves be limiting to algal production. The following modifications of medium B were made. Nitrogen sources, both KNO3 and Ca(N03)2'k320, were removed. Calcium nitrate was then added to the stream.at regular two- week intervals. A total of three additions were made with‘ each addition introducing one milligram.per liter of nitro- gen. Silica, not included in the orginal medium, was added in the form of silica gel. Since the entrance and growth of any organism was of interest in this project, the silica inclusion provided a possible basis for establishment Of diatom.communities. Ethylene diamine tetra-acetic acid (EDTA) was substi- tuted for sodium.citrate. Either of these chelating agents can be used to form.soluble complexes with various insoluble, 11 TABLE 1 COEUEETRATIOLS OF MAJOR SALTS ( g I“), MAJOR ELEMELTS ( mg I“), AKD MICROELEMENTS IN LUERIEHT KSDIUM B CONTAIHIKG A5 MIGROELEHsrTs STOCK SOLUTIOE ILGLUDIEG MODIFICATIurse. Salt Concentration1 Ion Concentration Mgsou'7H20 0.250 Mg 24.3 KHgPOg 1.000 K 287.0 Na2003 _ 0.700 Na 303.8.3 Fe(SOu)3-6H20 0.00h Si 03.6 Na203°Si02% 0.350 Fe 0.h EDTA* ’ 0.050 P 228.0 Microelements A52 1.0 ml CO3 395.2 ------- - - - $04 98.2 1 Concentration in 1 liter distilled water a Stock solution micronutrients ( g I‘): 4 B00 ' “ ’ 2.86 (k7 ) MO702h-MH20 - 0.18% MgClé'hHZO - 1.81 Cd§3_6- - -'- - - - 0.05% ZnSOg'TEZO - 0.22 00(N83)2'6H20 - - - O.u9% 3 Sodium from sodium silicate not accounted for 12 thus unavailable, nutrients and maintain a precipitate-free alkaline medium (Kratz and.Meyer 1955}- The mdcronutrient modifications included the substitu- tions.of (NHu)6Ho702h-I4.HZO for 85% Moo3 as a source of molybdenum, and anhydrous CuSOu fer Cu30u°5H20 . The as micronutrients were also supplemented with cobalt nitrate. Calcium initially was absent from the medium except as an impurity by the exclusion of calcium nitrate; but each addition of the nitrate salt introduced elemental calcium in a molecular concentration higher than that of the nitro- gen, i.e. 25 mgl" Ca(N03)2.4H20 contained 17.2 and 3.0 rag-1" of calcium and nitrOgen respectively. In general, calcium is required in minor amounts by organisms. Most nutrient media contain only traces except when the nitrogen source is calcium nitrate; a few such as that used by Warburg and Burk (1950) contain none. Bold (l9h2) reports that calcium is unnecessary for certain algae such as Chlorella. Allison, Hoover, and.Morris (1937) found calcium to be essential for nitrogen-fixation by Nostoc musorum but unimportant for growth. Media with higher magnesium.content Often reduces any requirement (Chu 1942). Still others found any limiting concentration to be far below that of nitrOgen (Gerloff, Fitzgerald, Skoog 1950, Kratz and Meyer 1955). All nutrients were placed into the stream.channel source as a solution or suspension. his prevented settling out in the reservior and facilitated solution of undissolved. 13 salts. Water Chemistry A sampling program set up for this project included analysis of alkalinity, pH, and conductivity at weekly intervals and analysis of total available nitrogen, ammonia nitrogen, and total phosphorus at two-day intervals. Total available nitrogen was also deterrincd immediately before and one hour after each addition of calcium nitrate. All samples were collected from the pool zone of the stream. alkalinity Phenolphthalcin and methyl orang, alkalinity were determined by titration methods described in Nelch (1948). Results were expressed in milligrams per liter of calcium carbonate. 1%; HydrOgen ion concentration was determined on a Beck- man Model H pH Meter. conductivity Electrical resistance was measured with an Industrial Instrument Company Model RC-7 portable conductivity meter. All readings were corrected to 18°C and expressed in units Of specific conductance as micromohs per centimeter (Industrial Instrument§_9perating_Manual). hardness Values of hardness in milligrams per liter were 14 obtained by the versonate method (Catalog No. 4, Each Chemical Company). Determinations were made only before EDTA had been added to the stream as EDTA was the titrant used in the versonatc method. silica Three determinations of silica were made: at the begin- ning, middle, and end of the experiment. A gravimetric method taken from the Chemical Laboratory Manual of the American Cast Iron Pipe Company, Birmingham, Alabama, was employed. Results were expressed in milligrams per liter. sodium and potassium Concentrations of these cations in milligrams per liter were determined from samples removed for silica deter- minations. Values were obtained from a Perkin-Elmer Flame Photometer. total phosphorus Values of total phosphorus were resolved by a color- imetric method described by King (1932). A Beckman Model B SpectrOphotometer at wavelength 860 mu was used in the pro- cedure. Results were obtained in milligrams per liter. total available nitrogen Total available nitrogen included all inorganic forms except atmospheric nitrogen. Determinations were made by using the reduction method described in Standard.Methods for the Ekamination of Water, Sewage, and InduStrifll Wastes (APHA, AWWA, FSIWA, 1955). These determinations were made .‘r ’.|. i"; . J b L‘ufld" 15 immediately after collecting the sample to prevent loss of ammonia. All results are eXpressed in milligrams per liter. ammonia nitrogen Ammonia nitrogen in.milligrams per liter was measured by the distillation method described in "Standard Methods" (APHA, AWWA, FSIWA, 1955). Determinations were begun on April 25. Periphyton Analysis samplingpprocedure Artificial plexi-glass substrates 7 mm thick with an exposed area of 150 cm2 were employed to sample the commun- ity of periphyton. These plates were held stationary in the stream by plastic coated wire racks. A total of Sh shingles received use in this project (Fig. 3a). The riffle and pool areas each contained 2h, and.the remaining 6 were placed into an unlighted zone preceding the riffle. The 2h shingles per area were sub- divided into 8 sets of 3 shingles each, 7 of which were incorporated into a 12-day overlap sampling system with one set remaining to be used as an extra. The 12-day sampling system.was devised to obtain an approach to the measurement of instantaneous primary pro- duction. To help visualize this procedure figure 3b is provided with the S two-week nitrOgen periods labled. Each of the 7 shingle sets was exposed to two weeks of stream "HESIT Figure 3. Diagrams of Substrate Arrargoment in tificial Stream (5-) azd Dates Sampled by Bach Shingle Set (b) ”oz 1.»? £3: :- 2 Q r m . 5. we 2 m. 2 h n on 71H- q . H _ i rah-+- : 1.2.1.3-}: __.i 1". garlil. - . — HIM-""I.III[1.II # 41 lLr‘- q u _ .1:gl'll1 I J I. fi 43 ...- ”H. - - - a. .u 1 1 III'I-lr'lu L5 I. -. tutti-.- .6 .1313... _ '4. L7 T voted m. voted NtoIoQ «to-:51 . .m a a If s r m in. a a a m n.- .w In L Hm I. 5 5 O u R w A W m gunman-hm qwmwhm mmw izsmflm mIMMwmm .m-M-w. _ a w . fl , . J. a L mew. mmman zasmm JIMMQMQ aim IE .- m in i x a 5mm - 1. 2 3 u: 5 6 7 6 an m. «U. n 2t3~0 U )3 56+ Number .‘ ~..'i J} 1.5-4.- 17 conditions, but overlapped in a.manner which allowed each set to sample two days longer than the preceding set. Before the first addition of nitrogen on April 12, two weeks were required to set up the Operation; set one being added 13 days before adding nitrogen; set two 11 days before; ... set seven one day before. Therefore, set one was removed after one day of eXposure to nitrogen, set two after three days, ... and set seven after 13 days. This process was kept in motion throughout the project by removal and re- placement of the designated set of shingles every two days from both riffle ani pool zones. All light shingles wqre collected after eXposure to periods 2 through 5. After the substrate set had been removed from the stream, shingles I and II were analyzed for phytOpigment concentration and organic nitrogen content. Each shingle was divided laterally with organic nitrogen and chlorOphyll being determined from the upstream.and downstream halves respectively. Shingle III was reserved as an extra. phytopigmont_ The growth on the entire surface including bottom.and sides was removed to obtain maximum.material in periods of low production. This material was scraped and washed with 95 percent ethanol into 250 ml beakers. Remaining shingle halves were stored in the freezer. Complete chlorOphyll extraction was insured by soaking the material for a period exceeding 2h hours in complete darkness. Tests indicate J} {#90 18 that samples can be stored in this manner for as long as 30 days without a loss of phytOpigment due to decomposition (Brehmer, PhD Thesis). Alchol was used as a solvent in preference to acetone since the latter dissolves plexi- glass. After soaking, the samples were filtered through glass wool, and the extract volume was adjusted to 50 ml by dilu- tion or evaporation. PhytOpigment concentration was deter- mined in a Klett-Summerson colorimeter using a 6h0-700 mu red filter. Brehmer and Grzenda (in press) have shown that the absorbency (6h0-7OO mu) of 95 percent ethanol pigment extracts is not linerally related to the concentration except at low values. Hence, they have provided a graph (Fig. A) for converting the measured absorbency into the theoretical absorbancy which follows the Lambert-Boer Law. The necessary corrections Were made by locating measured absorbency on the ordinate and following this value to a point of interception on the experimental line. A vertical extension of this point will intercept the lembert-Beer line, and the absorbancy unit directly Opposite this inter- cept represents the corrected absorbancy reading. The unit of adjusted absorbency is termed (AA) and is multiplied by 103 to avoid use of the decimal. Shingles in the area of diminished light were measured only for their chlorOphyll content. For comparison purposes Figure h. Correction Graph for Adjusting Measured PhytOpigment Absorbancy Values to Units Related to Con- centration 5.0 h 8‘ CLOL- 0 Z ‘0 I— S r. .5?— U) .0 r- $1.93” <32 __ k. ,g 1 l llLilll J LJIJIII I IO '00 Re/afi ve Concen+ra+ion 20 half of each shingle was analyzed. organic nitrOgen The remaining shingle halves, which had been frozen after chlorophyll determinations, were removed, thawed, scraped, and the aufwuchs was washed into 250 ml beakers with distilled water. Each was analyzed by the s;mi-micro Kjeldahl procedure described in "Standard Methods" (APHA, ANNA, FSIWA 1955). Results are expressed in milligrams organic nitrogen per 75 cm; (area of half’shingle). total gry_weight An estimate of the total dry weight of periphyton which accumulated within two weeks was obtained by removing the growth from a 37.5 cm2 section of shingle III. These deter- minations were made after completion of the project and only certain shingles were available at that time. The chlorOphyll was extracted upon removal as described in the section on PhytOpigment, except that gooch crucibles were used to collect the algal residue. This was necessary for weight ans ysis. The residue was dried overnight in an oven at 55°C then placed into a dessicator to cool. Succes- sive weight measurements wure conducted upon an analytical balance until a constant weight of;:’.5 mg was reached. The total dry weight values will tend to underestimate the actual values since chlorOphyll and lipids were removed with the filtrate. 21 Light Measurements of light intensity were obta'ned by using a PR-l General Electric EXposure Meter. This instrument was calibrated by direct comparison with an Eppley pyr- heliometer maintained at the Michigan Hydrologic Research Station on the Michigan State University campus. The cali- bration was conducted on a clear summer afternoon from 1:00 PM until sundown. Light meter readings were taken at 15 minute intervals. 1 The intensities recorded by the pyrheliometer were converted to gram-calories per square centimeter per minute by dividing the direct measurement which was recorded in millivolts by a constant, 1.71. A straight line relation- ship was obtained by plotting the exposure meter readings against gm-cal cr.r1.""rn.in'I on semi-log scale. The line of best fit was adjusted by eye and extrapolated to give an estimate of the energy received at lower light intensities. Two lines are indicated which represent the adjustment of the eXposure to read at high and low light irtensities. Correction Grapn for Converting EXposure Meter Readings to Gram- calories per Square Certimeter per Minute 3:5: Loo. 535.450 23cm. Lem mo..»o,ou\Ees0 5 0 I I m m. w a s m » ddAJfiqd A u .d‘——_ d — fl _—fiq__ II A L Low J3 l4” i5 J6 l7 ExPosurc Meier Readings RESULTS Species Composition The first attempt to establish an algal community in the artificial stream was successful using tap water en— riched with small amounts of water from both the Red Cedar River and the fish tanks in the laboratory. A few stones from the Red Cedar, well encrusted with aufwuchs, were introduced to seed the system. Diatoms were the dominant organism in this material with Navicula and Gomphonema as the major species. The stream was in continuous Operation two weeks before new growth appeared on the seed rocks and stream.bottom. This growth, primarily Chroococcus, first appeared in the pool. During the third week a commercial fertilizer rated 17-17—17 (NH3-P205-K) was added to the reservoir. Within three days a dense algal bloom of Chlorella_and Navicula had begun. Navicula seemed to dominate the riffle area while Chlorella was more abundant in the pool. This phase of the project indicated that a reproducing algal community could be established under atypical lotic conditions. On March 28 the quantitative experimental program described in "Methods" was begun. Seed.material scraped from.Red Cedar River stones was added on March 29. 2h The first green cells appeared in the pool on April 15, three days after the first addition of nitrOgen. No growth was noted in period 1. The pioneer community was essential- ly composed of diatoms although representatives of green and blue-green algae were present. From.April 15 to 25 unidentified unicellular blue-greens and diatoms dominated the pool area while a lesser pepula- tion of diatoms existed in the riffle. The diatom Navicula was most abundant. Filaments of Stigeoclonium.and Ulothrix also entered both zones about April 19 and remained until early May. After April 25 a major community change took place as colonies of Anabaena oscillarioides (identified by Dr. Francis Drouet) appeared. Thereafter this species of blue-green algae completely dominated the habitat, and eventually it formed a spongy mat of cells about one-fourth of an inch thick. By Jamao the mat had begun to break loose from the stream bottom. This was most likely due to death of the cells adjacent to the bottom and formation of gas bubbles under the material. Moreover, the community appear- ed to be senile; but production, described in later sec- tions, was still great and was visually evident by the rep0pulation of areas left bare after sloughing. Initially it seemed that Anabaena succession had elmin- ated other organisms, but careful examination found a large community of Navicula living within the blue-green mat. 25 Other species of algae such as Stichococcus bacillaris and Schizochalamys_§eliafiuh.were minor constituients of the community for a short time after April 25. About me312 filaments of Stigeoclonium again entered the area at the pool head. Throughout most of the experiment there was little evidence of an invertebrate community. Hewever, toward the end a pOpulation of midges was beginning to develop. Water Chemistry alkalinity Weekly determinations indicated that high levels of both phenolphthalein and methyl orange alkalinities were maintained throughout the experiment (Fig. 6). Initial high values were expected since the prOportions of mono- basic potassium.phosphate and sodium carbonated used buffer- ed medium B in the alkaline range (Kratz and Meyer 1955). It was also calculated that approximately hOO mgl“ of carbonate ion would exist in the system upon complete solu- tion of sodium carbonate (Table 1). Conversion of this ion to bicarbonate would be enhanced by eXposure of carbonate ion to carbon dioxide in the air resulting from the turbu- lent conditions and shallow depths of this stream. With the accrual of algal cells the concentrations of hho mgl“ bicarbonate ion and 96 mgl"monocarbonate ion were ‘reduced to lows of 257 and SO mgl“, respectively. A gradual [U 0‘\ TABLE 2 WA 1‘ 3i CE-iEK-i IS‘I‘RY "- Alkalinity (mg 1) Resistance Conductivity” Date pH Phth . 1'1. 0 . (ohms) (mi cromho s) h-é 8.h 75 375 1150 1610 u-a 8.h 96 LAO - - - - - u-16 8.a 80 386 ‘ 1300 1a25 a-23 8.5 80 370 1320 1u03 4-25 8.? 75 368 1380 1342 5-6 8.7 77 366 1280 Ian? 5-9 8.9 68 290 1630 1136 5-17 8.9 72 261 1670 1109 5-22 8.9 66 257 1950 950 5-29 “.7 56 26a 2000 926 6-6 8.8 58 278 1950 950 6-12 8.6 50 290 - - - - - Figure 6. Alkalinity Artificia h 1 we ,‘ b\-& Stream .56 9i E? 3 r r on he 3.. .9 2 h on m... 8.. h. 2 o q l a T q a a _ W q d d n .1 9w or f L 5 U8 .------.-.---: I x .8 n H 32 L m6 0 m A r a: z .1 it c x . . H. ...... «531?. ct L 8e ..... stiff Law a A W “10?.“ 190 m I I , my Sea , , J 5 .Mu 3m -Qo Sn 1mm 8: .. 12. I, a. 2... ; 1:6 28 rise in bicarbonate which occurred after May 23 plus con- tinued.monocarbonate reduction might indicate that the latter was being converted to the former by presence of increased carbon dioxide from organic decomposition. 'pg The hydrOgen ion concentration increased from a value of 8.u at the beginning of the project to 8.9, which re- mained for some time after the third addition of nitrogen. A decrease again occurred toward the project termination (Fig. 6). The slight fluctuation in pH seemed to follow produc- tion levels and possible nitrate assimilation. On May 29 production drOpped in both riffle and pool areas after a general increase prior to this date (Fig. 11). A produc- tion rise again occurred after this date to correspond with pH rise. In certain vessel cultures the assimilation of nitrate by growing cells was accompanied by an increase in pH (Rodhe 19118, Kratz and Meyer 1955). It is evident from.figure 6 that pH was inversely related to the bicarbonate alkalinity. This effect could be a product of many factors since alkalinity results from.the solution of NazCOB, and pH results from solution of both buffer salts plus the presence of sodium.silicate which was not included in the medium.used by Kratz and Meyer. Sodium silicate and sodium carbonate are often combined to main- tain an alkaline buffering capacity (Gerloff, Fitzgerald, 29 koog 1950). The pH and alkalinity, no doubt, have played an impor- .tant role in determining the presence of Anabaena and Navicula. Using a modified Chu 10 medium Gerloff, Fitzgerald, and Skoog (1952) noted optimum growth of Migra- systis aeruginosa was at pH 10. Cells of Anabaena vari- abilis experienced maximum growth at pH 6.9 to 9.0 (Kratz and Meyer 1955). Bold (1942) cites Gietler as finding it necessary to grow Navicula on agar of pH 9 in order to secure formation of aUXOSpores. conductivity_ Values of specific conductance, plotted in figure 7, implied that a constant decrease in total ionized consti- tuents in the water took place. As echcted, diminution followed aufwuchs increase. It is interesting to note that conductivity variations followed closely to those of bi- carbonate alkalinity. The alterations in alkalinity, pH, conductivity, and other specific ion concentrations were in part due to an irreparable leak which deve10ped at the filter lip on or about May h. The exact date it began was unknown but was estimated from observations of standing water under the filter. Corrections for this loss were made by collecting and measuring the liquid from the leak for a period of three weeks. The determined average leakage rate indicated that approximately 38 liters of solution would have been ")1; Figure 7. Specific Cordrctance of Artificial Stream Water L 8: L 29 l 8: l 83 1 c2... soqmoaogw 31 lost in 20 days--a length of time known to exceed the actual days of leakage. hardness Calcium.hardness (EDTA) was determined on April 8 to confirm.the fact that alkalinity was not due to calcium carbonate. Results indicated that calcium was present in amounts too small to be detected. On April 13 total hard- ness (EDTA) was measured twice and found to be 88 mg-”. Since the total hardness is lower than the alkalinity, it appeared that most hardness was due to magnesium carbonate. The calcium.hardness was rechecked on April 13, but results remained negative. total‘phosphorus The phosphorus levels were extremely high throughout the experiment, ranging from 100 to 170 mgl“, with only one exception which occurred on April 15 (Fig. 8). An even greater concentration of phosphorus would have occurred if complete solution of KHzPOA had taken place (Table 1). Therefore, the gradual rise from the minimal to maximal concentration from April 11 to 21 probably was due to in- creased disintegration and mixing of the undissolved salt. After April 21 it appeared that plant production began gradually to reduce the phosphate level to a low on May 25. Throughout this reduction several pulses of increase were observed. These pulses might stem from liquification of undissolved phosphate salt or regeneration and recycling of 32 TABLE 3 NITROGEK AKD PHOSPHORUS DETERNILATIOES IN MILLIGRAMS PER LITER .A Total Available Total Nitrogen Ammonia Phosphorus .043“ - - - - - - .130; - - - 100 .BAO - - - - - - 4-13 .310 - - - 120 u-lg .370 - - - 60 4-17 .570 — - - 1A0 h-l9 .260 - - - 160 h—Zl .015 - - - 170 u-23 .090 - - - 16h L-25 .OLO, .000 150 4-26 .080“ - - - - - - .250O - - — - - h—27 .L20 .000 150 h-29 .085 .000 170 5-1 .090 .000 150 5-3 011.10 .015 m0 5-5 .330 .060 160 5-6 .090 - - - - - - 5-7 .105 025 120 5-9 .150“ .005 120 5 1 .210" — - - - - - .970O - - - — - - 5-11 .180 .025 120 5-15 .120 .120 120 5-17 .150 .000 125 5-19 .160 .020 118 5-21 .110 .015 1&2 5-23 .105 .000 110 5-27 .100 .050 10h 5-29 .140 .000 100 5-31 .175 .000 120 6-2 .110 .000 122 6-6 .070 .000 120 6-8 .300 .090 130 6-10 .185 .030 136 6-12 .110 .080 126 6-lu .110 .060 l2u * Sample taken immediately before nitrogen addition 0 Sample taken one hour after nitrogen addition Figure 8. Total Phosphorus in Artificial Stream Water 8 o w 2 cmoquoqd 491,11 494 sw045mgw r I00 7— 501 I5 20 25 May IO '0 I5 20 2.5 30 IO Apr? | 31+ this nutrient in the plant material. Total phOSphorus levels began to rise again after May 25. This seemed to coincide with the sloughing of large algal fragments from the aged mat. If bacterial de- composition took place in the filter, phosphorus in organic form would be recirculated through the system. Losses of phosphorus due to leakage were estimated to be approximately 16 mgl't As was mentioned earlier, certain elements in the arti- ficial stream might be detrimental to growth as a result of their enormous concentrations. PhOSphorus is one such element; and since it plays a vital role in plant nutrition, the large stream supply needs consideration. Chu (1942) noted that strong phosphorus concentrations inhibited growth of certain diatoms and green algae, but it varied with the species. Later he found less inhibition when nitrate- nitrogen was used as a nitrogen supply (Chu l9h3). Osterlind (19h?) mentions that use of phosphate buffers in high concentrations are often injurious to algae. In direct conflict with these findings Kratz and Meyer (1955) varied KZHPOh from .25 to 1.5 g1“ without effect on growth of Anabaena, Anacystis, and Nostoc. Other nutrient media used by Ketchum, Lillic‘x, and Redfield (19m); Warburg and Burk (1950); and Harris (l9hl) to grow many algal species also contained phosphate in similar or even higher amounts than those occurring in the artificial stream. Views on this 35 subject conflict, but it seems much depends on tolerance ranges of the particular algae species and the phosphate concentration range with which one works. silica The depletion of silica was of interest since the ubiquitous diatoms remained throughout this study. Data from.three determinations (Table A) indicated that only about one half of the added silica gel went into solution (Table 1). This provided an initial concentration of 3h mg l“, which according to hrauss (1958) is about Optimum for dense cultures of Navicula. Certain greens and diatoms exhibit an Optimum growth in nutrient solutions containing 30 mg 1“ of silicon dioxide (Chu lQMZ). The concentration of silica was reduced 6 mg 1” during the experimental period, although leakage accounted for about two-thirds of this loss. Moreover, the probable solu- tion of salts, yet undissolved, complicated the picture. potassium and sodium The relative concentrations of 58.7 mg K 1“ and hh.8 mg Na 1“ gave evidence that monobasic potassium phosphate was more soluble than sodium carbonate at the beginning of the experiment (Table A). In table 1 it should be noted that complete solution of the two salts would insure a high pH. In view of this, pH should have been much lower than it actually was. The actual high pH value was probably a product of sodium silicate. TABLE u POTASSIUM, SODIUM, AND SILICA ION CONCENTRATIONS AND CORRECTED TOTAL REDUCTION OF EACH IN MILLIGRAMS PER LITER Date Potassium Sodium. Silica u-é 58.3 u5.o - - - a-e 58.8 . Qh-S 3h.0 5-5 35.0 35.8 32.0 6-16 31.3 25.0 28.0 Concentration ‘ Reduction 27.2 19.8 6.0 Leakage Loss h.75 4.9 u.3 Corrected 22.h5 lh.9 1.? Reduction Figure 9. Potassium, Sodium, and Silica Ion Concentrations of the Artificial Stream £0.34 Lem mEULmzzz o 0 O 0 r0 5 u.- 3 H «10 ---IO J— _. J Potassium ”'1 Silica Sodium June 16 F...... .............. Y. “ ........... . .. . V////////////////////////////A May 5 //w/ / / A Apr]! 6 38 As plant growth increased both potassium 1nd sodium were signi 'ficantly reduced (Fig. 9). When compared to silica their reSpective losses due to lea.tage were minor. The total corrected reduction of potassium was the 1 but analysis lflldlc tc d that sodium depletion was relatively gr reater when tin community was dominat d by Aaabaena osciHarioides. Potassium ion is found almost universally as the principle inorg anic cation of cells, whereas the sodium.ion is known to be dispensable for most plants with the exception of blue-green algae (Pruton ani DLHJOJdo 1959). It is necessary to point out that few conclusions can be drawn about these ioaic reductions when quantities of un- dissolved salts in the system provided a continuous source for nutrient replenishment. total available nitrggen The concen ruti01:.s of total available nitro en in the ar-;ficial s ream are illustr ted in figure 10. Successive alcium nitrate ::dditio 8, each introducing 1 mg I“ of nitrOgen, caused an immediate rise in the total available nitrOgon. Tiese high initial levels we re sigiificantly reduced by the growiz1g algal community. tion peaks in figure 11 followed each initial peak of in- organic nitrOQCn. It also should be noted that the fall of initial levels was more rapid as the st andin " crOp deVeIOped. A total of nine days pa seed before the April 12 nitrOgen supply fell to trace amelxxts (April 21), whereas the Figure 10. Total Available Kitrogen and Ammonia Nitrosen in the Artificial Strean o m :3: E < 59:5 «323% .3.» h float—Um —'——-——---—-—u_*----. r .2an 591:2.-. E“ as 1. ..2.< _ . _ . _ . m totem __ . . _ 5032...?” dofizu?‘ _ _ L _ _ I. _ _ I _ _ .... N motel _ — l _ . L _ :umogtz .-. m5.“ L fio..£v1< 0 z~ W 6 6 0 491.5“ Jed $044045ng ‘7 o 9 no period.u supplement on May 10 dropped from .,7 m3 14 to .18 mg 1“ in one day. A strict two day sampling would have missed this rise and fall. After the new sultplies of nitro3 n hzd been depleted the "normal" stream concentration remained in the vicinity *3. of .1 m3 1“. In periods a and 5 this is re icularly notic- able. Many authors working on the subject of nitrogen as a limiting factor and using many veri tics of 3135c found tiat the lower limit of this element for optimum growth occurs well above the value of .l me I4 (Gerloff, Fitzgerald, and Skoog 1950; Rodhe l€h83 Chu 19h3)o Moreover, Brehmer (PhD thesis) has interpreted nitrogen to be the limiting factor of the Red Cedar River where mean values ofi or3anic nitr03en are .7 m3 lf'above the sew 3e outfall. Ri - (lQLO) also indie: ted that the plankton of Linsley Pond becomes dominate d oy diatoms and blue-green alfiae in the summer . I“ ._ .1 ' | rzr (1'38 beta-Den .el and .01.}.1'13 l‘. montr1s wh en the nitrs -t- 0 These oeulations :nl concentrations comps e somewhat to 1 } those of the artifi ial stream. In conjunction with low levels of nitrogen the presence of Anabaena and diatoms in the artificial stream gives further evidence that nitrogen was the limiting factor. It was noted arrli er that the salts of the nutrient med- ium, did not dissolve completely. Calcium.nitrate was no I'D xc egtioz: to the rv 9. Less than 35 percent of the perioc (.1 dd Id. 5‘.) tion went into solution; and when al3a13rowth vegan to 41 affect a reduction on April 17, only 55 percent had been dissolved. Since these undissolved salts might always be present, the nutrient could be immediately assimilated upon release into solution. The solution of undissolved salts plus organic breakdown, nitrogen fixation, and release from livin3 plants mi31t also account for the fact that some nitrogen was present at all times in the stream. £333 ammonia Free ammonia determinations were begun after some growth had accumulated in hepes that pulses of this product would be indicative of organic decomposition by hetero- trOphic bacteria. The values ranged between .12 and O milligrams per liter for the entir' period (Fig. 10), and ( it appeared there might be a tendency for the concentration to rise toward the end of the CXperiment. Plant material was breakina 10032 and being washed into the filter at (I) this t ime . 4.:- [U Periphyton Analysis phytopigment Relative rates of primary production were obtaiaed by comparison of the phytopigment concentration per talf shingle. Theznea11 ph.yto1i3ment u;1its ara plotted versus time in figure 11. The first detectable shingle growth occurred six days after the first nitrogen had been added. He growth was noted in period 1. In view of these facts the introduction of nitrogen became a lechanism for triggering the reproduc- tion of cells which had lain dormant for almost three weef.‘ -s. Figurrg s 11 and 12 illus crate the t successive additions of calcium nitrate brou3ht about an increase in the relative production rates in all lighted areas of the artificial stream, Ni thin periods 2, 3, and h pigment concentration dr0pped after the ass 1nilation of: low nitrate (Fi3 . 10). This drOp mi 3ht be an actual decrease of "31 nt. Yentsch and Vaccaro (1958) report that nitrogen deficiency produces a decrease in chlorOphyll which may ee attributed to the decompositio:1 of the pigmzm tprotein complex Anotm1 r possible rose 01 for reduction of chlorophyll might merely be a result of the physical condition of the shingle. For exampl:, the surface of a two-ween substrate rem dOV d on /.pril lS could have been alt,rcd 3y the environ- ment to accep adhering cells. Cells throu3hout the stream at this time would be entering a period of rapid division, #3 o.mm mmpo mac coo o.¢m momfl moo o:o o o co” cod owm mate omo Hmum ooa em-: :.ms coma ems owe ~.H~ om: mmo gas m.a ms :1” .mofl owe oa-o oao oa-m omfl pm-: m.nm :sm mm: mom n.m~ com omo owe H.»H mam mm 0: me: wuo one ennm 0H mmtd H.~ om moo ooo m.mm com omm omm :.H m em mm oao 0-0 am ma-m om mun: 0.59 mmmo Hoe «mm m.:~ ooom 4m: om :.H m em oo 05 4-0 0: maum we Hm-: o 0 mm: mm m.e 0: mo: ma: w.:~ omm an mm mm m-e no: ”Hum ._:: oau: m.:ma enema so: me o.~m mmofl ”ma ma :.a m m N we antm om mum sa-: 0.9m 04m mmm an m.mm QJQ mom omm c.m m 0H m men mm-m sea ~-m ma man: m.m m oom son m.om mme pom wma ~.o so ma mm mom emnm omm m-m ma mfl-: mc.om Nymm om: mm s.»oa oomwm 04m omH o o o o w mm-m com m-m 0 «an: H.oo maem mm: oo o.m : NHH maa o o o 0 ma: mm-m maa ~-m o o”.: .>on .>on .>on .uum .aa> nae: moax<< opan .apm .u¢> nuns. moax<< opaa .apm .nas cues. moax<< open mmsmomxm mxmflg 338 mmam< m memes «wzomev oQ oon .35 .5; can: model .25 .35 38> :33 mod“: 33 .35 33> 532 mode}. 33 mmnwomfi 933;. 25H. mmamd <33. 8H2: mmm Noammona BZHEGHmOBMmm doam o mqmde Figure 11. Mean PhytOpigment Absorbancy Units per Unit Area After Two Weeks EXpo- sure. Arrows Indicate Bi-weekly NitroSen Additions mush xvi zit. dimwfismwr‘\.>me.~hd§tfl 4 J q lip-1! A _4 q _ 4 7 /.\ 1 . _ 1. _ . IIII. «SE null: .ool _ . _ \\’ _ \~ ”I _ x x _ x a. x . K V a) I \\\ /\ a . >’ a . s . 5, . < _ e .31.“: £232.19; E2121?“ i is? 52.2 853 u _ _ 1 .. _ . _ . _ _ u _ u . _ _ m‘uomgvl _ T V0201 “ m. .3781 _ N totem _. . r § 0 § :2 g S 90' x KauquOqu J § 33 ohm. 8% 16 thus reducing the available nitrogen content. At this time the shin3les which were to be removed on April 25 were not cone.itioned for growt oh and only became condit1011:3d durin3 the reduced nitrogen concentrations, therefore loweri113 the cumuhtive cell production to that date.. A two-way analysis of the varia: ce we used to ana— lyze the variability of phytOpi3ment concentrations between riffle and pool zones (location) anl between periods 2, 3, h, and S. The "F" values obtained from this test show that the; e was a si3r ificant difference between location, between periods, and the inte action between location and periods at the one percent level (Table 7). The relative production differences between the riffle and pool zones can be narrow- ed to one of three factors or the int:ractiens between the since other Variables are assumed constant. These three factors were variations in water velocity, variable light intensity striking the str cuim bottom, and c011pe Ht1on effects. Although all lights were placed eqin Md stant from1 the bottom, the variations in li3 ht intens itv are mentioned because light waves must pass through different depths of water. The effects of competition seem very plausible wiea considerin13 the differences between locations and between nitrogen periods. Briefly, seed material settles out in the pool and begins to grow. Riffle areas lag in cell establishment due to the phvsical effects of the current. TABLE 7 #7 ANALYSIS OF VARIAECE OF RIFFLE AND POOL AREAS FOR FOUR PERIODS OF SHINGLE EXPOSURE Phytopigment Sum of Decrees of Variation Squares Friedom Mean Square " F " Within 257,135 #5 5,357 - - - Cells l,696,h78 - - — - - - - - Location 46,633 1 Q6,633 8.705 Period 1,538,863 3 51,288 9.57M Interaction 110,982 3 36,99u 6.906 Total 1,953,613 - - - - - — - - Organic Nitrogen Variation Sum.of Degrees of Mean Square " F " Squares Freedom Within 5.51uo 18 .1119 - - - Cells 11.5851 - - - - - - - - Location 0.63h3 l .63u3 5.521 Period 9.7683 3 3.2561 28.339 Interaction 1.1825 3 .3912 3.131 Total 17.0991 - - - - - - - - TABLE 8 MEAN PHYTOPIGMERT ABSORBANCY UNITS AND NILLIGRAMS ORGANIC NITROGEN PER PERIOD Mean Mean Period PhytOpigment Organic Nitr03en Riffle * Pool Riffle Pool 1 33 73 0.02 0.11 2 l7u 196 0.27 0.31 3 5M3 402 1.17 0.77 a Q68 316 1.32 0.73 *q 2% t u 3 I"‘ r \l 12. 1‘; p U AA ru (.4- I ’1 *- n PhytOpigment Absorbancy ts per Period m9 x hogan. .3931 m m ~500 -— 2.00 D 4100 RENE: Pool 32%;; [... Period 50 Figure 11 shows that growth began earlier in the pool; and the histogram in figure 12, which contains period averages, illustrates that growth Was greater in the pool during ini- tial colonization. Eventually the riffle growth became equal to and ex- ceeded that of the pool because this area had the first Opportunity to extract nutrients coming from the reservoir. It is also interesting to note that pool pigment concentra- tion peaks preceded peaks in the riffle until the middle of period 3 (Fifi. 11). The standard deviations of two phytOpigmeat samples removed on the szme date are listed in tables 5 and 6 along with mean and variances. Pigment variability between these might be a result of sloughing algal cells H adjacent shingles at certain times, nigaly variab 3 growth on the shingle bottoms, and initial colonial growth of Anabaena. Single large Anabaena.colonics were noted on one of the tWO April 27 pool and May 3 riffle samples. It should be mentioned that variable growth on the shingle bottoms might have been a product of upper surface shading and/or adherence of cells which were torn loose from the stream mat. organic nitrogen The mean values of organic nitrogen in milligrams per half shingle (each exposed two weeks) are plotted versus the time exposed in figure 13. In conjunction with pigment production an increase in organic nitrogen followed each 51 aamm. ooaa. no. mma.a demo. :aoo. am. omw. . mo. amum mam. om-: smzo. mace. om. mmm. moaa. mmao. he. om. same. @000. mm. mom. mom. oa-o ode. oa-m now. am.: aao.a ooa.a ma.a mom. m Noam. ssdo. mm. 00». o 0 ea. -.- . w-o we». aaum «ma. mm-: masm. coma. om. 0mm. smos. owes. ma.a mom. mooo. once. on. omm. mam. o-o mom.a ma-m mmw. nm-: ammo. omoo. mm. mom. smwm. ease. co. mas. o o :o. u - mom. :-o mm». ma-m mmo. amus moo.a omo.a we. moo. a amao. aooo. co. m m. o o :m. - . mom. «no m o. aaum cam. man: maao. aooo. am. a mma. a sosa. omoo. mm. mm . o o . mo. - - mma.a amum co . o-m oao. aa-: demo. :aoo. em. com. . omna. :mao. on. new. - u - - - . oao. mom. om-m mm:. sum oao. ma-: mmso. amoo. mm. omm. maao. aooo. :m. wmm. - u - - - . oao. m :. am-m mam. m-m oao. ma-: amaa. adao. am. as». smso. maoo. mm. 0mm. - - u - u - oao. mam. mmnm oam. mnm oao. man: mafia. oomo.. ms. mm sane. :aoo. :a. mma. . u - - u . oao. $m nmnm moa. aum oao. can: .>oQ .>en .>00 .35 .3; :3: 23. 3.5 .Sm .8; 5oz 2% 33 8pm .5; 3...: awe 35 oa Roma 4 ...»m 23> can: z ma oven .dpm 33> use: a m: open .aum 23> 502 2 ma open <92 mthHm mam. 2H OOmeE mammomxfl Mum—3 038 < mmamca 5mm: 9H2: mam 2mmoxBHz UHzmmmo k0 msamqugHS 0a ”mamas“ Figure 13. Kean Organic Litr03en per Unit Area After Two weeks EXposure. Arrows Indicate Bi-week y Eitrogen Additions 055. he: €an m. x. 5 ma. ma. 2 .3 t a. n ea. bu Q t E r a a a W 1 a l 1 d 4 . \ _ ( ibfi u 16h. 1“”. Led :13...“ 595219: H a.omc£z..:wEfi SE22 calla? #233 :34 _ _ _ — o _ _ . aloha _ _ _ _ in: _ _ _ u h meted " r .3761 u 0.3th _ a point _ .N _ _ _ lllll .oon. new octm HIEW 11950.!“ N am 045 Figure 1h. Mean Milligrams of Organic Iitrogen yer Period a [.50 "‘ [-2.0 cmmfitz mEeem...:/>\ m ,0» m .1 fl ". g///////’///17/// , Poo! Riflle E] .V///////////////////¢///Al . 55 calcium.nitrate addition. Enormous fluctuations were also noted in periods a and 5. Figure 1h, which consists of values averaged for periods 1, 2, 3, and h, illustrates the most prOfOIfld in- crement of organic nitrOgen, which occured after period 2. The averages of periods A and 5 indicate that cellular nitrogen content may have reached a leveling off point, but this is only an assumption as enormous fluctuations occurred in these latter periods (Fig. 13). The point of change, just discussed, roughly corresponds to the entrance of Anabaena which later dominated the awfwuchs community. A two-way analysis of the variance indicated that there were significant differences between the location means and between th interaction of location and periods at the 5 percent level (T3510 7). Means between nitrogen periods are significant at the one percent level. The levels of significance show that the differences between periods were greater than differences between locations. The five per- cent level for location also indicates that differences between mean values of organic nitrogen were not as great as those between mean phytOpigment units in the riffle and pool areas. As with pigment the difference in riffle and pool areas was probably a product of current, light, and competition. Note that the trend again follows that of pthOpigment in that the early concentrations of organic nitrogen, hichest in the pool, were eventually overtaken 56 and succeeded ty a hi her co centr*tion in th riffle. 0 Tile differmw CD in nitro en periods, which result d from an increase of mean organic nitrogen concentrafions per period, appears to be a product of "itrate additions, although some increase could have occurred from nitr03en fixation by Anabrena. Within a tertic‘ler two- shinjle sample the veriability was of on quite large (Tables 9 and 10). The same factors that re re spo onsible for W riabilit; within chlorOphyll samples are believed to be lcrg 1;; reaponsible for this. However, the variability does not em large eno1ggh to de- stroy the value of thi proc: d re since the increases and decrea.ses in averaged period values of organic nitro;on and phytOpigment are closely related. ’3' C H 93 C'- H o O D 0) i“ [—l “FEtOni neLfi t—-orranic :itroren Ck A linear r33re331on WuS used to determine the pnytm i;- ment concentration--orranic nitrogen relationship (Fig. 15). his regression is given by the formula: § 3 a + bX where Y is the predicted value of organic nitrogen; X is the known phytOpig m;nt concentration; b is the lepe of the regression line; and a is the Y intercept. The lepe is found by the equatio on; zxy .. (EXHZYZ b : q (n ZX" - (nod :5 and the Y'intercept is calculated by the formula: Figure }- I ‘Jl . Regression Lines Expressing the Re- lationship Between PhytOpiQMOit Absorb— ancy Uni s “’ rum Milligrsms of Organic Nitrogen for Riffle and Pool Zones n 9 x @9583? q ''''''' NWoNL —°°l Psi vii ooh om? own cow. j I '0 6 L to ; é UQBOJHN 911045th 58 ZY-bZX n a : In the riffle and pool zones Y = -.089 +-.0027X and Y : —.023 +-.00215X,respectively. Although not determined, the variance of points about the regression appeared to be large. is coefficient of correlation between phytopigment and organic nitrogen is given by the formula: Z XY - '(ZX) (ZY) I‘ = n [e2 - is» (s2 - sea]? The calculated correlation coefficients for rifflezxnd pool regions are .73 and .63 respectively. It can be seen that a fairly good correlation was ob- tained, which probably was due to each variable's relation- ship to cellular weight. The higher correlation in the riffle indicates that for a given unit of chlor0phyll the riffle community contains slightly more organic nitrogen. This might mean that nitrogen pumped from the reservoir first becomes available to the riffle area. Eggal‘dgy'weipht-—pigment relationship The results of total dry weight--pigment analysis are listed in table 11. The relationship of these data is expressed by a linear regression of Y = 3.u7 - .0719x (Fig. 16). Most of these data came from the gnabaena cam- munity. Only the lowest points on the slepe were represen- TABLE 11 PHYTOPIGMENT PER UNIT DRY WEIGHT PER UNIT AREA AFTER TWO WEEKS OF EXPOSURE Riffle Pool Riffle Pool Date AAXlOJ AAXlO3 mg. dry wt. mg. dry wt. u-13 12 12 0.8 0.8 h-al 56 68 5.6 22.8 8-23 32 64 9.2 21.6 8-29 208 132 18.8 12.8 5-3 288 154 29.0 21.8 5-9 252 352 37.8 27.6 5-11 - - 51o - - - 79.6 5-17 SSA 510 ul-2 LO- 5-23 584 208 35.6 21.6 5-25 58h u86 u6.6 M3.6 5-31 - - 67h - - - 52-h 6-6 1112 A78 92.8 37.8 6-8 L9H 6&6 30.6 h7.0 Figsrrc 16 . Regression EXpressin; the Relation— ship of PhytOpigment Absorbancy and Milligrams of Total Dry Weight for All Artificial Stream Communities mO_ X >udoflaom£ < cot 00.5 no? can cow 0%... q a a _ J _ 4 ton. < o 23?. r®.nku-n-luuu$ ON é +H553M ’{JCI Slumszmw ' 1 3 J: 8 61 tative of the diatomrgreen algae pioneers since growth was slight during their presence. The calculation of separate slopes did not seem justifiable. The coefficient of correlation was calculated to be .89. A lower correlation of .616 between organic weight and chlor0phyll was given by Riley (1940). However, in certain Red Cedar River diatom communities Peters (M. S.) found that a common correlation of .93 occurred using organic weight versus ethanol pigment extracts. If it is assumed that total dry weight minus ethanol soluble compounds approaches Peters organic wei hts which were corrected for the loss of soluble compounds, then the correlation coefficient for blue-green algae is slightly less than that for diatoms. Riley (lQhO) indicates that partial correlations are slight- ly lower for the blue-greens. In view of the statements above, it is still interest- ing to note that two such diverse groups of a gae with different pigment characteristics are so closely comparable. This might in part be why deviations of all points from the common calculated SlOpe gave no trend to justify computation of sepa ate slepcs. organic nitroggn--total[dry_weieht relationship_ The relationship between organic nitrogen and total dry weight is eXpressed as percent organic nitrogen (Table 12). Only two shingles were analyzed per period in ques- tion. Of the two shingles one was removed near the beginning TABLE 12 PERCEKP CELLULAR NITROGEN PER UNIT AREA arzaa TWO NEEKS EXPOSURE AT Tee BEGINNINGS AND ENDS OF PERIODS 2, 3, AND a Date Pool Riffle Mg N Mg Dry % N Mg N :3 Dry % N Wt Wt h-13 trace 0.8 trace trace 0.8 trace h-23 .30 21.6 l.h .05 9.2 0.6 h-29 .31 l2.h 2.5 .12 18.h 0.8 5-9 .59 27.6 2.1 .76 37.8 2.1 5-17 .85 40.h 2.1 .97 ul.2 2.h 5-23 0&2 21.6 2.0 1.75 35.6 “-09 63 and one near the end of the period. This analysis was only conducted to obtain further insight 11to the nitrogen metab- olism within periods of nitrogen addition. Data for periods 2, 3, and h are illustrated in figure 17. It is evident that the cellular nitrogen content remain- ed below 2.5 percent most of the time. Only on May 23 in the riffle area was this value exceeded. Foga (lghh) states that nitrogen fix ng blu e-green algae contain a rather high organic nitroven content ne-a r? or 8 percent. Gerloff and Skoog (195A) have pointed out that an internal concentration of nitrOgen above four percent in cells of Microcystis is luxury consumption, w1me eas grow1.h below M113 amount is prOporti onal to the supply of the element. The organic nitrOgen in the pool zone increased until the beginnine of tariod 3. Thereafter a slight drOp in the level occurred. In h: riffle area the percent of nitrogen rose consistently thro oughout the experiment. This rise migat be accounted for bv the uroximity of t1:e riffle to the incomin n1 tro; en from the reservoir influent, with the riffle community depleting most of the nitro;en in solution before it reaches the pool. This eXplanaULOH would account for the stabilization or drOp of percent organic nitrogen in the pool after period 2. However, since Anabaena domin- ated the stream community from the middle of period 3 until the end of the project, nitrogen fixation may in part be responsible for incre:.ses withi 1n periods 3 d h. It is Figure 17. Percent of Cellul hitrogen 1 Ends Of ar at the Beginnings ar inmost. z +¢MUL®Q 5 HT 3 fl 4 . u. ..m MN a ”P: am AA... 3 1d O . .m P. m. a W ......w: J .m Mm ...m w 2 R P .w .m P 65 interesting to note that the percent nitrogen relationship between pool and r1 f1 “1e compares closely to mean period organic nitrogena nd 1mtoyi3nent values (Figures 12 and 1h). stream.primary production Up to this time only relative rates of primary produc- tion have been mentioned. The calculation of absolute rates treating ata from all samples would be difficult if not impossible since there was considerable overlap of shingle exposure. It was necessary to use only one set which was removed and replaced at two we3: intervals for the absolute estimate. This group chosen, set 1 of both riffle and pool, sampled each nitrogen period in er tircty and t1! 1us r pr;serts the accumulation of algae during that period. r1hese values should 3ive a close approx1nct1o to so Ial primary produc- tion since consumers were not noted to exist in the al,3al mat until shortlry before thceexper11e t termizlated. All chlorOphyll values were converted to total dry weight by emplovin3 the reg1e ssion in figure 16. Results of calcula- tions are listed in table 13. An increase in absolute production per period followed the cumulative additions of nitrate with a mean daily produc- tion based on 8 weeks of 250.2 and 201.u milligrams dry weight per square meter {er da y for riffle and pool reSpec- tively. The higher riffle production was due to a greatly accelerated rate during periods A and 5 (Figures 18 and 19). This more than made up for higher pool rates in the 7-,#.1.J33\Jv~fl« NT~<<§4LN~ =1ap¢a§5§u open seahom soaaozvo&m doom N.omm u u u n n u u u u u a u u s u7maevs asaewfiaafis nae: :.mao.:au . - u n - u u s u . szmsoop m¢s meuawfiaafia Hence H.mOH . u - -:asoos oveo ms maupwaaaae sauce .s s. em. 0.5 me m one on mm-m m m.o m.mmm.w :.ew mm m mmum op e-m M m.o a :.omfl.~ m.ea Hod m o-m on mm-: .o 5. es .umui mm H Immuz op HH-:I m :3: m ea Lee. ml: 18:. m .18 ms :3? m «.3 ms comonflz a we 83 823 atmnv ms. at hue we pi hue we moa N << o>apeasasu coauosUOEm OHHMHm zeHaoeaoxm MmeeHmm seamen queHeHsme ma mqm