THE EFFECT OF VARIED LEVELS OF NITROGEN, PHOSPHORUS, POTASSIUM, AND BORON IN SOIL ON THE YIELD AND CHEMICAL COMPOSITION OF GREENHOUSE TOMATOES By ANDREW J. WATSON A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Soil Science 1949 ACKNOWLEDGMENT The author wishes to express his appreciation to Dr. R.L. Cook for his guidance and assistance throughout the course of this investigation. He is also indebted Dr. C.E. Millar and Dr. J.F. Davis for their helpful sug­ gestions in the preparation of the manuscript and to the American Potash Institute, Inc., of Washington, D.C. for its fellowship which made this study possible. The photographs included were taken by Dr. Cook. TABLE OP CONTENT'S .GE INTRODUCTION 1 EXPERIMENTAL 1 A. Soil Testing Procedure 1 B. Methods of Soil Analysis 2 C. Calibration of the Fertilizer "Fixing Power" of Soil 3 D. Plan of Greenhouse Experiments 4 E. Descriptions of Soils Used 5 F. Analytical Procedures 8 SOME EFFECTS AND INTERRELATIONSHIPS OF NITROGEN, PHOSPHORUS, AND POTASSIUM LEVELS ON TOMATOES 10 A. Nitrogen, Phosphorus, and Potassium Levels on Oshtemo Loamy Sand 1. Experimental 2. Results and. Discussion a. Conductivity and pH of the Soil b. Dry Weight and Fruit Yield 12 12 15 15 19 B. Nitrogen, Phosphorus, and Potassium Levels on Brookston Silt Loam 1. Experimental 2. Results and Discussion a. Dry Weight b. Fruit Yield c. Nitrogen Content of Leaf Samples d. Phosphorus Content of Leaf Samples e. Potassium Content of Leaf Samples 48 48 50 50 60 62 64 64 DIFFERENTIAL VARIETAL RESPONSE OF TOMATOES TO POTASSIUM FERTILIZER 68 A, Experimental 70 B. Results and Discussion 71 EFFECT OF BORON LEVELS ON THE YIELD AND CHEMICAL COMPOSITION OF TOMATOES 78 TABLE OF CONTENTS PAGE A. Response to Boron Treatments on Overlimed Oshtemo Loamy Sand 1. Experimental 2. Results and Disctission B. Response to Boron Treatments onThomas Loamy Sand, Miami Loam, and Wisner Sandy Loam 1. Experimental 2. Results and Discussion 80 80 81 85 85 86 VI. SUMMARY AND CONCLUSIONS 94 VII. LITERATURE CITED 98 Many commercial growers of greenhouse tomatoes are attempting, by the use of soil and tissue tests, to control the nutrient status of their soils. Their large capital investment makes it important that all scientific aids available be used to help them obtain maximum yields. Fer­ tilizer costs are secondary to high yields of marketable fruit. Many of the growers in Michigan do their own soil testing with the Spurway Simplex Soil Testing Kit (32). The question arises regarding the most desirable nutrient levels or range of levels in soils for the elements required by tomatoes. Since the Spurway procedure for soil testing is commonly used in this area, it appeared that a nutrient level problem on the tomato, using this testing procedure, would have practical as well as scientific value. Accordingly this investigation was undertaken to pro­ vide further information regarding the proper range of con­ centrations of nitrogen, phosphorus, potassium, and boron in soils for greenhouse tomatoes. The interrelationships of varying levels of these elements were also studied. Chemical analyses of the plant tissues were made to determine the effect of treatment on the chemical composition of the plants. EXPERIMENTAL Soil Testing Procedure In this study soil testing was carried out on air dry soil according to the rapid methods of Spurway (32). Only his '•active1* tests were used. Briefly, this consisted of extracting the soil with approximately 0.018 N acetic acid, using a 1:4 ratio by weight of soil to extracting solution* Tests for nitrogen, phosphorus, and potassium were run on the filt©red extract* Nitrates were determined by use of a sulfuric acid solution of diphenylamine. Phosphorus was estimated by use of an ammonium molybdate solution with stannous chloride as the reducing agent. The sodium cobalti- nitrite procedure was used for potassium determinations. The colors or turbidities developed in the above tests were com­ pared with those of standard solutions carried through the same procedures. All soil test results are expressed as parts per million in the soil extract. Methods of Soil Analysis Mechanical analyses of the soils were determined by the hydrometer method proposed by Bouyoucos (5). Base ex­ change capacity was determined by leaching with neutral ammonium acetate as suggested by Peech, et al. (26). The procedure followed for the determination of exchangeable potassium was essentially that of Lawton (16). This is a colorimetric method which utilizes hexahitrodiphenylamine (dipicrylamine)* The carbon train technique was used for determining carbon. Per cent organic matter was Estimated by multiplying the per cent carbon by the factor 1.724, assuming that organ­ ic matter is 58 per cent carbon* 3 All pH measurements of the soils were made potentiometrically, using a Macbeth, alternating current pH m e t e r with glass electrodes. Calibration of the Fertilizer "Fixing Power" of Soil Two experiments were c o n d u c t e d which involved the maintenance of nitrogen, phosphorus, stant levels in the soil. and potassium at c o n ­ The p r o b l e m arose as to the amount of the various C.P. reagents needed in order to o b ­ tain the desired levels in the soil. These amounts were determined by means of a tumbler experiment set up in the laboratory. One hundred and f i f t y gram samples of the screened, air dry soil to be u s e d later in the greenhouse experiment were placed in 24 g l a s s tumblers. ments, in duplicate, were applied. amounts of the nitrogen, phosphorus, Twelve treat­ They included varied and potassium reagents. The amount of each reagent used progressed by treatment f r o m zero nitrogen, phosphorus, or p o t a s s i u m to treatments i n c l u d ­ ing sufficient quantities of e a c h to bring the soil tests to a value higher than any level d e s i r e d for that element in the future experiment. Lime w a s added to the Oshtemo soil in the tumblers at a rate equivalent to one ton per acre. The nitrate and potassium salts w e r e added in solution, w h i l e the calcium carbonate and primary monocalcium phosphate w e r e added as dry salts due to their low solubilities. ments were mixed thoroughly into The t r e a t ­ the soils which were then brought to their moisture equivalent with distilled w a t e r 4 and allowed to incubate at room temperature for ten days. Additional distilled water was added as needed to keep the soils at the desired moisture content during incubation* At the end of the incubation period the soil was removed from the tumblers, air dried, and tested for avail­ able nitrogen, phosphorus, and potassium according to the Spurway procedure. Using the results obtained, calibration curves were plotted for each of the three elements concerned. The test results were plotted as the ordinate and the milligrams of the respective salts added as the abscissa. By using these graphs the amount of a particular reagent required to bring any given amount of this soil to a desired nutrient level was readily calculated. During the experiments, when soil tests indicated that a nutrient level had dropped, the calibration curves were also used to calculate the amount of a reagent necessary to restore the soil to its proper level. Plan of Greenhouse Experiments Greenhouse experiments were conducted with tomato plants grown in closed bottom, glazed, earthenware jars. Surface soils of five common Michigan soil types were collected. The soils were air dried and screened through a three-eighths inch mesh rotary type screen before using. During the growing period soil moisture was maintain­ ed at approximately moisture equivalent by additions of dis­ tilled water. 5 Descriptions of Soils Used The following soils were used and the descriptions are essentially the same as given in the respective county soil survey reports published by the United States Depart­ ment of Agricultures 1, Oshtemo loamy sands A soil type with a yellowishbrown surface layer of loamy sand which occurs on nearly level land. Very little clay is present in the subsoil and loose dry sand or sand and gravel extend to depths of several feeb. This soil is naturally very drouthy, is low in organic matter, and is strongly acid in reaction,, Surface soil of this type was^obtained from the Bose Lake Wildlife Exper­ imental Farm, Clinton County, Michigan* 2. Brookston silt loams A soil developed on heavier materials of the flat basin lands, valleys, and depressions w h i c h were originally wet and swampy. It has a very dark gray or nearly black surface layer which is rich in organic matter and ranges from 6 to 10 inches in thickness. The surface layer grades into a gray or yellowish-gray layer, 4 to 8 inches thick, which is more coherent and has a higher clay content. Beneath this is found steel gray or bluish- gray plastic or sticky clay, slightly mottled with yellow and rust-brown. The substratum consists of clayey glacial till containing more or less lime. This soil is relatively h i g h in natural fertility when drained. The Brookston used was from the Michigan State College farm, Ingham County, East Lansing, Michigan. 30 Thomas loamy sands The surface layer of this soil is black loamy sand, high in organic matter, and contains a considerable amount of light colored fine sand. At a depth of about 12 inches the surface layer grades into a thin lay­ er of gray sandy clay which contains some fine gravel and some small shells or fragments of shells. Below 15 inches light brown, yellow, and gray plastic clay occurs. The soil is naturally poorly drained and is alkaline throughout the profile. Soil of this type was obtained near Saginaw Bay in Tuscola County, Michigan, 4. Wisner sandy loami This soil consists of very dark gray sandy loam to a depth of 4 inches. Between 4 and 8 inches the material is lighter in color, and gray clay loam occur3 at 8 to 15 inches. Small fragments of shells are present in the surface layers, A plastic clay, colored with light gray, yellow, and yellowish-brown, occurs at depths greater than 15 inches. It Is naturally poorly drained and is alkaline throughout the profile. Surface soil of this type was obtained from the Wiergorski farm, Tuscola County, near Wisner, Michigan, 5, Miami loamJ Under cultivation this soil type has a surface layer of light grayish-yellow loam which extends to an average depth.of 12 Inches. The subsoil, which ranges from 12 to 16 Inches in thickness, is dull yellowish-brown heavy clay. The substratum. Is light grayish-yellow or light gray heavy calcareous glacial till. Soil of this type was 7 obtained from Tuscola County, about 6 miles northeast of Caro, Michigan. A summary of some of the physical and chemical prop­ erties of these soils is shown in Table 1. TABLE 1.— Some physical and chemical properties of the soils used. Soil type Particle size distribution % sand % silt % clay (2 ) (3) (1 ) % organic matter Original pH Base exchange capacity (4) Oshtemo loamy sand 83.2 11.9 4.9 0.92 4.43 3.5 Brookston silt loam 28.0 64.8 7.2 5.06 6.10 22.7 Thomas loamy sand 88.0 10.9 1.1 6.89 7.00 20.4 Wisner sandy loam 58.0 39.0 3.0 5.96 7.40 13.9 Miami loam 49.8 47.0 3.2 2.29 7.25 10.2 (1) (2) (3) (4) 1.0-0.05 mm. 0.05-0.002 mm. < 0.002 mm. millieqiiivalents per 100 grams of soil. Analytical Procedures The plant materials analyzed were oven dried at 90 to 95°C«, then ground in a small Wiley Mill to pass through a 20 mesh seive. Before grinding, the tomato fruit was thin­ ly sliced, placed on enamel trays and dried in the oven. A one gram sample of the oven dried material was placed in a porcelain crucible, and 1 ml. of lsl H2SO4 was added followed by 3 ml. of distilled water to moisten the sample. It was then dried in an oven and ashed overnight in a muffle furnace at approximately 550°C. After cooling, the ash was moistened with a few drops of distilled water, fol­ lowed by 3 ml. of concentrated HC1, This mixture was heat­ ed to boiling, diluted with a few ml. of distilled water and filtered. The filtrate was collected in a 50 ml. volumetric flask which was brought to volume with small, successive washings of the residue. The solution was thoroughly mixed and stored in a tightly stoppered, soft glass bottle. Phosphorus, potassium, calcium, and magnesium were determined on aliquots of the plant ash extract. To deter­ mine phosphorus a colorimetric method using ammonium molybdate and the Fisk-Subbarow (10) reducing agent was followed. Potassium was determined colorimetrically by a modification of the method suggested by Lawton (16), which uses hexanitrodiphenylamine. For determining calcium a 10 ml. aliquot of the extract was precipitated as oxalate and then titrated with standardized KMn04 . ""The titan yellow procedure,-' proposed, b y magnesium and English (25), was followed for the determinations, Total according Peech to n i t r o g e n was determined on the dried material tine Kjeldahl-Gunning procedure (1). 10 SOME EFFECTS AND INTERRELATIONSHIPS OF NITROGEN, PHOSPHORUS, AND POTASSIUM LEVELS ON TOMATOES. Agricultural scientists have recognized for nearly a century the mutual effect of one element upon another in the nutrition of plants. In this period, especially during the past three decades, a great volume of literature has accu­ mulated on the subject. Thomas (35), In 1932, published an extensive review of literature regarding ionic antagonisms relative to plant nutrition giving particular emphasis to the reciprocal effects of nitrogen, phosphorus, and potassium. Prior to that time considerable attention had been devoted to the development of physiologically balanced nutrient solu­ tions with the objective of discovering the optimum ratios between various nutritive elements. Investigations Into the interrelationship of nitrogen, phosphorus, and potassium have been generally approached by the use of nutrient solutions. Plants were grown either directly In such solutions or in sand cultures with frequent applications of nutrient solutions. Plant yields correlated with chemical analyses of plant tissues from plants grown under many different nutrient conditions have provided con­ siderable insight Into the interrelation between essential plant nutrients. In recent years workers have emphasized the impor­ tance of maintaining the proper concentration and balance of available nutrient elements In the soil for the production 11 of maximum crop yields. Spurway (33), in considering the available nutrients in soils, stresses the importance of '•level and balance" for maximum production. This inves­ tigator points out that the same crop yield may be produced under many conditions of unbalanced fertility, but he sug­ gests fertilizing in such a way as to both raise the level of the required nutrients and to bring them into better physiological balance for the plant. The technique of foliar analysis has been valuable in studying and interpreting various interrelationships between nutrient elements. Salter and Ames (29), in 1928, suggested studying the availability of soil nutrients by means of chemical analyses of plants. Since that time there has been considerable interest in this technique as an aid in deter­ mining fertilizer needs. Lundegardh (17) has pointed out that the fundamental concept of foliar analysis is that the amouht of nutrient salts absorbed by the plants reflects the availability of these salts in the soil under the actual growing conditions of the plant. Thomas (36,37,38,39,40) has adopted the foliar anal­ ysis technique in the study of nutrient relationships of several crops. In the interpretation of such analyses, he recognizes the importance of two factors, "quantity" and "quality". For maximum growth each element must be above its critical concentration and at the same time be in proper 12 balance with all other elements. This concept of nutrient element balance is essentially in harmony with the concepts of Shear, Crane, and Myers (30) who have conducted extensive investigations on the nutrition of tung trees. Macy (21) and Ulrich (41) proposed the use of foliar analysis to determine the "critical percentage" or "critical concentration" of the various elements in plants. If plants showed a concentration below these critical values they would be expected to respond to an application of that el­ ement. There is need for further research regarding desir­ able nutrient levels and nutrient relationships in both field and greenhouse soils. Much work remains to be done with regard to these relationships for specific crops. This is especially true with intensively grown crops such as greenhouse tomatoes. Nitrogen, Phosphorus, and Potassium Levels on Oshtemo Loamy Sand Experimental i A 5x5x5 factorial greenhouse exper­ iment using Oshtemo loamy sand was conducted during the spring of 1948. Single plants, started In flats, were grown in 2-gallon glazed jars filled with 8 kilograms of screened, air dry soil. Nutrient levels in the soil extract were maintained at 0, 25, 50, 100 and 200 ppm of NO3 , 0, 2s-, 5, 10, and 25 ppm of phosphorus, and 0, 15, 30, 45, and 60 ppm of potassium. These nutrient levels are indicated by sub- 13 scripts throughout this discussion. In all cases the zero level was the original soil with no addition of the element concerned. The untreated Oshtemo soil was extremely low in all three nutrients. Because the Oshtemo soil had a pH of 4.4 an amount of precipitated 0aC03 equivalent to one ton per acre was added to each jar. Manganese sulphate and magnesium sulphate, each equivalent to 100 pounds per acre, and sodium tetra­ borate (NagB^Or^) equivalent to 5 pounds per acre were also applied to all jars. A summary of the varied treatments Is given in Table 2. Seeds of the Master Marglobe (Stokes) variety were planted in flats on January 3, 1948. Oshtemo loamy sand which received 200 pounds per acre of 2-16-8 fertilizer was used in the flats. The treatments were thoroughly mixed into the soil on January 9 to 12. Sufficient distilled water was added at this time to bring the soil in each jar to its moisture equivalent which was approximately 12 per cent moisture. On February 7 the plants were transplanted from the flats to the jars. The roots of each plant were washed free of soil when transplanted. At this time the plants were had uniform in size, about three Inches In height, andAstarted their second pair of true leaves. All plants were a normal green color and healthy in appearance. The soils were tested frequently during the growth of 14 the crop and additional nutrients were applied as needed to maintain the original levels. The fruits were picked as they ripened. When the experiment was terminated on July 6 the remaining green fruits were picked and included in the total yields, TABLE 2.— A summary of the nitrogen, phosphorus, and potassium treatments applied to Oshtemo loamy sand. Nutrient levfel N0 n 25 N 50 N100 N200 *0 C.P. reagent used P10 P25 K0 *15 k30 K45 K60 Pounds per acre equivalent 0.00 0.0 1.43 357,5 ii 2.86 715,0 n 5.72 1430.0 11.44 2860.0 G a (H 2PO4 )g •HgO e.oo 0.0 11 3.90 976.0 5.87 1466,5 11 8.82 2206.0 n 14.32 3580.0 K 2SO4 0.00 0.0 ii 2.56 640.0 ti 4.80 1200.0 ti 8.96 2240.0 11 13.12 3280.0 NH4N03 m ii p2lr *5 Gm. added per jar ii 15 Results and Discussion: Conductivity and pH of the Soil: Many of the fertil­ izer applications in this study were extremely high. To obtain information regarding the effect of total soluble salt concentrations on plant growth, conductivity measure­ ments were made on soil samples collected at the time of transplanting. Conductivity was measured with a Solu-Bridge, type RD, manufactured by Industrial Instruments Inc., Jersey City, New Jersey. To 50 grams of air dry soil, 100 ml. of distilled water was added and the mixture was stirred for 5 minutes. After settling for 2 minutes the liquid was decant­ ed and conductivity was determined on the decanted liquid. These results, presented in Table 3, show that at the time of sampling the nitrogen, phosphorus, and potassium fertilizers had little consistent effect on the pH of the soil, but they had a marked influence on the conductivity values. The potassium sulphate and ammonium nitrate treat­ ments caused noticeable increases in the conductivity values but the phosphate treatments did not. The only plants in the experiment which showed any indication of an excessive soluble salt concentration were those that received the highest application of ammonium nitrate. The specific conductivity of the decanted liquid from soil which received this treatment varied from 50 to 70 x 10”® mhos at the low potassium levels to 100 to 140 x 10**'® mhos at the highest levels of potassium. Within 16 TABLE 3.— The effect of treatment on conductivity and pH of Oshtemo loamy sand. Conductiv­ Treatment ity of soil solution Vr PH Conductiv­ Treatment ity of soil solution *■ pH N0P0K0 12 6.6 n 25p 0k0 17 6.4 N0PoKl5 19 6.5 N25p 0k15 30 6.2 n 0p 0k30 26 6.4 N25P0K30 38 6.2 n 0p 0k45 54 6.3 N 25P0K45 60 6.2 N o P0K60 69 6.3 N25P0K60 75 6.1 N0P2§K0 13 6.7 N25P2iEo 25 6.3 n 0p 2^k15 21 6.6 N 25p2^El5 31 6.2 N0p2iK3O 32 6.5 N25p2|-K30 43 6.1 n OP2|K45 49 6.4 N 25p 2iK4 5 47 6.1 N0p2iK60 55 6.4 N25p2iK60 58 6.3 N0p5K0 12 6.7 N 25p 5K0 20 6.2 N0p5K15 ‘ 28 6.5 N25p5Kl5 30 6.2 N0p5K30 29 6.4 n 25p 5k30 43 6.1 n 0 p 5k45 49 6.3 w 25p 5E45 50 6.1 n 0p 5k60 58 6.4 n 25p 5k60 73 6.1 N0p10K0 12 6.6 n 25p 10E0 3i 6.1 N0P10K15 19 6.5 n 25P10k15 40 6.1 N0P10K30 30 6.4 N25P10K30 61 6.0 n 0 p 10k45 46 6.3 n 25p 10E45 50 6.0 n 0p 10k60 60 6.4 N25p 10k60 65 6.2 n 0p 25k0 13 6.4 n 25p 25E0 24 6.1 N0P25K15 20 6.3 N25p 25Kl5 33 6.0 N0p25*30 30 6.3 N25P25K30 44 6.0 H0P25K45 50 6.2 N25P25K45 66 6.0 n 0 p 25E60' 65 6.2 n 25p 25k60 86 5.9 # Specific conductivity x 105 mhos at 25°C. 17 TABLE 3.— (continued) The effect of treatment on conductivity and pH of Oshtemo loamy sand. Conductiv­ Treatment ity of soil solution # pH Conductiv­ Treatment ity of soil solution pH N 50P0K 0 32 5.9 N100P0K0 ‘ 59 6.0 n 50?0K15 36 5.9 n 100 p 0k15 46 6.1 n 50P0K 30 47 5.9 n 100P0K30 71 6.0 w 50P0K45 84 5.9 N100P0K45 85 6.1 50p 0k60 84 6.0 n 100p 0k60 84 6.1 N50P2^K0 39 6.0 N100P2^K0 35 6.1 N 50p2 ^ i 5 52 6.0 N100p2iKi5 70 6.0 n 50P2^k30 58 6.0 N100p2-g-K30 75 6.1 N 50P2iK45 59 6.0 N100P2iK4'5 84 6.0 N50P2iK60 78 6.0 N100P2iK60 96 6.0 N50?5K0 36 6,0 n 100p 61 6.1 N50p 5k15 43 6.0 w 100p 5k15 62 6.1 50p 5k 30 53 5.9 N100p 5K30 73 6.1 N50P5K45 62 6.0 W100P5K45 89 6.0 N 50P5K60 65 6.1 N100P 5K60 85 6.0 N 50P10K0 37 5.9 Ni q o ^ 44 6.1 n 50P10k15 42 6.0 n100p10k15 55 6.1 W 50P10K30 47 5.9 n100P10K30 67 6.0 N 50P10K45 70 5.9 N100P10k45 72 6.0 n 50p 10k60 95 5.9 N100P10K6C 95 6.0 N50p 25k0 29 6,0 n 100 p 25k0 44 6.1 50p 25k15 45 6.0 n 100 p 25k15 47 6.1 N 50p 25k30 50 5.9 w 100p 25k 3C 65 6.0 N 50p 25K45 60 5.9 n 100p 25k45 75 6.0 n 50p 25k60 87 5.9 n 100p 25k60 95 6.0 n n n \ 5k0 o Kq * Specific conductivity x 10s mhos at 25°C. 18 TABLE 3.— (continued) The effect of treatment on conductivity and pH Of Oshtemo loamy 3and. Conductiv­ Treatment ity of soil solution pH CondtictivTreatment ity of soil solution Vr pH ^200^*0 70 6»3 N200p5K45 120 6.1 N200P0K15 82 6.2 N200P5K60 88 6.1 ®200P0K30 70 6.3 NgOO^oKo 57 6.2 n 200P0k45 100 6.2 N200p10K15 60 6.2 N200P0K60 110 6.4 N200P10K30 90 6.3 n 200p 2£kO 71 6.3 N200p10K45 140 6.0 n 200p 2|k15 73 6.4 N200p10K60 125 6.0 n 200P2^k30 77 6.2 N200p25K0 72 6.2 N200P2iK45 108 6.2 n 200p 25k15 68 6.6 N200p2iK60 110 6.2 N200p25K30 65 6.2 N200p5K0 67 6.2 N200p25K45 88 6.0 n 200p 5K15 50 6.3 n 200p 25k60 130 5.9 $200p 5k30 100 6.2 ^-Specific conductivity x 10s mhos at 25°C. 19 this range all plants showed similar toxicity symptoms. Since^ the lower values of this high nitrogen group were low­ er than the values obtained for some soils at lower nitrogen levels, where normal growth occurred, it was concluded that the toxicity was related to the actual NH4 or N03 concentra­ tion rather than to the total soluble salt concentration. Dry Weight and Fruit Yield: The dry weight and fruit yield per plant, as a result of each treatment, is presented in Table 4. Both factors were analyzed statistically by the analysis of variance. For both dry weight and fruit yield, the effects of nitrogen, phosphorus, and potassium levels were highly significant as were also the NP, NK, and PK interactions. The levels of nitrogen were more Influential on dry weight and fruit yields than were the levels of either of the other elements. There was a highly significant differ­ ence between each of the five levels of nitrogen on both fac­ tors. As Table 5 and Fig. 1 show, N50 resulted in the highest average dry weight. However, during the early stages of growth the plants at N 25 were largest, as illus­ trated in Fig. 2. Part of this difference in final dry weight is attributed to the difficulty encountered in keep­ ing the nitrogen level from dropping below 25 ppm when the plants were large and actively growing. There was a marked depression of growth where nitrogen levels were above 50 ppm in the soil extract, as Fig. 1 shows. Severe toxicity 20 TABLE 4* — The effect of treatment on dry weight and fruit yield of tomato plants# Treatment N0Pq Ko Dry weight Yield of Dry weight Yield of of plant fruit fruit Treatment of plant (gms.) (gms.) (gms.) (gms.) 377.2 6 35 55.5 N0Po K15 8 39.4 N25p0K15 47 442.7 n 0p 0k30 6 64.3 N25p0K30 50 261.2 N o I*Ok45 7 76.3 N 25p0K45 34 366.3 *0*0*60 11 .51.4 N25p0K60 49 207.3 * 0p2|K0 6 21.5 N25p2iF0 28 70.2 N0p2iK15 6 77.3 N25p2iK15 84 555.9 N0P2^K30 13 0.0 N25p2iK30 96 599.0 Nop2iK45 7 90.2 N25p2 p 4 5 70 941.3 n 0P2^K60 9 99.7 N 25P2^K60 82 784.9 N0P5K0 3 84.6 N25p5K0 19 220.1 N0P5K15 9 93.2 N25p5Kl5 84 513.2 N0p5K30 8 63.2 N25P5K30 99 721.4 N0p5K45. 7 50.2- N25p5K45 78 945.2 N0P5K60 7 88.5 N25P5K60 85 625.4 N0P10K0 7 59.1 N25p10K0 21 68.7 N0p10K15 8 102.6 N 25P10K15 75 940.0 n 0 p 10k30 10 96.4 n 25p 10k30 77 979.4 n 0p 10k45 6 74.7 N25P10K45 81 878.3 n 0 p 10k60 9 81.4 ^FlO^O 43 925.0 Nop25K0 , 5 0.0 N25P25K0 30 177.8 N0P25K15 9 99.5 N25p25K15 78 949.8 n 0p 25k30 7 193.9 n 25p 25k30 91 650.7 N0p 25k45 6 128.4 N25p 25k45 82 1015.0 No^^O 5 0.0 *25*25*60 73 924.4 21 TABLE 4, — (continued) The effect of treatment on dry weight and fruit yield of tomato plants. Treatment n 50p 0k0 Dry weight Yield of Dry weight Yield of of plant fruit Treatment of plant fruit (gms. ) (gms.) (gms •) (gms.) 92.8'' N100P0K0 1 13 0.0 N50?0k15 8 0.0 N100p0K15 26 35.2 n 50p 0k30 33 108.3 n 100p 0k30 3 0.0 w 50p0*45 35 133.9 Nl00p0K45 3 0.0 N50P0K60 10 61.5 N100P0K60 3 12.5 N50p2iK0 26 102.2 Nl00p2iK0 30 190.3 N50P2£K15 92 465.1 N100p2iK15 55 287.5 N50p2^K30 93 615.3 N100p2iK30 N50P2£K45 106 425.6 N100p2iK45 66 518.3 N50p2iK60 84 628.7 Nl00p2iK60 71 389.0 n 50p 5k0 22 142.7 N100p5K0 20 158.2 n 50p 5k15 86 488.9 NlOOp5K15 50 300.0 n 50P5k 30 107 617.4 n 100p 5k30 64 532.2 n 50p 5k45 111 734.9 Nl00p5K45 72 685.3 Nso^Keo- 96 572.9 Nl00p5K60 82 453.8 n 5o p i o Kq 22 120.9 N100P10K0 34 114.0 n 50p 10k15 71 811.6 N100p10K15 75 535.4 n 50p 10k30 111 705.7 N100p10K30 96 365.2 n 50P10k45 92 958.6 n 100p 10k45 85 610.8 N50P10K60 82 950.2 N100P10K60 100 429.9 ^O^KO 23 101.2 N100p25K0 21 43.4 N50P25K15 77 848.0 m 100p 25k15 32 513.9 N50P25K30 99 577.9“' n 100p 25k30 69 541.1 50p 25k45 121 659.3 n 100p 25k45 ?3 357.9 N50p25K60 126 529.9 N100P25K60 83 388.3 n 271.2 - . TABLE 4,—-(continued) The effect of treatment on dry weight and fruit yield of tomato plantse Treatment Dry weight Yield of Dry weight Yield of of plant fruit Treatment of plant fruit (gms.) (gms.) (gms.) (gins.) N20o Po K0 0 0.0 N200P5K45 66 256.4 N200p0K15 0 0.0 N200p 5k60 - 54 340.6 n 200p 0k30 0 0.0 N200p10K0 39 91.0 n 200p 0k45 0 0.0 n 200p 10k15 61 n 200P0k60 0 0.0 N200p 10k30 62 317.6 N200p2iK0 42 120.9 N200p10K45 31 43.6 N200P2iK15 55 153.8 N200P10K60 52 242.3 n 200p 2^k30 51 328.2 n 200p 25k0 24 28.8 N200p2iK45 60 115.1 n 200p 25k15 44 182.5 N200P2iK60 53 353.5 N200p25K30 50 521.5 n 200P5k0 40 110.8 N200p25K45 15 153.7 n 200P5k15 55 215.8 n 200p 25k60 45 ^ 265.9 n 200p 5k30 24 63.6 _ 241.8 23 TABLE 5*— Analysis of variance of dry weight of tomato plants grown on Oshtemo loamy sand. Mean square D.P. S. S. 124 150,663.4 4 62,046.2 15,511.6** 4 Nq vs. N254N5o +Ni o o +n200 1 46,214.5 46,214.5** 3 N25 vs• N 5 0 +N1 0 O+N2 0 O 1 2,225.0 2,225.0** 2 N50 1 10,956,8 10,956.8** 1 2,649.9 2,649.9** 28,883.6 7,220.9** 28,758.5** Source Total N levels vs . N100+N20o N 1 0 0 vs. N 2 0 O P levels 4 P0 vs. P2^+p5+p10+p25 1 28,758.5 3 P2 i vs. E>5+Plo+p25 1 25.8 25.8 2 P5 vs. Pio+p25 1 22.4 22.4 P10 vs. P25 1 76.9 76.9 20,600.2 5,150.1** 1 19,807.2 19,807.2** 1 630.8 630.8* 1 162.2 162.2 1 0.0 0.0 NP 16 13,736.8 858.6** NK 16 12,963.4 810.2** PK 16 4,733.2 295.8** NPK (error) 64 7,700.0 120.3 K levels 4 Kq 3 K15 vs Ki5+K30+K45+K60 . vs . K3 0 +K4 5 +K6 O 2 K3 0 vs. K4 5 +-K6 O K 45 v s . KgQ * Significant at 5% point. Significant at 1% point. 24 (1) Dry weight (2) l/lO bf fruit yield GRAMS 25 50 100 200 NITROGEN LEVEL (PPM N03 IN SOIL EXTRACT) FIG. 1.— Effect of nitrogen levels on the average dry weight and fruit yield of tomatoes. 25 PIG. 2.— The effect of nitrogen levels on plants adequately supplied with phosphorus and potas­ sium. Treatments ares 1-Nq P i o k 45» 2-N25Pi o k 45» 3-n 50p 10k 45» 4“n 100p 10k 45» 5“N200F10K45* symptoms occurred at the N 200 level. The plants were ex­ tremely stunted in growth, had a soft, semi-wilted appear- . ance, and were deep green in color. This is illustrated in Fig. 2, jar 5. Symptoms of nitrogen starvation occurred only on those plants that received the W q treatment. Six weeks after transplanting, the color of all plants in this group began to fade from green to light yellow. The symptoms appeared on the older leaflets first and progressed to the younger ones. Where there was only a moderate deficiency the leaflets were uniformly light green in color. In the advanced stages, however, the veins were distinctly purple in color while the interveinous tissue remained a faded yellow. During this advanced stage of nitrogen starvation the plants had a very tough, wiry texture in comparison to plants which received nitrogen. The occurrence and severity of these symptoms were unrelated to phosphorus and potassium treatments. The marked difference obtained between the Nq and N25 treatments is apparent from Fig. 3. FIGr. 3.— Plants on the left received no nitrogen while those on the right received the N25 treat­ ment. Both groups received various phosphorus and potassium treatments. 27 ”Fruit yields were closely related to the size of plants in this experiment* Pig. 1 and Table 6 show that N25 resulted in the highest fruit yield per plant and that no nitrogen, as well as excessive nitrogen, resulted in decreased fruit production. A marked response to phosphorus was obtained in this experiment. When dry weights are considered, the only significant difference between levels was between none and some phosphorus. This is shown in Table 5 and Pig. 4. Levels of Pgi, P5, Plo> 5111(1 p25 wer® equally effective in production of dry weight. Fruit yields were also increased by phosphorus treatments. As Table 6 and Pig. 4 show, P5 , PlO, and Pg5 resulted in the highest yields, yet these three levels were equally effective. Distinct phosphorus starvation symptoms occurred within ten days after transplanting on all plants that did not receive phosphorus fertilizer. These plants were spin­ dling, stunted, and dark green with distinct purpling of the under side of the leaves. The purpling extended over both the veins and the interveinous tissue. As the roots became better established, many, of the phosphorus starved plants lost their purple color and grew to a fair size. Symptoms similar to these have been described by MacGillivray (20) and Cook and Millar (8 ) for phosphorus starvation of tomatoes. A definite NP interaction was present in this experi­ ment. Increased amounts of phosphorus tended to counteract TABLE 6.— -Analysis of variance of fruit yield of tomato plants grown on Oshtemo loamy sand. D.F. S.S. 124 10,974,223,5 4 4,685,941.7 1,171,485.4** 4 N q vs. N25+n 50+n 100+w200 1 1,948,065.8 1,948,065.8** 3 Ngs vs. N5o+Nioo+n200 1 1,640,861.8 1,640,861.8** 2 N50 vs. Nioo+n200 1 825,282.2 825,282.2** 1 271,731.9 271,731.9** 4 1,746,054.3 436,513.6** 4 Pq vs © Y2%+*b+T10+?25 1 1,594,656.3 1,594,656.3** 3 1 95,815.4 95,815.4** 2 P5 vs. Pio+P25 1 39,690.7 . 39,690.7 PlO vs. Pg5 1 15,891.9 15,891.9 Source Total N levels N100 vs. NgQO P levels vs. P5+P10+P25 . Mean square 4 1,541,390.4 385,347.6** 1 1,502,232.5 1,502,232.5** 3 K]_5 vs. K30+K45+K60 1 13,895.1 13,895.1 2 K30 vs. K45+K6O 1 8,616.9 8,616.9 1 16,645.8 16,645.8 NP 16 690,314.8 43,144.7** NK 16 855,183,7 53,449.0** PK 16 647,019.5 40,438.7** NPK (error) 64- 808,319.1 12,630.0 K levels 4 Kq vs. Ki5+K3o+K45->-Kgo K45 vs. Kgo ## Significant at 1% point (1) Dry weight (2) l/lO of fruit yield 60 t- - 50 40 GRAMS 30 20 10 0 5 10 2.5 PHOSPHORUS LEVEL (PPM IN SOIL EXTRACT) 25 PIG. 4.— Effect of phosphorus levels on the average dry weight and fruit yield of tomatoes. 3° the toxic effects of excessive nitrogen. It can he seen in Fig. 5 that in the presence of excessive nitrogen, at least during the early stages of development, growth was consist­ ently improved by the higher phosphate levels. results have been reported by Emmert (9). Similar This Investigator, studying the nutrition of the tomato, found that the effects of nitrogen are governed by the levels of phosphorus in the plant. His experiments also showed that there is a tendency for the harmful effects of high nitrates in the growth me­ dium to be offset by the presence of a high phosphorus level. FIG. 5.— The effect of Increased phosphorus levels In the.presence of adequate potassium and exces­ sive nitrogen. Treatments are: 1-^00^0^30> 2-N 2O0p2i K30» 3-N 200p 5k 30> 4,i"N200p10K30» 5-N 200p 25K30* A marked NP interaction was also present at low phos­ phorus levels. At Pq the phosphorus deficiency symptoms occurred earlier and became more severe as nitrogen was in­ creased. The plants in Pig. 6 present a good example of the inverse relationship of these two elements. The potassium FIG. 6.— The effect of increased nitrogen levels in the absence of adequate phosphorus and potas slum. Treatments ares 1-NoPoKo> 2-N25PoK0> 3-N50P0K0, 4"N 100i>0K0 » 5-NgooPoK0« levels had no apparent influence on this relationship. As the experiment progressed the plants that received N q Pq lost their purple color, developed normally to the limit of their available nitrogen supply, and produced a single small fruit on each plant. These plants were essentially the same size as the others within the N q group. In this case nitrogen was the most limiting growth factor. Where moderate amounts of nitrogen were applied with­ out the addition of phosphorus the plants eventually lost their purple leaf colorations, made more growth than where nitrogen was lacking, and produced a few fruits. However, where the nitrogen level was kept at 100 or 200 ppm, in the presence of very low available phosphorus, the plants contin­ ued to show extreme phosphorus deficiency symptoms throughout the study. The five plants within the N 200P0 group made little growth and died before the experiment was completed. Pig, 7 illustrates these relationships. PIG. 7.— The effect of increased nitrogen levels in the absence of adequate phosphorus. Treat­ ments are: 1-Nq Po k 60» 2-“n 25?0k 60* 3-N5q P()K60» 4-N100PoK60> S-NgooPoKeO* This picture was taken two months later than the one in Pig. 6. Statistical analyses of dry weight and fruit yield data showed that the NP interaction was highly significant in both groups of data. A segregation of the 16 degrees of freedom of each of these interactions is shown in Tables 7 and 8 respectively. The combinations of levels which caused the significant NP interaction are indicated in these tables. With the combinations designated as significant it can be concluded that the plants responded differently to nitrogen levels in the presence of the phosphorus levels concerned. In considering these interactions one must be cautioned that these data represent the results of only one crop of tomatoes grown on one particular soil type with extreme treatments applied. Nevertheless, they do show the harmful effect of excessive nitrogen, especially when it occurs in the absence of adequate phosphorus. Fertilizing with potassium resulted in significant increases in both dry weight and fruit yield. During the early stages of growth, those plants which received K45 were la.rger than those at any other potassium level, as indicated in Fig. 8 . It can be seen from Tables 5 and 6 and Fig. 9 that nearly all of the variance with both factors at time of harvest was between Ko and the average of the other four levels. As the fruit yield data show, potassium levels greater than K15 were not effective in causing further in­ creases In yield. The dry weights were greatly increased with the addition of the first increment of potassium, and 34 TABLE 7.— A statistical analysis of NP interaction of dry weight of tomato plants grown on Oshtemo loamy sand. Source NP D.P. S.S. 16 13,736.8 1 No-N25-N50“N100-N 20cj g Po-P2i“P5-PlO-P25] 1 Mean square 858.6** 7,284.7 7,284.7** 0.01 0.01 " g P2i-P5-Pl0“p2l 1 n d P5-P10-P21 1 10.1 10.1 1 2a4 2.4 " gl0-p2g g N25-N50-N10o -N20o |5 p0“p2^"p5-p10“:p2i 1 2,502.7 2,502.7** g p2|-p5-p10"p2| 1 47.5 47.5 g P5-PlO-P2i 1 179.2 179.2 1P10_P2^ 1 662.7 662.7* 1 371.3 371.3 g N50-Nl00-N20cj @ g0“p2|-P5“p10“p2g " g P2i-p5-p10-p2l 1 51.4 51.4 " g P5_P10“P2i 1 32.9 32.9 glO“P2^ 1,653.8 1,653.8** 1 169.3 169.3 g p2i-P5“P10-P2| 1 307.2 307.2 g P5"P10”P2d 1 360.2 360.2 1 101.3 101.3 JFl00“N20^ 5 p0-p2i“p5-p10'p25l " 1 P10-P23 ■ft Significant at 5% point. ## Significant at 1% point. 35 TABLE 8.— A statistical analysis of NP Interaction of fruit yield of tomato plants grown on Oshtemo loamy sand. Source D.F. 16 NP H N0-Ng5-N50-N10O“N2o3 5 p0“p2&“p5-p10-?2i 349,880.6 349,880.6** 10.880.5 [2 P5-pl0-p2| 1 7.052.1 7.052.1 1 4.318.9 4.318.9 1 141.9 141.9 [3 p2i-p5-p10-p2l 1 4.062.2 4.062.2 (2 p 5-p10~p25| 1 33.458.5 33.458.5 1 3.042.1 3.042.1 1 91.259.3 91,259,3**1 I? P2i”p5’P10“P23 1 24.948.4 24,948.4 H p5-p10-p 2i 1 35,960.9 35,960.9 tp10"p25] 1 46,305.7 46,303.7 1 59,426.3 59,426.3* |3 p2^-p5-p10-p2| 1 6,736.5 6,736.5 I? p5“p10”p25l 1 3.735.1 3.735.1 1 9.107.9 9.107.9 5 N5o-N10o-^203 5 P0"P2-|“P5"P10“P25] ?100“N2o 3 I? P0”P2-|-“P5“P10-P25] ” ■ 43,144.7** 10.880.5 (?10-p2| ” 690,314.9 1 (3 N 25-N5o -Ni o o -N20^ 5 p0“p2i-p5“p10-p2^ '' Mean square (? p 2i - p 5-p 1 0 - p 2 i gl0-p2i » 1 S.S. lPlO-p2S # Significant at b% point. Significant at 1% point. 36 FIG. 8 .— The effect of increased potassium levels in the presence of adequate nitrogen and phos­ phorus. Treatments ares 1-N25P2^Kq, 3”N25P2l?K30» 4-N25P2iK45> 5-N25P2iE60* there was some indication of further “benefit from the second increment. Amounts above this level resulted in no further increase in growth. The plants tolerated the two higher levels but they did not require them. Deficiency symptoms of potassium occurred only on those plants that did not receive potassium fertilizer. These symptoms appeared within three weeks after transplant­ ing, Plants which showed potassium starvation were drawfed in appearance, had unusually short internodes, and were yellow at the tips and margins of the leaflets. These 37 (1) Dry weight (2) l/lO of fruit yield 60 50 40 30 20 10 0 15 30 45 POTASSIUM LEVEL (PPM IN SOIL EXTRACT) 60 FIG. 9.— Effect of potassium levels on the average dry weight and fruit yield of tomatoes. deficiency symptoms started on the older leaflets and pro­ gressed to the younger ones as the deficiency became more severe. When potassium starvation occurred at high nitro­ gen levels, white spots were found in conjunction with the marginal yellowing of the leaflets. These spots were similar to those described by Cook and Millar (8 ) as char­ acteristic of potassium starvation on clover and alfalfa. The condition is apparently caused by a low potassium and high nitrogen relationship since legumes are generally high in nitrogen and this disorder occurred on the tomatoes only in the presence of high nitrogen. Symptoms of potassium deficiency on older plants are considerably different from those on young plants. A normal mature plant that becomes deficient in potassium shows a yellow mottling of the interveinous tissue on the older leaflets, while the veins remain green. at the edges and progresses Inwardly, The mottling starts This is illustrated in Pig. 10. There were highly significant potassium interactions with both nitrogen and phosphorus in this study. It was observed that potassium deficiency symptoms occurred earlier and with greater severity on plants with high phosphorus and (or) with high nitrogen. The plant which showed the most severe symptoms of starvation received the NgogPggKg treatment. The phosphorus-potassium relationship is shown in Pig. 11. Statistical breakdowns of the PK Interactions for, dry 59 FIG, 10,— Tomato leaflets showing potassium defi­ ciency symptoms at different stages of growth. The leaflets on the left developed the symptoms at fruiting time. In the center is a younger one from the same plant. On the right is an older leaflet from a young plant. FIG, 11, — The effect of increased phosphorus levels on the severity of potassium deficiency symptoms. Treatments are? l-N25PoK0> 2-N 25p2iK0> 3“N25P5K0> 4 “n 25p 10k 0» S-NgsPssKo* weight and. fruit yield into their individual degrees of freedom are presented in Tables 9 and 10 respectively. With both groups of data it can be seen that this interaction was significant at the 1 % point only when the extreme levels were involved. Statistical analyses of the NK interactions showing a breakdown into their individual degrees of freedom for dry weight and fruit yields are presented in Tables 11 and 12 respectively. These data show the combinations of treatments which resulted in significant interactions. With both factors the significant interactions occurred where K q was involved. In other words, the plants responded differently to nitrogen levels at Ko than they did at higher levels of potassium. The data from this experiment emphasize the impor­ tance of maintaining proper nutrient balance in the soil for maximum production of tomatoes. Low nutrient levels which were well balanced gave much better results than did high levels which were unbalanced in one or more elements. For example, high nitrogen in the absence of adequate phosphorus resulted in death of the plants, yet when both elements were low they- grew fairly well and produced a few fruit. Another unbalanced condition was observed within the No group when both phosphorus and potassium were high. Plants which received NqKoq with Intermediate phosphorus levels showed no toxicity symptoms nor did those which received NqP25 with 41 TABLE 9.—-A statistical analysis of PK interaction of dry weight of tomato plants grown on Oshtemo loamy sand. Source S.S. 16 4,733.2 295.8** 1 3,333.4 3,333.4** 1? K15“K30"K45’K63 1 252.1 1 K3o-K45-K63 1 0.03 0.03 1 4.5 4.5 1 48.8 48.8 1 K15-K3o -K45"K6^ 1 70.9 70.9 1 K30~K45"K60l 1 7.5 7.5 1 17.6 17.6 1 36.5 36.5 § K15-K30-K45-K6^ 1 16.0 16.0 1? K30-K4 5-K65j 1 325.4 325.4 1 35.3 35.3 1 7.2 7.2 PK 1? P0-p2&-p5“PlCrP25] |? K0“K15“K30"K45~K60| " '* |?45“K60] |? p2i“p5-p10-p2l 5 Ko-KlS-KsO-^S-Keo] " P45-K6q 1 H p5-p10“p2| |§ K0-K15-K3CTK45-K60] “ " Mean square D.P. (K45-K6(3 [p10“p2l 5 Ko-Ki5-K3o-K45-K6(2 252.1 " 1 KX5-K50-K45-K6Q] 1 208.0 208.0; " |2 K30>K45-K6^| 1 273.1 273.1 1 96.8 CO o CO Oi " f45”*K60| •» Significant at 5% point. Significant at 1% point. TABLE 10.— A statistical analysis of PK interaction of fruit yield of tomato plants grown on Oshtemo loamy sand. Source D.P. 16 PK & p0“p2i-p5-p10-p25] H K0-Ki5-K30-K45-K60l 1 S.S. 647,019.6 Mean square 40,438.7*« 409,409.1 409,409.1** " [3 K15-K30-K45-K6Q] 1 7,937.2 7,937.2 " 1? K30-K45-K6q ) 1 1,438.1 1,438.1 1 470.3 470.3 1 23,686.5 23,686.5 [3 K15-K30-K45-K6Q] 1 17,152.2 17,152.2 [? K30-K45-K6q ] 1 5,292.3 5,292.3 1 7,202.9 7,202.9 1 51,884.7 51,884.7* " g45-K6«3 [3 P2i“p 5~p10“p25] I Ko-Ki5-K3o-K45-K6oj " l?45-K6Ql H p5"p10“p2| i Ko-Ki5-K3o-K45-K6o] " § Ki 5-K3o -K45-K6q ] 1 67,889.4 67,889.4* ”■ I? K3o-K45-K6(5] 1 17,928.1 17,928.1 17,978.2 17,978.2 1 717.8 717.8 n {K45-K60l [?10~P25| 5 ^ “K15”K30"K45“K60] ” [? K15**K30“K45"K60l 1 3,393.1 3,393.1 " (2 K30-K45-K6Q] 1 11,032.4 11,032.4 1 3,607.3 3,607.3 " . C^S-KSO] * Significant at 5% point. Significant at 1% point. 43 TABLE 11o— A statistical analysis of NK interaction of dry weight of tomato plants grown on Oshtemo loamy sand. Source NK D.F. S *S » 16 12,963.4 B N0-N25-N5crN;L0CrN203 5 KO-KIS-KSO-^S-KSO] 165.0 165.0 I k 30-k 45-k 6o ] 1 12.3 12.3 |?45“K60l 1 8.0 8.0 1 307.1 307.1 g K15-K30-K45-K65] 1 337.6 337.6 g K3o -K45-K6(3 1 572.5 572.5* 1 16.1 16.1 (3 N 25-N50“N10o -N20 o 1@ Ko«-Ki5-K3o-K45-K6(2 (?45-K651 1 (2 N50-Nioo-N20Qj B 3,635.9 3,635.9** g Kl5-K30-K45~K6a 1 590.3 590.3# g k30- k 45- k 651 1 32.1 32.1 1 707.3 707.3* i 45-K65] 1,562.4 1,562.4## [F10O-N20Q] B K o -K ig -K so -^ s-K g o l " 4,196.3 4,196.3** 1 " " 810.2*# (? K15-K30-K45-K60] " " 1 Mean square 806.0* g K i 5-K30-K45-K6q ] 1 806.0 g K3Q-K45-K6Q] 1 11.3 11.3 1 3.2 3.2 [K45-K6QI # Significant at b% point. ** Significant at 1% point. 44 TABLE 12.— A statistical analysis of: NK interaction of fruit yield of tomato plants grown on Oshtemo loamy-sand. Source D.F. 16 NK (4 N0-N25-N5O-N10O-N200)§ Ko-El5“K30-K45-K60l g Kl5-K30-K45-K65| [K45“K 6<3 g N25-N50-Nl00”N200| (I Ko“Ki5-K30-K45"K6o| " g " |? K30”K45”K60l " g k 15"K 30~k 45-k601 p45“K60l n 50"N100“n 20 q 15 K o -K15-K3o -K45-K65] 855,183.8 Mean square 53,449.0** 287,220.5 287,220.5** 1 4,915.5 4,915.5 1 - 4,325.0 4,325.0 1 874.9 874.9 1 201,836.5 201,836.5** 1 180.9 180.9 1 34,556.8 34,556.8 1 30,175.2 30,175.2 1 160,563.6 160,563.6** " g K15“K30"K45“K60l 1 91.2 91.2 " g K30-K45-K6Q] 1 7,695.0 7,695.0 1 2,317.6 2,317.6 1 42,457.0 42,457.0 1 3.1 3.1 1 24,688.9 24,688.9 1 53,282.2 53,282.2* (K4 5-K6^ [?100“N20o ]5 K0”K15~K30“K45“K60] g K30-K45-K63 " 1 S.S. g45-K60] Significant at 5% point. Significant" at 1% point. intermediate potassium levels. But the plant which re­ ceived N0P25K60 showed definite symptoms of nutritional unbalance as can be seen in Fig. 12, jar 4. When these FIG. 12.— The effect of high phosphorus and potas­ sium levels both with and without nitrogen. Treatments are: I-NqPoKo* 2-N0P25KQ, 3“NqPqK6o> 4-N0P25k60» 5- n 25p25k6 0 high phosphorus and potassium levels were In proper bal­ ance with nitrogen, the plants grew very well. A close- up of one of the lower leaves from the plant which showed the unbalanced condition is presented in Fig. 1-3. The margins of the most severly affected leaflets were yellow and necrotic, symptoms which are similar to potassium 46 PIG. 13.— A tomato leaf, from the plant which received Nq P25K60> showing symp­ toms of nutritional unbalance. deficiency. Tissue tests verified the fact that the plant was extremely high in potassium. This suggests that foliar toxicity symptoms of an element can easily be mistaken for deficiency symptoms of that or of some other element unless tissue tests are used in diagnosing nutritional disorders. The results of this experiment indicate that for greenhouse tomatoes, grown as a spring crop, the nitrogen level should be maintained between 25 and 50 ppm In the 47 soil extract, phosphorus between 5 and 10 ppm, and potassium between 15 and 30 ppm. In the presence of a continuous supply of nitrogen the phosphorus and potassium levels can exceed these limits considerably without injury to the plants. However, levels above those suggested result in luxury consumption and have no particular advantage. Nitrogen, Phosphorus, and Potassium Levels on Brookston Silt Loam Experimentalt In order to get further information pertaining to the nutrition of the tomato a 4x4x4 factorial experiment was conducted using Brookston silt loam soil. The treatments were duplicated. The general procedures followed were the same as in the previous experiment, although the nutrient levels were different. Nutrient levels in the soil extract were maintained at 0, 50, 100, and 200 ppm of NO3 , 0, 10, 20, and 40 ppm of phosphorus, and 0, 25, 50, and 100 ppm of potassium. sented the untreated soil. The zero level repre­ However, the untreated Brookston soil, tested at the time of setting up the experiment, had 75 ppm of NO3 , 3/4 ppm of phosphorus, and 5 ppm of potas­ sium. A general treatment was applied to all soils. This included manganese sulphate equivalent to 100 pounds per acre, magnesium sulphate equivalent to 500 pounds per acre, and sodium tetraborate (NagB^^) equivalent to 10 pounds per acre. A summary of all nitrogen, phosphorus, and potassium applications is shown in Table 13. Treatments were mixed evenly throughout the soil before the distilled water was added. Tomato plants of the Improved Bay State variety were started in flats four weeks before the beginning of the experiment. A sandy loam soil of medium to low fertility was used in the flats. The transplanting was done on August 10, 1948. At that time the plants were of uniform size and showed no visual deficiency symptoms, llae roots were washed free of all soil "before they were transplanted. TABLE 13.— A summary of the nitrogen, phosphorus, and potassium treatments applied to Brookston silt loam. Nutrient level n 50 N100 n200 *0 *10 ?20 p40 K0 k 25 K50 K ioo Gm. added per jar Pounds per acre equivalent NaNC>3 0.00 0• 0 N0 C.P. reagent used tt 0.00 0.0 n 1.14 285.0 n 5.05 1262.5 Ca(H2P04)2.H20 0.00 0.0 H 10.67 2667.5 II 20.27 5067.5 II 37.33 9332.5 0.00 0.0 II 11.94 2985.0 II 19.99 4-997.5 II 38.37 9592.5 KgS04 The soils were tested frequently and nutrients were added as needed to maintain the original levels of nitrogen, phosphorus, and potassium. When the entire crop was harvested on November 7, 1948 the first fruits were just beginning to ripen. Fruit yields recorded included all fruit on the plants at harvest time* At this time the seventh to sixteenth leaves, Inclu­ sive, from the base of the plants were collected to repre­ sent the leaf samples. These samples were later analyzed in the laboratory for nitrogen, phosphorus, and potassium* Results and Discussion: The average results, by treatment, for dry weight, yield of fruit, and per cent composition of nitrogen, phosphorus, and potassium in the leaf samples Is presented in Table 14. Dry Weight: Nitrogen, phosphorus, and potassium each exerted a highly significant Influence on the dry weight of the plants as the summary of analysis of variance in Table 15 shows. Fig. 14 as well as Table 15 Indicates that the N200 level resulted In the greatest growth under the condi­ tions of this experiment. It Is of interest to compare the effect of the heavy nitrogen treatment on this soil with its effect on the Oshtemo loamy sand. On the Brookston silt loam, plants which received the N200 treatment showed toxicity symptoms soon after its application. During the early stages of development growth was Inhibited to some extent and the plants possessed the soft, succulent, slightly wilted appear­ ance which was described as nitrogen toxicity in the pre­ vious experiment. The condition was less severe on the Brookston soil, however, and the plants soon overcame the toxicity, while the plants on the Oshtemo soil never fully 51 TABLE 14#— The average result by treatment for dry weight, yield of fruit, and per cent composition of nitrogen, phosphorus, and potassium in the leaf samples. Treatment Yield of fruit per plant(gm.) Dry weight per plant (gm.) Chemical composition of leaf samples# % N % P % K N0P0Ko 231.0 22.5 0.99 0.15 2.15 N0P0K25 180.0 25.5 0.97 0.15 3.70 n 0P0K50 207.5 24.0 1.01 0.16 3.98 N0P0K100 314.5 24.0 1.06 0.14 5.05 N0P10K0 288.0 28.5 1.18 0.51 2.33 N0P10K25 291.5 32.5 0.95 0.27 3.68 N0P10K 50 261.0 29.5 1.02 0.32 3.98 N0P10K100 329.5 35.0 1.11 0.25 4.78 n0P20k0 247.0 31.5 0.98 0.52 1.55 n 0 p 20k25 170.5 25.5 0.78 0.41 2.93 n 0 p 20K50 277.5 35.5 1.00 0.29 3.65 221.0 25.5 1.06 0.34 4.65 279.0 30.0 1.08 0.53 2.00 n 0p 40k25 324.0 26.0 1.01 0.37 3.58 n 0P40k50 285.0 28,0 0.94 0.38 3.65 n 0 p 40k100 280.0 26.0 1.18 0.39 4.58 20^100 »0p40*b # Percentage is based on oven dry weight. i 52 8 TABLE 14.-— (continued) The average result by treatment for dry weight, yield of fruit, and per cent composition of nitrogen, phosphorus, and potassium in the leaf samples. Treatment Yield of fruit per plant(gm.) Dry weight per plant (gm.) Chemical composition of leaf samples^ % N % P 1 % K N50P0K0 446.5 60.0 1.97 0.16 0.83 n 50p 0k25 575.0 54.5 1.89 0.15 2.85 n 50p 0k50 488.5 53.5 2.43 0.15 3.90 483.0 63.0 2.41 0.16 4.15 N 50p 10e0 662.0 52.5 2.11 0.53 0.65 N50P10E25 560.0 70.0 1.75 0.26 3.50 n 50p 10e 50 616.5 54.0 1.96 0.26 3.00 n 50P10e100 646.5 53.0 1.82 0.23 3.80 N50P20K0 556.0 51.0 2.10 0.88 0.65 n 50p 20k25 691.5 66.0 1.82 0.36 3.00 517.0 66.0 1.80 0.29 3.40 n 50P20e100 644.0 49.0 1.75 0.29 3.83 n 50p 40k0 603.0 55.5 2.06 0.82 0.45 n 50p 40e25 655.5 60.0 1.89 0.38 3.15 n 50P40e50 676.5 58.5 2.01 0.42 3.15 n 50p 40k100 584.5 65.0 2.39 0.32 3.93 n 50P0K100 N50P20K 50 v, Percentage is based on oven dry weight < 53 TABLE 14.— (continued) The average result by treatment for dry weight, yield of fruit, and per cent composition of nitrogen, phosphorus, and potassium in the leaf samples. Treatment Yield of fruit per plant(gm.) Dry weight per plant (gM»)___ Chemical composition of leaf samples# % n I i P T i k ~ NlOO^oKo 498.0 41.0 2.63 0.16 1.13 N100P0K25 536.0 54.5 2.44 0.15 2.23 n 100p 0k50 618.5 50.0 2.42 0.14 2.53 N100p0K100 593.0 52.0 2.38 0.15 3.28 W100P10K0 527.0 56.0 2,53 0.53 0.48 N100P10K25 553.0 63.0 2.31 0.31 2.78 W100P10K50 611.5 66.0 2.36 0.29 3.40 N100P10K100 615.5 60.5 2.37 0.29 3.43 N100P20k0 639.0 53.0 2.44 0.73 0.45 N100P20K25 653.0 53.0 2.38 0.34 2.98 n 100p 20k 50 620.0 62.0 2.31 0,32 2.80 n100p20k100 520.5 67.5 2.56 0.30 3.85 n 100P40K0 613.0 51.5 2.50 0.76 0.55 N100P40K25 650.0 63.0 2.35 0.40 2.35 N100P40K50 659.0 58.0 2.29 0.38 2.85 n 100P40k100 749.5 58.5 2.34 0.33 3.35 • # Percentage is based on oven dry weight. 54 TABLE 14.-— (continued) The average i*eault by treatment for dry weight, yield of fruit, and per cent composition of nitrogen, phosphorus, and potassium in the leaf samples. Treatment Yield of fruit per plant(gm.) Dry weight per plant (gm.) Chemical composition of leaf samples* % N % P 1 % E ^ O o P o 15*) 383.0 53.0 3.00 0.17 1.30 N200P0K25 450.0 54.0 2.90 0.17 2.25 n 200p 0K50 447.0 53.0 2.75 0.15 2.80 n200P0K100 475.0 54.0 2.85 0.15 3.70 ^ o o P i o 2^ 614.5 58.0 2.74 0.60 0.65 n 20C)P10k25 490.0 71.0 2.62 0.29 2.65 n 200P10k50 539.5 72.5 2.62 0.27 3.08 N20C)Pl0K100 657.0 66.0 2.66 0.25 3.80 N 200P20K0 593.0 56, 5 2.81 0.65 0.85 N200P20K25 581.0 76.0 2.58 0.34 3.00 n 200p 20k 50 543.5 70.5 3.11 0.35 3.00 w200p20K100 554.0 72.0 2.65 0.32 3.85 n 200p 40k0 604.0 53.0 2.63 0.64 0.60 N200P40K25 707.0 69.0 2.61 0.35 2.60 N200P40K50 597.5 69.0 2.71 0.37 3.03 N 200P40K100 550.5 69.0 2.85 0.40 3.73 * Percentage is based on oven dry weight. 55 TABLE 15.— Analysis of variance of the dry weight data of tomato plants grown on Brookston silt loam. Mean square D.P. S.S. 127 32,450.2 63 29,781.2 472.7** 3 24,518.8 8,172.9** 3 N q vs. NgQ+NiQO+^OO 1 23,719.6 23,719.6** 2 N 50 vs. NioO+n200 1 82.7 1 715.6 715.6** 3 1,350.5 450.2** 3 P q vs. Pl0+E>20+p40 1 1,298.0 1,298.0** 2 P1Q vs* P20+P40 1 26.3 26.3 1 26.3 26.3 3 929.0 309.7** 3 Kq vs. Kgs+Kso+K^OO 1 894.3 894.3** 2 K25 vs. K50+KI00 1 28.5 28.5 1 6,3 6.3 NP 9 693,7 77.1 NK 9 569.4 63.3 PK 9 337.7 37.5 27 1,382.1 51.2 1 16.5 16.5 63 2,652.5 42.1 Source Total Treatment N levels N100 vs» N200 P levels p20 vs* P40 K levels K50 vs. K10o NPK Between replicates Within replicates (error) * Significant at b% point. ■»# Significant at 1% point. 82.7 56 (1 ) (2 ) (3) (4) (5) l/lO of dry weight in grams 1/100 of fruit yield in grams 10 times per cent phosphorus Per cent potassium Per cent nitrogen Jll 5.0 4.0 3.0 2.0 1.0 0 100 50 NITROGEN LEVEL (PPM N03 IN SOIL EXTRACT) 200 FIG. 14.— Effect of nitrogen levels on the average dry weight, fruit yield, and per cent nitrogen, phosphorus, and potas­ sium in the leaf samples. recovered. Eventually, those grown on Brookston soil at this high level of nitrogen made more growth than did those that received less nitrogen. The Brookston silt loam used in this experiment was higher in organic matter and had a much higher base exchange capacity than the Oshtemo loamy sand. These properties gave the Brookston soil a much greater buffering capacity which apparently accounted for the inconsistent results between the two experiments. Symptoms of nitrogen starvation appeared seven weeks after transplanting on all plants which did not, receive nitrogen. These symptoms were the same as those described in the previous experiment. No phosphorus or potassium deficiency symptoms occurred during the experiment. Growth response to phosphorus treatments is shown in Table 15 and Pig. 15. They show that the only significant difference between levels was between none and some phos­ phorus. The second level, Pio» was fully as effective as the heavier applications. There was no indication that the highest phosphorus level, which was equivalent to over 9300 pounds per acre of G.P. primary monocalcium phosphate, was detrimental to the growth of the plants. Potassium treatments caused results similar to those obtained for phosphorus as is apparent in Table 15 and Pig* 16. It is shown that Kgg resulted in a significant response in growth over no potassium and that there was neither pos­ itive nor negative growth response to additional applications 58 (1 ) (2) (5) (4) (5) 1/10 of dry weight in grams l/lOO of fruit yield in grams 10 times per cent phosphorus Per cent potassium Per cent nitrogen 6.0 (2 ) 5.0 4.0 3.0 (4) 2.0 i 1.0 0 20 10 PHOSPHORUS LEVEL (PPM IN SOIL EXTRACT) 40 FIG. 15.— Effect of phosphorus levels on the average dry weight, fruit yield, and per cent nitrogen, phosphorus, and potassium in the leaf samples. 59 (1) (2) (3) (4) (5) l/lO of dry weight in grams l/lOO of fruit yieldin grams 10 times per cent phosphorus Per cent potassium Per cent nitrogen 6.0 (2) 5.0 4.0 3.0 (3) 2.0 1.0 0 50 25 POTASSIUM LEVEL (PPM IN SOIL EXTRACT) 100 FIG. 16.— Effect of potassium levels on the average dry weight, fruit yield, and per cent nitrogen, phosphorus, and potassium in the leaf samples. of this element. Fruit Yields High fruit yields were obtained only from large, healthy plants. It is apparent from the data in Table 16 and Fig. 14 that Ng0 and N^oo resulted in significantly higher fruit yields than did No or N200 even though N200 resulted in the largest plants. The decreased fruit yield at Ng0Q was evidently due to an unbalanced ratio of carbohydrate to available nitrogen. Since this crop was grown in the fall, during a period of low light intensity and short days, the explanation is applicable. According to Eraus and Kraybill (19) excessive nitrogen in the presence of low carbohydrates results in low fruit production of the tomato. The importance of an adequate supply of available phosphorus is indicated in Table 16 and Fig. 15. A response in fruit yield which was significant at the 1% point was obtained by the addition of phosphorus fertilizer. The P^q and P20 levels were equally effective in promoting fruit production. Even though the yields between P20 anetween treatment means required for significance at the 5% point. 00 TO 83 of NagB407 was at the rate of 200 pounds per acre was not significant. Judging from green weight yields, applications of NagB40y could be as great as 50 pounds per acre, but from fruit yields it is evident that the limit should be placed at 10 pounds per acre. Foliar symptoms indicating nutritional disorders appeared on plants which did not receive boron but the defi­ ciency was not severe enough to cause death of the terminal growing points as described by Johnston and Dore (13). Boron starved plants in this experiment showed a yellowing of the leaflet tips with the affected area later dying. In advanced stages of the deficiency the entire area of the lower leaflets became yellow and, as the marginal tissue be­ came necrotic, the edges of the leaflets curled upward. The disorder appeared on the older leaflets first and progressed to the younger ones. Toxicity symptoms occurred on all plants which re­ ceived 50 pounds per acre or more of NagB4C>7. The symptoms were slight on the plants which received 50 pounds per acre, but were severe where 200 pounds was applied. Boron toxicity symptoms appeared as yellowish-brown areas at the tips and along the margins of the older leaf­ lets. The tissue in these areas later died resulting in brown necrotic regions. These symptoms were similar to those described by Johnston and Dore (13) and Johnston and Fisher (14). They were also similar to boron toxicity symptoms of soybeans, as described by Muhr (23). Chemical analyses of the soil and plant tissue samples revealed that in both the boron concentration in­ creased with applications of boron. As shown in Table 22 the water soluble boron in the soil increased from 0,14 ppm where boron was not applied to 9,46 ppm where 200 pounds per acre of NagB^Oy was applied. Likewise, the concentration of total boron in the tissue samples increased from 8.0 ppm to 249.9 ppm as a result of the 200 pounds per acre treatment. The concentrations of nitrogen, phosphorus, and potas­ sium in the tissue samples were each significantly affected by the boron treatments. The percentage of each element varied inversely with growth and fruit yield. There was a tendency for them to accumulate where boron was either defi­ cient or present in toxic concentrations. Where growth and fruit yields were highest the concentration of each of the three elements was at a minimum. The percentages of calcium and magnesium in the tissue samples were not affected by the boron treatments. A.s is shown in Table 22, the calcium-boron ratio of the samples decreased from 2175 where boron was not applied to 77 where 200 pounds per acre of NagB^Oy was added. Nor­ mal plants, which received the 10 and 25 pound per acre applications, had a calcium-boron ratio of 645 and 335 re­ spectively. Those plants grown with the 50 pound treatment showed slight toxicity symptoms and had a calcium-boron ratio of 242. These values are in fair agreement with those obtained by Brennan and Shive (6). They investigated the calcium-boron relationship in tomato plants and found that the ratio of these elements in the leaves of normal plants ranged from 201 to 593. They also observed that any great deviation from this range resulted in an unbalanced condi­ tion which caused physiological abnormalities in the plants. Values much above 593 were associated with boron deficiency symptoms while those much below 201 were associated with boron toxicity symptoms. Response to Boron Treatments on Thomas Loamy Sand, Miami Loam, and Yfisner Sandy Loam Experimental: This experiment was similar to the preceding one but was expanded to include three soil types, Thomas loamy sand, Miami loam, and Wisner sandy loam. These soil types were selected because they ranged from neutral to alkaline in reaction and because s^^ch crops as sugar beets had shown favorable responses to applications of borax on them. For this experiment Improved Bay State tomatoes were grown. The seeds were planted in flats on July 22, 1948. When the plants were about four inches high they were trans­ planted following the same procedure as was used in the pre­ ceding boron experiment. The general treatment in this experiment was different from that used in the preceding experiment. The reagents applied to these soils included? Reagent Pounds per acre equivalent NH4N03 1325 K2S04 1325 Ca(HPC>4 )2*1*20 2500 MnS04»4H20 100 MgS04•7H20 500 Nitrogen was added as needed during the course of the experiment* The boron treatments consisted of Nagl^O^ at rates equivalent to 0, 10, 25, 50, and 100 pounds per acre. The treatments were in triplicate. Soil samples to be analyzed for boron were taken from each jar four weeks after transplanting. The experiment was terminated on November 7 at which time separate leaf samples were collected from each plant. The seventh to sixteenth leaves, inclusive, from the base of the plants made up the leaf samples. At the time of harvest the fruit were well, formed but none were ripe. Individual measurements and determinations were made on each plant and analysis of variance was run on fruit yield, dry weight of plants, and composition of the leaf samples. Results and Discussion? The average result by treat- ment for fruit yields, dry weight of plants, chemical comp­ osition of leaf samples, and boron content of the soil samples for each of the soils, Thomas loamy sand, Miami loam, and Wisner sandy loam, is shown in Tables 23, 24, and 25 respectively. It can be seen from these data that fruit yields were not affected by treatment on any of the soils. Likewise, the dry weight of the plants on the Thomas and Wisner soils was not affected by treatment. However, the plants on Miami soil were significantly inhibited in growth where 100 pounds per acre of ^ 264.07 was applied. As with the other two soils there was no appreciable increase in grow­ th due to boron applications on the Miami soil. Symptoms of boron starvation did not occur on any of the plants in this study. Toxicity symptoms occurred on all plants which received either of the two higher rates of Na2B40y. These symptoms were the same as those described for boron toxicity in the preceding experiment. As can be seen in Pig. 20, these symptoms were slight where 50 pounds per acre was applied but were severe where the application was 100 pounds per acre. It Is of interest that even though severe toxicities occurred at the highest rate of applica­ tion, fruit yields were not reduced and growth was inhibited only on the Miami losm. The water soluble boron in the soil and the total boron in the leaf samples increased with increased boron applications, as was true in the preceding experiment. The boron content of the fruit changed only slightly with treat­ ment. Fruit of normal plants contained between 11 and 15 TABLE 23.— Effect of boron treatments on the yield and chemical composition of tomato plants grown on Thomas loamy sand. Na?B407 added fibs./acre) Ave. dry wt.fern.) 0 548.7 49.3 2.06 69.2 11.5 497 2.34 0.57 1.03 3.44 0.90 10 629.3 46.7 2.65 72.8 12.7 488 2.25 0.56 0.90 3.55 0.90 25 565.3 46.0 3.01 89.2 13.2 381 2.51 0.53 0.97 3.40 0.98 50 559.3 46.3 3.74 150.9 14.5 209 2.54 0.59 0.95 3.15 0.90 100 494.7 44.3 9.22 377.3 12.3 103 2.73 0.64 1.10 3.88 1.03 L.S.D.(2) n.s. n.s. 0.28 n.s. n. s. n. s. &• s# 0 Ave. boron content(ppm) soil leaves fruit Ca/B in leaves Ave. fruit vieldfem.) Ave. composition of leaf samples fl) * N % K 1 t Ca % Mg* P (1) Percentage is based on oven dry weight. (2) Least difference between treatment means required for significance at the point. oo oo TABLE 24.— Effect:of boron treatments on the yield and chemical composition of tomato plants grown on Miami loam. Ave. boron content(ppm) soil leaves fruit Ca/B in leaves 1.52 55.0 11.5 1167 2.88 0.59 1.25 5.85 0.99 42.0 1.75 59.1 10.9 755 2.74 0.68 1.55 4.55 1.10 267.0 49.5 2.27 62.4 14.2 615 2.60 0.65 1.52 5.84 0.95 50 575.0 58.5 5.47 124.5 21.0 556 2.76 0.76 1.40 4.45 1.05 100 570.7 54.7 5.76 257.9 19.4 185 2.94 0.71 1.85 4.77 1.08 L.S.D. (2) n.s. 7.6 0.15 n.s. 0.27 0.59 n.s. Ave. fruit yieldfgm.) Ave. dry wt.(gm.) . 558.5 42.0 10 555.5 25 Na«B.07 added (lbs./acre) 0 Ave. composition of leaf samples(.I)1% N % P % K % Ca % Mg (1) Percentage is based on oven dry weight. (2) Least difference between treatment means required for significance at the 5$ point. 00 to TABLE 25.-— Effect of boron treatments on the yield and chemical composition of tomato plants grown on Wisner sandy loam. NapB407 added fibs./acre) 1 Ave. fruit yieldfem.) Ave. dry wt.(gm.) Ave. boron content(ppm) soil leaves 1 fruit Ca/B in leaves Ave. composition of leaf samples(l) % N % K < Car: . % Mg % p 0 517.7 50.0 1.30 50.1 14.5 866 2.27 0.43 0.97 4.34 1.22 10 552.0 53.0 1.62 76.0 14.5 571 2.25 o»46 1.07 4.34 1.18 25 530.7 47.0 2.53 80.7 17.9 496 2.44 0.51 0.98 4.00 1.27 50 517.3 46.7 4.12 121.5 14.7 337 2.36 0.51 1.02 4.09 1.37 100 564.3 46.3 4.80 290.9 18.9 157 2.38 0.55 0.98 4.57 1.42 L.S.D. (2) n.s. n.s. n# S« 0.07 n.s. n.s. n.s. (1) Percentage is based on oven dry weight. . (2) Least difference between treatment means required for significance at the 5% point. co o 91 FIG. 20.— Leaves from tomato plants grown on Thomas loamy sand which received increased rates of Na2®4®7« The applications in pounds per acre were: 1-0, 2-10, 3-25, 4-50, 5-100. ppm of boron on a dry weight basis. The mineral composition of the leaf samples in this experiment was not influenced by treatment as much as it was where the plants were grown on Oshtemo loamy sand. The composition of plants grown on the Thomas loamy sand was not affected by the sodium tetraborate applications so far as phosphorus, potassium, calcium, and magnesium in the leaves was concerned but there was a slight nitrogen accumulation where the boron application was the highest. This is shown in Table 23. Similar results were obtained with the plants grown on Wisner sandy loam. As Table 25 shows, boron treatments did not affect the concentrations of nitrogen, potassium, calcium, or magnesium in the leaf samples, hut did cause a slight accumulation of phosphorus where the highest boron treatment was made, A greater variation in the composition of the leaf samples occurred in plants grown on the Miami loam than in those grown on the other two soils. In these samples the greatest boron application caused significant increases in per cent nitrogen, potassium, and calcium while the phos­ phorus and magnesium concentrations were not affected. This result is indicated in Table 24. Considering the data from both boron experiments, the percentage of nitrogen, phosphorus, and potassium in the leaf samples was markedly changed only where growth was in­ hibited, due either to a starvation for boron or to an ex­ cess of boron. Leaf samples from plants which developed normally were essentially the same in their composition of nitrogen, phosphorus, potassium, calcium, and magnesium regardless of boron treatment. At the extreme levels of boron, where growth was inhibited, there was a tendency for nitrogen, phosphorus, and potassium to accumulate in the leaves. Within the limits of the boron applications used in these experiments the calcium concentration in the leaves of the boron deficient plants was not increased. Its concen­ tration was increased by toxic amounts of boron only on the Miami loam. The percentage of magnesium in the leaves was not affected by any of the boron treatments. During the course of the experiment it was observed that toxicity symptoms were more severe on the Thomas loamy sand than on either the Miami loam or the Wisner sandy loam. Chemical analyses later showed that the plants grown on the Thomas soil contained less calcium and more boron than did those grown with the same boron treatments on the other two soils. The water solxible boron, also, was generally higher in the Thomas soil. These conditions resulted in lower calcium-boron ratios in the leaves of the plants grown on this soil. As a result more severe toxicity symptoms oc­ curred at the high boron levels on the Thomas loamy sand than on the other two types. These data show that the calcium-boron ratio in the leaves of boron deficient tomato plants was greater than 1167 and the ratio in the leaves which showed toxicity symptoms was approximately 375 and below. The soils studied were relatively high in calcium and tomato plants grown on them developed normally when the concentration of water soluble boron in the soil ranged from 0.50 to 2.75 ppm. When the concentration varied greatly in either direction from, this range deficiency or toxicity symptoms were noted. SUMMARY AND CONCLUSIONS The results of a study of the effects of nitrogen, phosphorus, potassium, and boron levels in the soil on the yield and chemical composition of greenhouse tomatoes are presented. Plants were grown in the greenhouse in pot culture. Levels of nitrogen, phosphorus, and potassium were based on parts per million in the soil extract according to the Spurway soil testing procedure. Two factorial experiments involving nitrogen, phos­ phorus, and potassium were conducted. In the first, the soil was Oshtemo loamy sand with' nitrate levels maintained at 0, 25, 50, 100, and 200 ppm, with phosphorus at 0, 2-gr, 5, 10, and 25 ppm, and with potassium at 0, 15, 30, 45, and. 60 ppm. In the second, the soil used was Brookston silt loam and nitrate levels were maintained at 0, 50, 100, and 200 ppm, phosphorus at 0 , 10, 20, and 40' ppm, and potassium at 0, 25, 50, and 100 ppm. In both experiments fruit yield and dry weight were obtained from each plant. Leaf samples were collected in the second experiment and were analyzed for total nitrogen, phosphorus, and potassium. Statistical analyses of these data were made. Differential varietal response to potass5_um fertiliz­ er was studied^ Plants of the Improved Bay State and Spar­ tan Hybrid varieties were grown at potassium levels of 0, 5, 10, 25, 50, and 100 ppm. Leaf and fruit samples were collected and analyzed for phosphorus, potassium, calcium, and magnesium. Tomatoes were grown on Oshtemo loamy sand, Thomas loamy sand, Miami loam, and Wisner sandy loam with boron treatments ranging from 0 to 200 pounds per acre of Minimum requirements as well as the tolerance range for the element were studied. The relationship of boron to the mineral composition of the leaves was investigated. The results of these experiments may be briefly summarized as follows: 1. A method devised for determining the fertilizer "fixing power" of soil proved useful in establishing and maintaining nutrient levels. 2. Highest yields of fruit resulted where the nitrogen level was maintained between 25 and 50 ppm, phosphorus be­ tween 5 and 10 ppm, and potassium between 15 and 30 ppm. 3. Nitrogen levels above 50 ppm in Oshtemo loamy sand resulted in a decrease in both fruit yield and dry weight. 4. Phosphorus and potassium levels as high as 25 and 60 ppm, respectively, on Oshtemo loamy sand and as high as 40 and 100 ppm, respectively, on Brookston silt loam were not toxic to the plants where the nitrogen supply was adequate. 5. Deficiency symptoms of nitrogen, phosphorus, and potassium were observed. With nitrogen deficiency the older leaflets were uniformly yellow. In advanced stages the leaflet veins were distinctly purple while the interveinous tissue remained a faded yellow. Phosphorus deficiency symptoms appeared as uniform purpling of the under side of the leaflets while the upper side became very dark green. Plants which were starved for potassium showed marginal yellowing of the younger leaflets and yellow mottling of the older ones. In the presence of soluble nitrogen character­ istic white spots accompanied the yellowing where the plants were starved for potassium. 6. In the presence of a low supply of available potassium, symptoms of potassium starvation appeared earlier and with greater severity as either or both nitrogen and phosphorus levels were increased. 7. In the presence of a lowavailable phosphorus supply, symptoms of phosphorus starvationwere progressively more severe with increased nitrogen levels. 8 . High phosphorus levels tended to counteract the harmful effects of excessive nitrogen. 9. The percentage of nitrogen, phosphorus, and potas­ sium in the dry leaf samples increased with each addition of their respective salts. 10. Low nitrogen levels favored the accumulation of potassium in the leaves., but had no effect on the phosphorus content, 11. Phosphorus levels were unrelated to the concentra­ tion of nitrogen and potassium in the leaves. 120 Low potassium levels favored the accumulation of phosphorus in the leaves,but had no effect on the nitrogen content. 13. High potassium levels caused a reduction in the percentage of calcium and magnesium in the leaves. 14. Improved Bay State tomatoes required a higher concentration of potassium in the leaves than did the Spartan Hybrid plants in order to prevent the occurrence of leaf yellowing due to potassium starvation. 15. Tomatoes showed a favorable response to boron when grown on overlimed Oshtemo loamy sand. 16. The concentration of boron in the plant tissues was directly related to the amount in the soil. 17. Boron treatment had a significant effect on the percentage of nitrogen, phosphorus, and potassium in the tissue samples only when it was low or high enovigh to cause restricted growth. accumulated. In these cases the three elements No relationship was found between boron treat­ ment and magnesium concentration. The percentage of calcium in the leaves was increased by boron applications in only one of the four trials. 18. Tomato plants developed normally when the concen­ tration of water soluble boron in the soils ranged from 0.50 to 2.75 ppm. When the concentration varied greatly from this range deficiency or toxicity symptoms were noted, 19. Applications equivalent to 50 pounds per acre or more of NagB^jOy resulted In boron toxicity symptoms on the plants. Literature Cited ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS. Official and tentative methods of analysis, Ed. 5. Washington, D.C. 1940. BATJER, L.P., and MAGNESS, J.R. Potassium content of leaves from commercial apple orchards. Proc. Amer. Soc. Hort. Sci., 36S197-801. 1938. BLAKE, M.A., NIGHTINGALE, G.T., and DAVIDSON, O.W. Nutri­ tion of apple trees. N.J. Agr. Exp. Sta. Bui. 626. 1937. BERGER, K.C., and TRUOG, E. Boron tests and determination for soils and plants. Soil Sci., 57:25-36. 1944. BOUYOUCOS, G.J. Directions for making mechanical analyses of soils by the hydrometer method. Soil Sci., 42:225-229. 1936. BRENNAN, E.G., and SHIVE, J.W. Effect of calcium and boron nutrition of the tomato on the relation between these elements In the tissues. Soil Sci., 66:65-75. 1948. CAROLUS, R.L. Tomato fertilization II. The effect of different fertilizer ratios on the chemical composition of tomatoes. Va. Truck Exp. Sta. Bui. 81. 1933. COOK, R.L., and MILLAR, C.E. Plant nutrient deficiencies. Mich. Agr. Exp. Sta. Spec. Bui. 353. 1949. EMMERT, E.M. Plant-tissue tests as a guide to fertilizer treatment of tomatoes. Ky. Agr. Exp. Sta. Bui. 430. 1942. FISKE, C.H., and SUBBAROW, Y. The colorimetric determina­ tion of phosphorus. Jour. Biol. Chem., 66:375-400. 1925. GREGORY, F.G., and CROV/THER, F.A physiological study of varietal differences in plants. Part I. A study of the comparative yields of barley varieties with different manurings. Ann. Bot., 42:757-770. 1928. — — ,----- — . A physiological study of varietal differ­ ences In plants. II. Further evidence for the differen­ tial response in yield of barley varieties to manurial deficiencies. Ann. Bot., 45:579-592. 1931. JOHNSTON, E.S., and DORE, W.H. The Influence of boron on the chemical composition and growth of the tomato plant. Plant Physiol., 4:31-62. 1929. 99 14. ------- and FISHER, P.L. The essential nature of boron to the growth and fruiting of the tomato. Plant Physiol., 5:385-392. 1930. 15. LAME, C.A., and SALTER, R.M. Response of wheat varieties to different fertility levels. Jour. Agr. Res„, 53:129143. 1936. 16. LAWTON, K. The determination of exchangeable potassium in soils using hexanitrodiphenylamine. Soil Sci. Soc. Amer. Proc., 10:126-128. 1945. 17. LUNDEGARDH, H. Leaf analysis as a guide to soil fertility. Nature, 151:310-311. 1943. 18. LYNESS, A.S. Varietal differences in the phosphorus feeding capacity of plants. Plant Physiol., 11:665-688. 1936. 19. KRAUS, E.J., and KRAYBILL, H.R. Vegetation and reproduc­ tion with special reference to the tomato, Ore. Agr. Exp. Sta. Bui. 149. 1918. 20. MacGILLIVRAY, J.H. Effect of phosphorus on the composi­ tion of the tomato plant. Jour. Agr. Res., 34:97-127. 1927. 21. MAGY, P. The quantitative mineral nutrient requirements of plants. Plant Physiol., 11:749-764. 1936. 22. McCALLA, A.G., and WOODFORD, E.K. Effects of a limiting element on the absorption of individual elements and on the anion:cation balance in wheat. Plant Physiol., 13: 695-712. 1938. 23. MUHR, G.R. Plant symptoms of boron deficiency and the -effects of borox on the yield and chemical composition of several crops. Soil Sci., 54:55-65. 1942. 24. PARKS,R.Q,. , LYON, C .B . and HOOD, S.L. Some effects of boron supply on the chemical composition of tomato leaf­ lets. Plant Physiol., 19:404-419. 1944. 25. PEEGH, M., and ENGLISH, L. Rapid microchemical soil tests. Soil Sci., 57:167-195. 1944. 26. ------- , et al. Methods of soil analysis for so51-fertility investigations. U.S.D.A. Cir. 757. 1947. 27. PURVIS, E.R., and HANNA, W .J. Boron studies: I. The suscep­ tibility of various plants to boron toxicity as influ­ enced by soil type. Soil Sci. Soc. Amer. Proc., 3:205209. 1938. < RALIEGH, G.J. Fruit abnormalities of tomatoes grown in various culture solutions. Proc. Amer. Soc. Hort. Sci., 37:895-900. 1939. SALTER, R.M* , and AMES, J.W. Plant composition as a guide to the availability of soil nutrients. Jour. Amer. Soc. Agron., 20 #808-836. 1928. SHEAR, C.B*, CRANE, H.L., and MYERS, A.T. Nutrient-element balances A fundamental concept In plant nutrition. Proc. Amer. Soc. Hort. Sci., 47:239-248. 1946. SNEDECOR, G.W. Statistical Methods. Ames, Iowa. The Col­ legiate Press. Ed. 4. 1946* SPUR1AY, C.H. Soil testing, a practical system of soil fertility diagnosis, Mich. Agr, Exp. Sta. Tech. Bui. 132. (3rd r e v . ) . 1944, ------ . Soil fertility diagnosis and control for field, garden, and greenhouse soils, (published by the author) Lithoprinted, by Edwards Bro., Inc. Ann Arbor, Mich. 1948. STRINGFIELD, G.H., and SALTER, R.M. Differential response of corn varieties to fertility levels and to seasons. Jour. Agr. Res., 49:991-1000. 1934. THOMAS, W, The reciprocal effects of nitrogen, phosphorus, and potassium, as related to the absorption of these ele­ ments by plants. Soil Sci., 33:1-20. 1932. ------ . Present status of diagnosis of mineral r e q u i r e ­ ments of plants by means of leaf analysis. Soil Sci., 59:353-374. 1945. ------ and MACK, W.B. Salient features of the method of foliar diagnosis. Proc. Amer. Soc, Hort. Sci. 37:253260. 1939. ------ ,------- • Foliar diagnosis of differentially ferti­ lized greenhouse tomatoes with and without manure. Jour. Agr, Res,, 60:811-832. 1940. ------ ,------- and COTTON, R.H. Nitrogen, phosphorus, and potassium nutrition of tomatoes at different levels of fertilizer application and of irrigation. Proc. Amer. Soc. Hort. Sci., 42:535-544, 1943. t end FAGAN, F.N. Foliar diagnosis: An ap­ proach to t h e control of the nutrition of aoole t r e e s . Proc. Amer. Soc. Hort, Sci., 47:97-106. 1946, ULRICH, A. Plant analysis as a diagnostic procedure. Soil Sci., 55:101-112. 1943,