Journal of Agricultural Science in Finland Maataloustieteellinen Aikakauskirja Vol 61:323—360 THE ROLE OF FERTILIZATION PRACTICES IN THE YIELD AND QUALITY OF CARROT (Daucus carota L.) Selostus: Lannoitusmenetelmien vaikutus porkkanan satoon ja sisäiseen laatuun AINO-MAIJA EVERS Kemira Oy, Espoo Research Centre Luoteisrinne 2, 02270 Espoo, Finland Academic Dissertation TO BE PRESENTED. WITH THE PERMISSION OF THE Faculty of Agriculture and Forestry of the University of Helsinki, for public criticism in Auditorium 12, Unioninkatu 34, Helsinki, on November nth 1989, at 12 o’clock noon SUOMEN MAATALOUSTIETEELLINEN SEURA • HELSINKI https://www.c-info.fi/en/info/?token=vPnmUETY6DWrDcfo.qINBolnUKgNS480a96KRRw.hn9d2glKcUTyFloVck7MTjsQlG-JezHBVo2XAIHhNnX1quCXvcFM9UBaHvuCwups3rsiT64BcVceHX7jVKZFJu-qS2aZHQdQOVoCCCbTHW3DVnqfVATrg9J1pt6DZv2-AgQ_AR6e3otEKXH92hRrlEWbxevldI1d0t92lj7d 325 Preface This research was carried out at Kemira Oy’s Espoo Research Centre during 1985—1989. I am deeply grateful to my teacher and supervisor, Professor Erkki Kau- kovirta, Head of the Department of Horticulture, University of Helsinki, for his sup- port and most valuable instructions and advice during every stage of this study. I am most grateful to Professor Eero Varis and Professor Antti Jaakkola for reading the manuscript and for their useful suggestions. I wish to express my warmest gratitude to Professor Anna-Liisa Varis for her kind support and guidance. I owe my deepest gratitude to Dr. Heikki Rosenqvist, Head to the Espoo Research Centre, and to Kemira Oy for providing excellent working facilities. The encouraging attitude of Dr. Ari Lokio, former Head of the Espoo Research Centre, decisively contributed to the start of my postgraduate studiesand to this thesis. 1 would like to express my sincere appreciation to him. My special thanks are due to Miss Oili Uusitalo for excellent technical assistance. Without her intelligence, great diligence and organizing ability the project would not have been carried out in this form. I am also deeply grateful to Mrs. Kirsti Karhunen for creating a friendly working environment and for smooth cooperation, to Marina Niemi, Lie. Phil., for her interest in my work, for encouragement and for construc- tive criticism of the manuscripts, and to Katri Tiittanen, M.Sc., for introducing me into scientific writing, and for warm interest in my work and never-failing support. I am grateful to Ulla and Uljas Lehtonen for providing me with two organically cultivated field areas and carrots grown on them. Our discussions were most stimu- lating, and they broadened my points of view. I also wish to thank all my collabora- tors, especially Mrs. Ann-Marie Korhonen for skilful typing and for her patience with the corrections, Markku Yli-Halla, Lie. Agr., for constructive criticism of the manuscript, Dr. Pauli Seppänen for valuable discussions, Börje Björkqvist, M. Sc., for advice concerning organic analyses, Veli-Matti Taavitsainen, Lie. Phil., for ad- vice concerning statistical analysis, and the personnel of the Kemira Espoo library for providing me with all the necessary information. I also thank Sevastiana Kuusa- mo, M.A., for revision of the English text of the articles and this thesis. My dearest thanks I owe to my husband likka for his encouraging attitude, lov- ing care and never-failing sympathy, and to my parents and brothers with families for their benevolent care and encouraging support of all my attempts throughout the years, which I greatly appreciate. I feel special respect and gratitude to my school teacher Väinö Voutilainen for his imposing native language lessons and for his last- ing impact on my academic career. This work was financially supported by the Academy of Finland and Kemira Oy, which is gratefully acknowledged. Finally, I would like to thank the Scientific Agricul- tural Society of Finland for including my study in their journal. Espoo, September 1989 Aino-Maija Evers 326 List of original articles The original articles summarized here are: 1 Effects of different fertilization practices on the growth, yield and dry matter content of carrot. J. Agric. Sci. Finl. 60: 135—152. II Effects of different fertilization practices on the carotene content of carrot. J. Agric. Sci. Finl. 61: 7 —14. 11l Effects of different fertilization practices on the N03 -N, N, P, K, Ca, Mg, ash and dietary fibre contents of carrot. J. Agric. Sci. Finl. 61: 99—111. IV Effects of different fertilization practices on the glucose, fructose, sucrose, taste and texture of carrot. J. Agric. Sci. Finl. 61: 113—122. V Effects of different fertilizationpractices on the quality of stored carrot. J. Agric. Sci. Finl. 61: 123—134. References to these articles are made in the text by citing the appropriate Roman numerals. 327 Contents ABSTRACT 329 LIST OF ORIGINAL ARTICLES 326 L INTRODUCTION 331 1.1. Vegetable quality criteria 331 1.2. Fertilization practices 331 1.2.1. Broadcast vs. placement fertilization 331 1.2.2. Fertirrigation 332 1.3. Effects of fertilization on carrot quality 333 1.4. Background and aim of the present investigation 334 2. MATERIALS AND METHODS 335 2.1. Field experiments 335 2.2. Fertilization practices 335 2.3. Carrot samples 336 2.4. Soil samples 337 2.5. Weather conditions 337 2.6. Statistical analyses 337 2.7. Organically cultivated carrots 338 3. RESULTS 340 3.1. Effects of fertilization practices on carrot growth and quality 340 3.1.1. Growth, yield and dry matter content 340 3.1.2. Carotene 341 3.1.3. Nitrate-nitrogen 342 3.1.4. N, P, K, Ca, Mg and ash 342 3.1.5. Dietary fibre 343 3.1.6. Glucose, fructose and sucrose 343 3.1.7. Taste and texture 343 3.1.8. Quality of stored carrots 343 3.2. Organically cultivated carrots 344 4. DISCUSSION 346 4.1. Growth and quality criteria influenced by fertilization practices 346 4.1.1. Placement fertilization 346 4.1.2. Single vs. split application 347 4.1.3. Fertirrigation 348 4.2. Quality criteria not or only slightly influenced by fertilization practices 349 4.2.1. Carotene 349 4.2.2. Glucose, fructose and sucrose 350 4.2.3. Quality of stored carrots 351 4.3. Organically cultivated carrots 351 5. SUMMARY AND CONCLUSIONS 353 REFERENCES 356 Maataloustieteellinen Aikakauskirja Vol. 61: 329—360, 1989 The role of fertilization practices in the yield and quality of carrot (Daucus carata L.) AINO-MAIJA EVERS Kemira Oy, Espoo Research Centre Luoteisrinne 2, 02270 Espoo, Finland Abstract. The influence of different fertilization practices on the growth, yield and inner quality of carrot (Daucus carola L.) was studied. The field experiments were carried out in Southern Finland, on the experimental farm ofKemira Oy, in 1985 and 1986. Unirrigated and irrigated placement and broadcast fertilization, NPK fertirrigations without or with basic fer- tilization, and PK placement with N fertirrigations were compared. Placement and broadcast fertilizations were carried out by single application, whereas NPK or N fertirrigations were performed as split applications. The highest yields were obtained when the total phytomass production was high. The yield increased by the placement of NPK fertilizer as compared to broadcast fertilization, and by the placement ofP and K as compared to treatments where P and K were broadcasted or fer- tirrigated. The PK placement with N fertirrigations increased the root dry matter content as compared to NPK fertirrigations. The lowest NO s-N contents of roots were obtained by unir- rigated placement fertilization. NPK fertirrigations caused higher root NOrN contents than any other fertilization practice, and irrigation as such increased the NO,-N contents. Root N and P contents were increased by placement of PK or NPK fertilizers, but fertilization prac- tices had no effects on K, Mg and Ca contents. Single application increased ash and dietary fibre contents as compared to split applications. In conclusion, the placement of NPK or PK fertilizers had positive effects on the above mentioned quality criteria and yield, and thus in future research and in commercial vegetable production these fertilization practices should be utilized. Fertilization practices had no or only slight effects on the carotene and sugar contents, taste, texture and the quality of stored carrots. Carotene content increased as the growing sea- son progressed, and was 41 % higher in the climatically more favourable year 1986 than in 1985. Total sugar contents were higher in the climatically less favourable year 1985 and in un- fertilized treatments than in 1986 and in fertilized treatments. It was hypothesized that carrot sugar contents are related to growth and total phytomass production. The indicative results showed that the NPK fertirrigations without basic fertilization had a slight positive effect on carrot taste and texture as compared to placement fertilization but, on the other hand, the NPK fertirrigations without basic fertilization showed a tendency to impair the storage ability as compared to single application. In conclusion, carotene and sugar contents are strongly in- fluenced by genotype and climate and thus it is logical that the effects of fertilization practices remained slight. The results of carrot taste do not allow any final conclusions to be drawn; further investigation on the subject is needed. Index words: fertilization practices, carrot, quality, yield, dry matter, carotene, nutrients, nitrate, ash, dietary fibre, sugars, sensory evaluation, storage 329 JOURNAL OF AGRICULTURAL SCIENCE IN FINLAND 1. INTRODUCTION 1.1. Vegetable quality criteria Modern vegetable cultivation aims to even and high-quality products. Quality can be classified as to outer quality (size, form, colour) and inner quality (nutritionally impor- tant factors, nutritionally negative factors, taste and texture). The quality classification is modified case by case, according to each specific purpose for which the vegetables are produced. The criteria of outer quality are defined in national and international quality and grad- ing standards, and the vegetable marketabili- ty and trade are based mainly on these stand- ards. Lack of rapid, inexpensive and relia- ble measurement technology has made it im- possible to classify and price the products ac- cording to inner quality. The nutritionally im- portant inner quality criteria include nutri- tional fibre, sugars, proteins and essential aminoacids, minerals, vitamins and provi- tamins, organic acids, essential fatty acids and aroma compounds. The nutritionally negative quality criteria include nitrate content, pesti- cide residues, heavy metal and radioactivity contents, oxalic acid, free amino acids and aroma compounds that give a bad taste, mak- ing e.g. the carrot bitter or harsh. Alternative cultivation methods emphasize the quality criteria differently, and the inner quality is given high priority (Schuphan 1974). Some of the quality factors can be measured by methods known to the natural sciences, but others are more difficult to de- fine and are characterised by a holistic, non- analytical view of nature (Dlouhy 1981). The quality of vegetables is influenced by genotype, climate and soil as well as by culti- vation practices, harvest, storage and market- ing. Different plant species have different op- timal growing conditions, and thus the pro- duction of high-quality vegetables is an op- timization task. Some of the variables can be influenced by a grower, some by plant breed- ers and some by nature. 1.2. Fertilization practices 1.2.1. Broadcast vs. placement fertilization Broadcast fertilization is the most common fertilization practice applied in vegetable cul- tivation in Finland, though placement fertili- zation has been shown in many agricultural studies (Aura 1967, Kähäri and Elonen 1969, Pessi et ai. 1970, Varis and Lannetta 1974, Varis 1975, Lyngstad 1977, Esala and Larpes 1986a, 1986b) to increase the yield and the efficiency of fertilizer. Broadcast fer- tilization for row crops is easy to perform in spring before sowing or planting, but after the foliage has grown, each time a tractor drives through the field some photosynthetasing leaves are damaged. And with common dis- tribution practices, it is not easy to accomplish an even distribution so that every plant ob- tains an equal amount of nutrients. Techniques where a solid fertilizer is placed under the soil surface are called row applica- tion, banded fertilization or placement fertili- zation. In these techniques, the fertilizer gran- ules are placed in a band at varying depths un- der the soil surface, where the moisture con- ditions remain optimal for granules to dissolve over a longer period of time than in broad- cast fertilization. In addition, placement fer- tilizationcan increase the fertilizer efficiency, it provides an economical and convenient operation for precise and even distribution, and considering the distance from fertilizer to 331 seeds or plants, it can prevent salt damage dur- ing germination or early growth (MacLeod et al. 1975). Various nutrients are differently suitable for broadcast vs. placement fertilization. Both ammonium and nitrate nitrogen applied to the surface of loam, silt, silt clay, and fine sand clay soils have been shown to remain in the top inch for a considerably long period in dry conditions (Kaila and Hänninen 1961), but in irrigated plots the nitrate nitrogen moves downwards (Aura 1967). This is an obvious phenomenon, as N0 3-N is unreactive with any ion retaining complexes of soil, it is read- ily mobile in the soil solution, and it moves primarily vertically as soil water moves. Am- monium nitrogen reacts with the cation ex- change complex, becoming absorbed on the surface of soil particles. Mobility is thus diminished (Randall et al. 1985). Conse- quently, the effect of placement fertilization varies from year to year and is closely related to precipitation or irrigation and to nitrogen application (Aura 1967, Lyngstad 1977). it is generally accepted that orthophosphate in soil reacts rapidly with cations, hydrous ox- ide coatings, or cations in solution to form in- soluble compounds, therefore making the mo- bility of phosphorus insignificant. Thus broadcasted phosphorus concentrates in the surface soil, is susceptible to loss by erosion (Timmons et al. 1973), and may influence the surface water quality (Young et al. 1985). The plant roots grow in deeper, moist soil lay- ers, and thus the efficiency of broadcasted phosphorus is poor. The placement of phos- phorus has been shown to increase the uptake of phosphorus (MacLeod et al. 1975, Mulkey et al. 1979) and the efficiency of phosphorus (Pearson and Kirkham 1980, Peterson et al. 1981). Placement of phosphorus has been reported to be most effective at low soil phos- phorus levels (Fox and Kang 1978, Peterson et al. 1981), in dry soil conditions (Chaud- hary and Prihar 1974, Ekeberg 1986), at low soil temperatures (Gingrich 1964, Power et al. 1964), and with granular phosphorus fer- tilizer (Engelstad and Terman 1980). It is well established that potassium as a ca- tion is absorbed to cation exchange surfaces and becomes available to plants from soil so- lutionand by contact exchange. Potassium is relatively immobile in soil, having a moder- ate cation exchange capacity, but in coarse soils it can be leached downwards by rain or irrigation. Potassium in most cases does not react to become unavailable to plants. In dry conditions, broadcasted potassium is not tak- en up by plants, but in moist conditions plants can utilize it if the roots come into contact with it. Calcium and magnesium are slightly mobile in the soil. They bind to the exchange complex of the soil and act very much like potassium. The plant uptake mechanisms are similar to those for potassium. 1.2.2. Fertirrigation The application of fertilizers in irrigation water is called fertirrigation or fertigation. The advantages of this method are the low costs of energy, labour and equipment, great flexibility in the choice of time and rate of fer- tilizer application, precise application and dis- tribution of nutrients with some low-pressure systems, and reduced leaching losses of mo- bile nutrients if the proper application tech- niques are used. The potential disadvantages are that fertilizer distribution is determined by water distribution; precise application rates are not always possible with some of the wa- ter application equipment in current use; fertilizers can be lost in runoff water; nitro- gen may be lost by volatilization from NH 4 + -containing fertilizer applied to the sur- face of calcareous soils, stratification on the soil surface or localized concentration near the emitter of immobile nutrients (e.g. phospho- rus, potassium and some micronutrients) are observed, and precipitation problems may re- sult when some forms of nutrients are in- troduced into irrigation waters high in Ca2+ , Mg 2+ and bicarbonate HC03 ~ (Randall et al. 1985). In spite of disadvantages, the fer- tirrigation is becoming increasingly popular in the USA (Randall et al. 1985). In the Mer- 332 333 die countries this fertilizer application meth- od is not widely used, mainly because of the variable weather conditions, the need for ir- rigation varying from year to year and from month to month. Also the irrigation equip- ment now in use is not especially good for maintaining an even distribution (Gregersen 1978). Aura (1983) has proposed that in Finnish conditions fertirrigation is most suit- able for plants having a long growing period and taking up nutrients relatively late. Thus vegetables and potatoes grown on coarse soils can produce good yields if fertirrigation is used. 1.3. Effects of fertilization on carrot quality The carrot root dry matter content is un- affected by increasing amounts of nitrogen (Habben 1972, Dragland 1978, Aura 1985, Vuorinen and Takala 1987), potassium (Habben 1972), organic or inorganic fertiliz- er (Nilsson 1979), or by irrigation (Aura 1985, Vuorinen and Takala 1987). A drought period before harvest increases (Dragland 1978) the carrot root dry matter content and placement fertilization decreases it (Ekeberg 1986), and the differences be- tween years have been greater than the differ- ences between fertilization treatments (Dragland 1978, Aura 1985, Vuorinen and Takala 1987). In the literature there are contradictory reports of the influenceof macronutrients on the carotene content. An increasing nitrogen amount either increases the carotene content (Freeman and Harris 1951, Habben 1972), has no effect (Dragland 1978), or it may de- crease the carotene content (Southards and Miller 1962). An increasing amount of potassium has no influence on the carotene content (Gallagher 1966, Habben 1972) or increases it (Southards and Miller 1962). The fertilizer level or the use of organic vs. inorganic fertilizers has no effect on the caro- tene content (Nilsson 1979). In fertilization experiments, the N0 3-N contents (Dragland 1978, Nilsson 1979, Lehtinen 1984) and the total nitrogen con- tents in carrots (Bishop et al. 1973, Nilsson 1979, Lehtinen 1984) have increased system- atically with increasing nitrogen amounts. The phosphorus content in carrots has not in- creased with increasing phosphorus fertiliza- tion, but increasing potassium fertilization in- creases potassium contents in carrots (Bishop et al. 1973, Nilsson 1979). Reports of the ef- fect of fertilization on the calcium and mag- nesium contents in carrots have not been pub- lished. The effects of fertilization on the ash and dietary fibre contents have been investigated only a little, but because of the vital role of dietary fibre in human health (Spiller and Freeman 1981), this subject ought to be well understood. Increasing nitrogen fertilization decreases the dietary fibre content and increas- ing potassium fertilization increases the diet- ary fibre content, but the effects have been relatively small (Habben 1972). Reports of the influence of fertilization on carrot sugar contents in the literatureare con- tradictory. The carrot glucose content in- creases by increasing nitrogen fertilization (Barnes 1936, Habben 1972), decreases by in- creasing potassium fertilization (Habben 1972), or remains unaffected by increasing nitrogen (Dragland 1978), phosphorus (Barnes 1936) or potassium fertilization (Gallagher 1966). The carrot fructose con- tent increases by increasing nitrogen fertiliza- tion and decreases by increasing potassium fertilization (Habben 1972). The carrot su- crose content decreases by increasing nitrogen fertilization (Barnes 1936, Dragland 1978), the sucrose content remains unaffected by nitrogen (Habben 1972) or potassium fertili- zation (Gallagher 1966), and the sucrose content increases by increasing phosphorus (Barnes 1936) or potassium fertilization (Habben 1972). In Nilsson’s (1979) study the glucose, fructose and sucrose contents were unaffected by either the type of fertilizer (or- ganic vs. inorganic) or the amount of fertiliz- er applied. The flavour of raw carrots is influenced by genotype and environment (Simon et al. 1982). Sugars and volatile terpenoids are the two major components of carrot flavour (Freeman and Simon 1983), but carrot aroma is very complex, and it is influenced by many compounds (Simon 1985). Results concerning the effects of fertilization on carrot taste and texture have not been found in the literature. 1.4. Background and aim of the present investigation There are several reasons why the develop- ment of fertilization practices is important. Economists have predicted that fertilizers will undoubtly become more expensive in future due to escalating costs of natural gas, labour, mining, exploration, processing and transport (Randall et al. 1985), and thus it is very im- portant to improve the efficiency of fertiliz- ers. Trade on national and international vegetable markets requires large quantities of products of even quality. At harvest, each in- dividual vegetable should be of similar size and in the same developmental stage. More precise dosage and application of fertilizer promotes even growth and supplies the plant with nutrients needed at any given time. Growing concern is being voiced by the pub- lic towards the impact of fertilizers on the en- vironment. Modern application techniques can play a significant role in preserving the quality of the environment while still produc- ing optimum crops. Techniques to reduce en- vironmental concern almost always result in improved fertilizer efficiency (Randall et al. 1985). The primary aim of the present inves- tigation was to study whether the development of fertilization practices affects some of the inner quality criteria of carrot. In public discussion there is increasing in- terest in human health and nutrition, and of- ten vegetable quality is considered to depend mainly on modern growing practices and mineral fertilizers. However, also genetic and environmental factors have a certain effect on the quality. The secondary aim of the present investigation was to find out which quality criteria are mainly dependent on genotype and climate, and which on fertilization. 334 2. MATERIALS AND METHODS 2.1. Field experiments The field experiments were carried out on the Kotkaniemi Experimental Farm of Kemira Oy in Vihti, Southern Finland (60°22'N, 24°22'E) during the growing seasons of 1985 and 1986. The carrot variety grown in this ex- periment was cv. Nantes Duke Notabene 370 Sv which is commonly grown for thefood in- dustry in Finland. Coated seeds (I) were sown 1 cm deep at a row distance of 50 cm and plants were thinned later to 30 plants/m. Weeds were sprayed with promethryn twice per growing season. Sypermetrin was used for protection against carrot psyllit, Trioza apica- lis F. The carrots were harvested manually, and 8 kg of sample per plot in 1985 and 15 kg in 1986were stored in refrigerated storage (0... + I.5°C, RH 90—95 %) for six months and four months, respectively. 2.2. Fertilization practices In the field experiments, ten treatments were compared (Table 1). In the case of unir- rigated and irrigated controls, no annual fer- tilization was applied. In the original articles I—V these have been called unfertilized treat- ments. In broadcast and placement practices (Ta- ble 1), all nutrients were given as a single ap- plication in spring, before sowing. Placement fertilization was done with a fertilizer drill. Table 1. The fertilization treatments and the total amounts of nutrients and irrigation water. Treatment Number and time of Macronutrient Irrigation water fertilizer applications amounts in 19863 amounts in 1985 kg/ha and 1986 N P K mm Unfertilized Unirrigated 0 0 0 0 0 Irrigation 0 0 0 0 3x 10 NPK placement Unirrigated Once before sowing 80 35 133 0 Irrigation Once before sowing 80 35 133 3x 10 NPK broadcast Unirrigated Once before sowing 80 35 133 0 Irrigation Once before sowing 80 35 133 3x 10 NPK fertirrigations No basic Three times during the season 80 29 160 3x 10 Half the basic 1 Once before sowing and 80 32 142 3x 10 three times during the season PK placement 2 3N-fertirrigations Once before sowing and 81 56 133 3x 10 three times during the season 4N-fertirrigations Once before sowing and 155 56 133 4x 10 four times during the season ' Half of the nutrients were given in basic placement fertilization and half in NPK fertirrigations. 2 Phosphorus and potassium were given in basic placement fertilization and nitrogen in fertirrigations. ' The nutrient amounts were 30 % higher in 1985 than in 1986. 335 The fertilizer row distance was 50 cm, and the seeds were sown afterwards, 3 cm beside the fertilizer row. In the broadcast method the fer- tilizer was spread on the soil surface manual- ly and harrowed in with a rotary cultivator. The NPK fertirrigations and PK placement with N fertirrigations (Table 1) were split ap- plications, in other words, the nutrients were given periodically three to five times during the growing season. In NPK fertirrigations, a water-soluble fertilizer was dissolved in wa- ter, and all or half of the nutrient amount was spread with a self-made boom. In PK place- ment with N fertirrigations, the phosphorus and potassium were given by the placement method with a fertilizer drill in spring before sowing, and nitrogen was given by nitrogen fertirrigations during the growing season. All fertilized treatments received the same amount of nitrogen except PK placement with 4N fertirrigations which received a supraop- timal amount of nitrogen (Table 1). The PK placement treatments received a greater amount of phosphorus owing to the phospho- rus content of the granular fertilizer used in the experiments. The amount of potassium was somewhat greater in the NPK fertirriga- tion treatment owing to the fertilizer used. The micronutrient fertilizations are given in the original article I. Irrigated placement and broadcast fertilized treatments received the same amounts of water as the fertirrigated treatments (Table 1). 2.3. Carrot samples In 1985, carrot root and shoot samples were collected three times during the growing sea- son (Table 2) for determinationof fresh and dry weight and N0 3 -N, total N, P, K, Ca, and Mg contents. Carrot root samples were collected at harvest for yield, for determina- tion of carotene, glucose, fructose and su- crose, and for sensory evaluation, e.g. taste and texture. After the storage period, the mar- ketable yield and weight loss were weighed, the carotene content was determined, and sen- sory evaluation for taste and texture was per- formed. In 1986, carrot root and shoot samples were collected three times during the growing sea- son (Table 2) for determination of fresh and dry weight and NOrN content, and root samples were collected for determination of carotene content. Carrot root and shoot sam- ples were collected at harvest for determina- tion of N, P, K, Ca, Mg and ash contents, and root samples for determinationof yield, die- tary fibre, glucose, fructose and sucrose con- tents, and for sensory evaluation, e.g. taste and texture. After the storage period, the mar- ketable yield and weight loss were weighed, the carotene, NOrN, N, P, K, Ca, Mg, ash, glucose, fructose and sucrose contents were determined, and sensory evaluation for taste and texture was performed. The sample size for yield was 8 m 2, i.e. 16 row metres per plot. Other sample sizes are shown in Table 3. Table 2. The schedule for fertilization, sowing, sampling, harvest and storage. Fertilizer experiment Organic cultivation 1985 1986 Location 1 Location 2 Basic fertilization 4 June 1985 6 June 1986 14 May 1986 21 May 1986 First fertirrigation or irrigation 14 June 1985 29 June 1986 Second fertirrigation or irrigation 5 July 1985 10 July 1986 Third fertirrigation or irrigation 19 July 1985 24 July 1986 Fourth fertirrigation or irrigation 2 Aug. 1985 7 Aug. 1986 Sowing 8 June 1985 9 June 1986 14 May 1986 21 May 1986 First plant sampling 20 Aug. 1985 12 Aug. 1986 14 Aug. 1986 14 Aug. 1986 Second plant sampling 10 Sept. 1985 3 Sept. 1986 I Sept. 1986 I Sept. 1986 Harvest 30 Sept. 1985 6 Oct. 1986 25 Sept. 1986 25 Sept. 1986 End of storage 2 April 1986 23 Dec. 1987 23 Feb. 1987 23 Feb. 1987 336 Table 3. The carrot sample sizes in the experiments of 1985 and 1986. Sample size Carrots Carrots per per plot treatment 1985 Root and shoot fresh and dry weight N03-N 20 m m 80 20* 20*N, P, K, Ca, Mg Root carotene 12*6 10*glucose, fructose, sucrose taste and texture 5 10 20* 1986 Root and shoot fresh and dry weight NO,-N 20 80 20 80 20N, P, K, Ca, Mg, ash Root carotene 5 6 24 20dietary fibre 5 5glucose, fructose, sucrose taste and texture 20 4010 Carrots were collected from two blocks only. In 1985, samples for determination of qual- ity were collected from two blocks only. A heavy rain in June 1985 caused a crust on the soil surface, as a result of which the planta- tion was uneven in two blocks which were omitted. The methods used to analyse the plants are described or referred to in the following origi- nal articles: dry weight (I) carotene (11) NOj-N, N, P, K, Ca, Mg, ash and dietary fibre (III) glucose, fructose, sucrose (IV) taste and texture (IV) 2.4. Soil samples The soil in the experimental field was fine sand with 15—30 °7o clay and a humus con- tent of 12—20 °7o. Soil samples were collected across the carrot row from a soil layer 15 cm deep. Samples were taken two days before sowing and 6, 27, 41, 54, 76, 96 and 116 days after sowing in 1985. In 1986, the samples were taken one day before sowing and 18,31, 46, 65 and 117 days after sowing. Soil sam- ples were collected and analysed as described in the original article I. The chemical charac- teristics of the soil in spring before fertiliza- tion are given in the original article I, Table 4, those during growing seasons are given in the original article I, Figure 5. 2.5. Weather conditions The weather conditions were unfavourable in June and July 1985 (I/Table 3). The mean day temperature and the number of sunshine hours were lower than the long-term averages. On 16 June, heavy rain came only two days after the first irrigation, when seedlings had begun to emerge. The heavy rain caused a crust to form on the soil surface, and the crust hindered the emergence of seedlings. In Au- gust and September 1985, weather conditions were moderate, but the soil was too wet. In 1986, June was very warm and sunny. The mean day temperature and the number of sunshine hours were above the long-term averages (I/Table 3). The precipitation was only 28 mm, and the soil was dry. July was favourable for growth. August and Septem- ber were cold and rainy, and thus unfavoura- ble for growth. Altogether, 1985 was a less favourable year for growth than 1986. 2.6. Statistical analyses The field experiments were set up accord- ing to the method of completely randomized blocks. There were four blocks, ten treatments and thus totally 40 plots. The results were studied statistically by con- trast analysis (Steel and Torrie 1980). The two years were studied separately (II—V), ex- cept in the original article I. The partitioning of treatments into contrasts are presented in Table 4. The unfertilized, placement and broadcast fertilized treatments were also stud- ied by the analysis of variance as a factorial experiment. The fertilization levels were 1) no 337 fertilization, 2) placement fertilization, and 3) broadcast fertilization. The irrigation levels were 1) no irrigation and 2) irrigation. The sig- nificance of differences between means were tested with the Duncan test or with Student- Neuman-Keul’s test (Steel and Torrie 1980). The Pearson’s product moment correlation coefficients (r) were calculated to measure the strength and direction of the linearity between the variables. The differences were considered to be statistically significant at the 5 % level, and were marked with asterisks as follows: * significant at 5 % level, p<0.05 ** significant at 1 % level, p