Effect of soil compactness on the growth and quality of carrot Liisa Pietola AgriculturalResearch Centre ofFinland, Institute ofCrop and Soil Science, Jokioinen, Finland Department ofApplied Chemistry and Microbiology, University ofHelsinki, Finland ACADEMIC DISSERTATION To be presented, with the permission of the Faculty ofAgriculture and Forestry of the University ofHelsinki, forpublic criticism in Auditorium XII, Aleksanterinkatu 5, Helsinki, on June 14th, 1995, at 12 noon. AGRICULTURAL SCIENCE IN FINLAND https://www.c-info.fi/en/info/?token=HUHO-McbZcE9anTf.I8uJlIppD9AenJ0VmFiiyw.kcOBAEg9s1mbh7dV2EpW2ScR1LQ0gVSBMTxJ0to9o_hslOirr13a4op4ZgF07U1YGEsGjzp2g63kDCgN16Ljh3JegW4TdIqOnuaGnvhJz-RxKJwvad3HtB4_KyQcG2FL71hL8yBEiGYKV3q7jB4h1kwO0XHNSBHDAXb3yewGn9iU80ajE9u3msK-9WPfhb8p3qvcvUjXtChIViqgIxwJRkhBHE3jQFnIFqkw_YgbhraNT_qxOcKUpVGqkUZ2p_wqIofsJQ AGRICULTURAL SCIENCE IN FINLAND Preface The experimental part of this study was mainly carried out at the Agricultural Research Centre of Finland in 1989-1992, the root analyses being completed in 1993 at Michigan State University, USA. The work was finalized in 1994 at the University of Helsinki, Sec- tion of Agricultural Chemistry and Physics of the Department of Applied Chemistry and Microbiology. I wish to thank Dr. Esko Poutiainen, Director General of the Research Cen- tre and Dr. Paavo Elonen, Professor of Agricultural Chemistry and Physics at the Research Centre, for providing me the main financing and facilities for the experiments. I am very grateful to Dr. Alvin J. M. Smucker, Professor of Soil Biophysics at Michigan State Univer- sity, East Lansing, for allowing me to join his laboratory for root analyses, and Dr. Antti Jaakkola, Professor of Agricultural Chemistry and Physics at the University of Helsinki, for his constructive criticism and support at the various stages of the work. I wish to thank Docent Irma Voipio and Docent Erkki Aura for the valuable suggestions to the manuscript. I am grateful to the staff of the Agricultural Research Centre, especially to Mr. Tapio Salo, M. Sc., Mr. Risto Tanni, Mrs. Erja Äijälä and Mrs. Ritva Niemi at the Institute of Crop and Soil Science, for the skilful technical assistance in the experiments. I greatly appreciate the guidance and technical support of Mr. John. C. Ferguson, B.Sc. at Michigan State University. I also wish to thank the staff of the Department of Applied Chemistry and Microbiology at the University of Helsinki for the analysis of soil air com- position and Dr. Inge Håkansson, Professor of the Department of Soil Sciences at the Swedish University of Agricultural Sciences, for allowing the determinations of soil reference bulk densities to be made in his department. The Figures (6-35) were drawn by Mr. Ari Törmä, M.Sc. and the English manuscript was revised by Mrs. Sevastiana Ruusamo, M.A., and edited by Mrs. Sari Torkko, M.Sc., which work I greatly appreciate. This investigation was financially supported by the Academy ofFinland and the TiuraFoundation, which is grate- fully acknowledged. I would also like to thank the board of the Agricultural Science in Finland for including this study in their journal. Finally, my warmest thanks are due to my family whose great support made it possible for me to complete this work. East Lansing, February 1995 Liisa Pietola AGRICULTURAL SCIENCE IN FINLAND Contents Abstract 1 Introduction 145 1.1 Importance of soil compactness to plant growth 145 1.1.1 Consequences of excess soil compaction and loosening 145 1.1.2 Optimum soil compactness 146 1.2 Significance of soil properties to carrot rooting and yield quality 148 1.3 Efficiency of root system 148 1.4 Background and aim of the present study 149 2 Materials and methods 150 2.1 Experimental fields 150 2.2 Treatments 152 2.2.1 Irrigation 152 2.2.2 Soil mechanical treatments 152 Soil loosening 152 Soil compaction 152 2.3 Establishment and management of field experiments 153 2.3.1 Fertilization and spring tillage 153 2.3.2 Sowing and management during growth 153 2.3.3 Sampling 154 2.4 Weather conditions 154 2.5 Soil measurements 154 2.5.1 Soil water content 154 Electrical resistance 154 Sampling and drying 155 2.5.2 Soil dry bulk density 155 Gravimetric analyses 155 Gamma ray transmission 155 2.5.3 Degree of soil compactness 156 2.5.4 Soil penetrometer resistance 156 2.5.5 Soil pore size distribution 156 2.5.6 Soil air composition 157 2.6 Plant measurements 157 2.6.1 Sampling 157 2.6.2 Yield 158 2.6.3 Tap root quality 158 External quality 158 Internal quality 159 2.7 Studies on carrot fibrous root system 159 2.7.1 Destructive sampling 159 2.7.2 Dry weight 160 2.7.3 Root morphology 161 Video recording 161 Image analysis 162 2.8 Statistical analysis 163 3 Results 164 3.1 Effects of treatments on soil physical properties 164 3.1.1 Soil moisture 164 3.1.2 Soil dry bulk density 167 Gravimetric analysis 167 Gamma ray transmission 170 Gravimetric analysis vs. gamma ray transmission 171 3.1.3 Degree of surface soil compactness 171 AGRICULTURAL SCIENCE IN FINLAND 3.1.4 Penetrometer resistance 172 Relation to soil moisture and dry bulk density 175 3.1.5 Soil porosity and water retention capacity 177 Pore volume 177 Water retention 179 3.1.6 Soil air composition 182 3.2 Response of carrot growth and yield quality to treatments 183 3.2.1 Yield and external quality of tap roots 183 Biomass accumulation 183 Final yield 186 Root size 187 Root splitting and branching 190 Tap root length and diameter 193 Root shape 193 3.2.2 Internal quality of tap roots 196 Dry matter content 196 Crude fibre content 197 Juice content 197 Juice dry matter content 197 Sugar content 206 Carotene content 206 Internal quality vs. root size 206 3.3 Response of whole-root system and shoot growth to differently treated soil profiles in PVC cylinders 209 3.3.1 Distribution of weights 209 Tap root and shoot weight 209 Fibrous root weight 209 Fibrous root weight to tap root weight ratio 210 3.3.2 Distribution of fibrous root length 213 Root length per plant 213 Root length to tap root weight ratio 215 3.3.3 Distribution of fibrous root surface area 217 Root surface area per plant 217 Root surface area to tap root weight ratio 218 3.3.4 Fibrous root width 219 4 Discussion 221 4.1 Effect of soil loosening, compaction and irrigation on soil physical growth factors 221 4.2 Effect of soil physical growth factors on carrot growth and yield quality 223 4.3 Role of fibrous root system in carrot response to soil compactness 226 5 Conclusions 228 References 230 Selostus 236 Appendixes 1-5 AGRICULTURAL SCIENCE IN FINLAND Effect of soil compactness on the growth and quality of carrot Liisa Pietola Agricultural Research Centre ofFinland, Institute of Crop and Soil Science, Jokioinen, Finland Present address: Department ofApplied Chemistry and Microbiology, P.O. Box 27, FIN-00014 University ofHelsinki, Finland Field experiments were performed in Southern Finland on three soil types: fine sand (1989-1991), clay (1989) and mull (1990-1991). The following soil mechanical treatments were applied to autumn ploughed land: soil loosening by ridge preparation (ridge distance 45 cm), rotary harrowing (to a depth of 20 cm, clay 15 cm), and soil compaction track by track by a tractor weighing 3 Mg (1 or 3 passes, wheel width 33 cm) before seed bed preparation. One plot was untreated. These treatments were set up in April (on clay in May) under moist soil conditions. Sprinkler irrigation (one applica- tion of 30 mm) was applied to clay and fine sand when soil moisture in top soil had decreased to around 50% ofplant-available water capacity. PVC cylinders (r = 15 cm, h = 60 cm) were fixed in the experimental areas during the growing periods. At harvest, these cylinders were removed for specific analysis of tap and fibrous roots of carrot. Length and width of fibrous roots were quantified by image analysis in the USA. The impacts of soil loosening and partial compaction were determined by measuring soil physical parameters to a depth of 25 cm in mineral soils, and to greater depths in organic soil. Dry bulk densities of the plough layers increased with increasing tractor passes by 8%, 10% and 13% for fine sand, mull and clay soils, respectively. The lowest dry soil bulk density in the plough layer was obtained by rotary harrowing to a depth of 20 cm. Comparison of gamma ray transmission and gravi- metric analysis indicated that dry soil bulk density was slightly lower when determined by gravimet- ric analysis. Increased soil bulk densities were reflected by increased water retention capacity (mat- ric suction 30 pm). Soil compaction promoted crop establishment and early growth as compared with loose soil beds. Optimum soil compactness for carrot yield (D = 82) was observed only in clay field where excess loosening or compaction affected yield quantity adversely at different stages of growth. During bio- mass accumulation, excessive penetrometer resistances limited tap root growth in compacted fine sand without irrigation. Water applications promoted shoot growth, but did not affect final shoot and tap root yield. Among the three soil types tested in this study, compaction of mull soil had the least effect on carrot growth and external quality. © Agricultural Science in Finland 144 Pietola, L.: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND This paper presents evidence that the internal quality of carrots is only slightly affected by chang- es in soil physical properties, while the adverse effects of soil compaction on carrot external quality (short, deformed and conical tap roots with greater maximum diameters) are clear. Even though com- pacted clay soil greatly limited the biomass accumulations in the tap root, which had a high crude fibre content, the carotene (10 mg/100 g carrots) and sugar contents (5%) reached acceptable levels. The lowest carotene contents (4 mg/100 g carrots) were observed in loose mull, following a cool late summer in 1990. The effect of irrigation on carotene content varied from one year to another. High sugar and carotene contents appeared to respond to the high below-ground absorption sur- face. The fibrous root system of carrots, consisting of mostly very fine roots (diameter 0.15 mm), had total lengths of 150 m in loose fine sand at a soil depth of 0-50 cm (rotary harrowed), 200 m and 300 m in fine sand and mull soils subjected to 3 passes by a tractor wheel. The maximum dry weight (60 pg), length (1.2 cm) and surface area (0.05 cm 2 ) of the fibrous root system per soil volume (cm 3) were observed in compacted or irrigated soil to a depth of 30 cm, and also in relation to tap root dry weight. This suggests a capacity of carrot plant for high below-ground absorption potential and opti- mal biochemial maturation of tap root tissue even when surface soils are compacted. This is support- ed by higher leaf area, as the early shoot growth was promoted by partial soil compaction. Soil compaction affected the soil physical properties and carrot external quality in agreement with previous studies. Carotene and sugar contents appeared to be unaffected or were slightly in- creased in riper and firmer carrots of compacted soils. This is consistent with the earlier information about the internal quality of carrot which is shown to be highly dependent on genetic factors and developmental stage of carrot. The present study emphasizes the surface area of carrot fibrous root system as a beneficial factor for maintaining high levels of carotene and sugar contents in tap roots after partial soil compaction. Key words: tillage, traffic, soil physical properties, carotene, sugar, root length, root surface area, image analysis Introduction 1.1 Importance of soil compactness to plant growth 1.1.1 Consequences of excess soil compaction and loosening Poor aeration and high mechanical impedance are the major stress factors affecting the growth of most plants in compacted soil. Tractor wheel traffic, withaxle loads of no more than 3 Mg on wet clay soils, destroys the total macroporosity (pore diameter >3O pm) to a portion below 10% of soil volume (Aura 1983) which is considered to be the critical limit for soil aeration maintain- ing plant growth (Glinski and Stepniewski 1985). This detrimental effect on soil porosity reflects in altered soil air composition (Eavis 1972, Si- mojoki et al. 1991) and in an oxygen diffusion rate below 30 pg nr2 s' 1 which is too low for plant growth (Erickson 1982, Asady et al. 1985). In highly compacted soils the root to soil contact may be so intense that aeration of the root tissue is completely dependent on the internal air chan- nels in the root (Veen et al. 1992). Further, as soil compaction has an adverse effect on satu- rated water retention capacity and infiltration (Blake et al. 1976,Reicosky et al. 1981, Ankeny et al. 1990), poor aeration of compact and flood- ed soil inhibits root growth and nutrient uptake (Grath and Håkansson 1992). Under localized anoxia, compensatory root growth causes inef- ficient utilization of carbon and additional up- take of water and nutrients (Schumacher and Smucker 1984, 1987). 145 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Mechanical impedance is another component of soil compaction often implicated in poor growth. Pohjanheimo and Heinonen (1960) re- ported that the hardening and drying out of clay loam soil may entirely inhibit the penetration of barley roots into the subsoil. Soil strength on some roots is reflected also as reduced growth rate of other plant parts (Massle and Passioura 1987). In the absence of continuous large pores, soil resists the local deformation caused by roots, and there is a definite upper limit to a pressure near 1 MPa in the axial direction and 0.5- 0.9 MPa in the radial direction which can be exerted by roots of a given species (Pfeffer 1893, ref. Gill and Bolt 1955, Misra et al. 1986). Ra- dial expansion of a root behind the tip causes a lowering of external mechanical impedance ahead of the elongating root (Abdalla et al. 1969, Richards and Greacen 1986). Mechanical impedance measured by pene- trometers should be regarded as comparative, not absolute values. The static and additive frictional components of penetrometer probe results in greater soil resistance values than for a root (Far- rell and Greacen 1966, Whiteley et al. 1981). According to Materechera et al. (1991), the soil penetrometer resistance of4.2 MPa corresponds to an external mechanical resistance on the roots of approximately 1.14 MPa. The pressure at the root apex cannot, how- ever, be the only determining factor regulating the root elongation rate, as a clear relationship has been observed between the penetrometer resistance in the top layer and root growth rate in the subsequent loose layer (Bengough and Young 1993). The restricted root growth with morphological deformities, such as diameter growth and compensatory growth of laterals, is a general symptom of too high soil strength un- der field conditions (Wiklert 1960, Voorhees et al. 1975). Mechanically impeded roots, in turn, exhibit an increased respiration rate per unit root length (Schumacher and Smucker 1981, Atwell 1990a, b). Poor water supply and small root to soil con- tact area of loose soil can also contribute to poor growth. Soil needs some compaction, i.e. soil compression (Smucker and Erickson 1989), to be the most productive (Håkansson 1966, John- son et al. 1990). This has been proved particu- larly during drought periods, due to lower volu- metric water content (Boone et al. 1978, Dom- zal and Hodara 1992) and, thus, a weaker un- saturated hydraulic conductivity in loosely packed fine textured soils than under slightly compacted conditions (Kemper et al. 1971, Voor- hees et al. 1979, Mehta et al. 1994). On the oth- er hand, soil compaction reduces water conduc- tivity of coarse textured soils at matric poten- tials between 0 and -60 kPa (Lipiec and Tarkie- wicz 1984). According to Agrawal (1991), this improves the productivity of sandy soils with too high water infiltration by reducing losses of water and nutrients. Compaction of sand has been reported to increase yields by as much as 30-50%. The poor contact area between root and loosely packed soils has been discussed recent- ly: Kooistra et al. (1992) measured by a thin- section tehnique average root to soil contacts of 60, 72 and 87% for sandy loam of a porosity of 60, 51 and 44% (v/v), respectively. They also proved that water absorption and nitrate uptake per unit maize root length decreased with loos- ening soil and with decreasing root to soil con- tact (Veen et al. 1992). Similarly, Huang (1990) demonstrated that the greater root to soil con- tact at the high bulk density of water deficit plants increased water absorption per unit of roots growing in clay soil. These results are well in agreement with the earlier measurements of Lipiec et al. (1988, 1992)which indicated a faster and higher water absorption by a plant growing in compacted soil than in loose media. 1.1.2 Optimum soil compactness Measurements for soil compactness include analysis of soil dry bulk density and soil poros- ity. Different sizes of penetrometers have been used to indicate soil strength of dense layers (Bengough and Mullins 1990) or profile charac- teristics after various tillage operations and com- 146 Pietola, L : Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND paction (Carter 1988,Akkeretal. 1994). A meas- urement for soil property which is related to the variation in continuous pore size distribution, like saturated water conductivity, should be in- cluded in the analysis of soil compactness as well (Hartge 1992). To be able to compare the results of soil compactness between different soil types, soil compactness has been determined also as a relative value by means of the reference state. Håkansson (1990) specifies therelative soil com- pactness as a “degree of soil compactness” (D) which is given by the formula; D= 100 (p d /p dp ) (1) where p d is the dry bulk density of a soil (g cm 3) and pd p the dry bulk density of the same soil (g cm 3 ) after excessively compacted to the ref- erence state by an unaxial pressure of 200 kPa until drainage and compaction ceases (after 1 week). Initial soil water contents are near field capacity (matric suction =° 10kPa). Optimum soil compactness is reached by avoiding excessive soil loosening and compac- tion. According to field experiments in Sweden during the last three decades, barley yield is high- est at D around 87 in all mineral soil types. This compactness is achieved with one pass of a trac- tor wheel on moist ploughed land during spring tillage. The finding is well in agreement with results from Norway (Riley 1988) and Poland (Lipiec et al. 1991). The variation in seasonal precipitation ex- plains well the differences between yields at- tained at the optimum and non-optimum states ofsoil compactness (Lipiec et al. 1992). The dis- advantages of soil compaction are most promi- nent after excessive rains which cause oxygen stress. Similarly, lower optimum soil densities are needed during rainy seasons because there is a smaller need to retain all of the water (McKyes et al. 1979, Medvedev 1992). Thus, according to Håkansson (1992), the relation between plant growth and soil compactness is understandable only if soil moisture status has been measured along the growing period. He outlined the criti- cal values of soil air content, penetrometer re- Matric water tension (kPa) sistance and unsaturated hydraulic conductivity for plant growth in relation to both soil moisture and degree of compactness (Fig. 1). The data in Figure 1 is based on the analysis of two Polish loam soils by Lipiec et al. (1991) but, according to the authors, the information is generally ap- plicable to all mineral soils. In organic soils, or- ganic particles may be irreversibly deformed during prolonged loading, resulting in an over- estimationof the reference bulk density (Håkans- son 1990). The optimum compactness varies according to plant species, and precipitation at the time of the most intensive yield production is a decisive factor (Soane 1992). In the Nordic countries, most of the precipitation during growing seasons falls at the end of the summer. Consequently, a high state of soil compactness may affect most adversely crop production of late species, such as potatoes and root crops. For these crops, the disadvantages of soil compaction are common- ly prevented by preparing very loosely packed raised beds or narrow ridges (Millette et al. 1981). The bed cultivation system is often car- ried out with a controlled traffic system using permanent traffic lanes (Monroe and Taylor 1989). Water deficit may, however, occur soon- Fig. I. Critical limits for plant growth in relation to degree ofsoil compactness and matric waterpotential of theplough layer (Håkansson 1992). Shaded area indicates adverse soil physical properties for plant growth. 147 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND er in raised beds than on flat land (Mahrer and Avissar 1985). This is well in agreement with the above mentioned low hydraulic conductivi- ty and water retention capacity of loosely packed soils. 1.2 Significance of soil properties to carrot rooting and yield quality Carrot growth and yield quality are excellent in- dicators of soil compactness, as the external quality ofcarrot tap root is sensitive to soil com- paction, even on organic soils (Strandberg and White 1979). Mechanical impedance and poor aeration of compacted soil restricts the carrot root size and modifies the root, making it coni- cal. Followed by young carrot root buckling, branching and thickening (Strandberg and White 1979), the number of short, thick and branched tap roots is high under compacted soil conditions (White 1978,Taksdal 1984, Kesik 1990). On the other hand, tap root yield (Olymbios and Schwabe 1977,Agung and Blair 1989) and seed- ling emergence (Strzalka 1990) are reduced by excess soil loosening. The role of soil chemical and physical prop- erties in the internal quality of carrots is minor or may not be well understood. Although place- ment fertilization (Evers 1989a) and drought (Dragland 1978) have been found to be favoura- ble to carotene production, the developmental stage of carrot has a major influence on the in- ternal quality factors of carrot tap root, such as dry matter, sugar (Platenius 1934, Hole and McKee 1988) and carotene contents (Barnes 1936,Evers 1989a). These and tap root firmness increase along the growing season in develop- ing tap root, indicating root biochemical matu- rity, i.e. ripeness (Fritz and Habben 1974). Also genotype and weather conditions have a clear impact on carrot internal quality (Bradley and Smittle 1965, Simon et al. 1982, Miedzobrodz- ka et al. 1993, Evers 1994). Agung and Blair (1989) reported of poor ex- ternal quality of carrot tap root and decreased fibrous root length density caused by highly compacted soil but, unfortunately, no data on the internal quality were collected. The total length of carrot fibrous root system increased signifi- cantly during the last 50days before harvest (153 days after sowing) in all pots. The pots were filled by packing homogeneous soil to different dry bulk densities to create artificial soil pro- files of very great variation in soil compactness which hardly exist under field conditions with a normal field traffic. Moreover, this study was carried out without any biopores or cracks which contribute to root penetration, but have a limit- ed effect on average soil physical properties, such as mean bulk density or penetrometer re- sistance. As the thickness of the compact layer represents usually less than 25% of the final root- ing depth of cultivated plants (Tardieu 1994), the effects of localized compact zones on the root- ing characteristics of whole-root systems should be emphasized. 1.3 Efficiency of root system The response of growth to soil compactness is determined also by root characteristics of a giv- en species. These are root dimensions (length, diameter, surface area) which affect therequire- ments of carbon (Eissenstat 1992), nutrient and water uptake (Barber and Silberbush 1984), abil- ity to penetrate into soil (Richards and Greacen 1986, Materechera et al. 1991) and capacity to maximize root to soil contact area (Veen et al. 1992). Root length is the most frequently measured plant property influencing water and nutrient uptake (Nye and Tinker 1977, Molz 1981, Noor- wijk and Willigen 1991). Root diameter is also important, as species which have root systems with numerous fine roots containing only a few layers of cells produce more root surface area with less photosynthesized carbon (Eissenstat 148 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot 1992). Moreover, roots with a smaller diameter will have a better nutrient uptake as demonstrat- ed by Itoh and Barber (1983a) for theroot diam- eter range of 0.03-0.3 mm. In addition, fine roots remove more water (Eissenstat and Caldwell 1988a) stored in the soil profile than do larger roots of similar root biomass and distribution with depth. According to Richards and Greacen (1986), fine roots are also less affected by high mechanical impedance, even though in very compacted homogeneous soil (penetrometer re- sistance 4.2 MPa) there is a positive correlation between root diameter and elongation (Mater- echera et al. 1991). Noordwijk et al. (1993) found, however, no relation between root-soil contact and root diameter or roundness. The absorptive capacity of the root system is a product of root surface area and root permea- bility. According to Fiscus and Markhart (1979), the system size seems to be the dominant and root permeability the minor factor. Also age of the root seems relatively unimportant for the nutrient uptake (Clarkson and Hanson 1980). Root hair lengths less than 0.4 mm which are common for field crops have a minor effect on ion uptake and water entering as short root hairs do not increase significantly the absorbing sur- face of roots in the soil (Jones et al. 1983, Bar- ber and Silberbush 1984, McCully and Canny 1989). Fibrous roots are seldom distributed uniform- ly within a soil layer (Logsdon and Allmaras 1991,Aiken 1992). Roots tend to preferentially colonize loose zones, between soil aggregates, biopores or other planes of weakness by buck- ling as a root grows across a macropore and meets a solid surface (Dexter and Hewitt 1978, Pietola 1991). Roots congregate also in the fer- tile (Eissenstat and Caldwell 1988b) and wetter zones (Smucker 1993). Thus, root characteris- tics should be quantified from numerous hori- zontal and vertical soil planes or layers. While non-destructive root sampling by rhizotrons and minirhizotrons (Taylor et al. 1990) offers possi- bilities to do research on spatial and temporal root dynamics, i.e. root growth and turnover rates, destructive samples are useful for quanti- fying root biomass, length, diameter, surface area and volume at each sampling time (Smucker 1993). Early root length analyses rely upon New- man’s line-intersect method (Newman 1966). Also a core-break method has been used for root length, but it provides only a low precision esti- mate of root length density (Bland 1989). Auto- mated root intersect counting by video camera imaging (Voorhees et al. 1980) has been extend- ed to root image analyses (length, width, branch- ing) by using a scanner and a microcomputer (Zoon and Tienderen 1990, Pan and Bolton 1991, Ewing and Kaspar 1993). For recording project- ed root area or root surface area, a video camera and a computer system for digitizing and ana- lysing video images have been developed for clean washed roots (Harris and Campbell 1991, Kokko et al. 1993) and also for washedroot sam- ples with residues (Smucker 1993). The image analysis permits examination of root length, di- ameter and surface area which all are needed for the evaluation of the efficiency of root systems on a morphological basis. 1.4 Background and aim of the present study Growth of above-ground biomass in compacted soil has been the subject to much research over many years in the Nordic countries, but mainly on small grains on clay soils. Information about the optimum state of soil compactness for other plant species on different soil types is still need- ed. For understanding the relations between soil compactness and plant growth, responses of dif- ferent physical growth factors ofsoil to both soil loosening and compaction should be quantified on different soil types under varying soil mois- ture conditions, including impacts on below- ground growth. The disadvantages of soil compaction are better known than the negative effects of inten- 149 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND sive soil loosening, especially in vegetable pro- duction where the bed cultivation system has been applied extensively. This cultivation tech- nique aims at even and high-quality products. However, even if the external quality of carrot benefits from loose soil conditions, the question so far unanswered is: “To what extent and how is the carrot internal quality affected”. Since water and nutrients are absorbed through roots, an extensive functioning root system is of pri- mary importance to plant growth. If the distri- bution of fibrous root surface area were not ad- versely affected by moderate soil compaction, comparison of loose and slightly compacted soil conditions would suggest more intensive ripen- ing (i.e. carotene and sugar synthesis) in slight- ly compacted soil where the soil to root contact and the soil moisture retention capacity are high- er. Therefore, quantitative information about the interaction of soil compactness and moisture on carrot rooting and yield quality under field con- ditions is desirable, with special reference to in- ternal quality factors. The objective of this study was to evaluate the response of carrot growth and quality to the soil compactness of different soil types under different soil moisture conditions. Field experi- ments were conducted: 1) To determine the alterations of soil compact- ness and physical growth factors (soil water, mechanical impedance, aeration) induced by soil loosening, partial compaction and irri- gation. 2) To identify the changes in carrot growth and quality caused by soil compactness, empha- sizing tap root size, shape and ripeness (dry matter content, firmness, sugar and carotene content). 3) To evaluate therole of carrot fibrous root sys- tem in the response ofcarrot biochemical ma- turity to the soil compactness. 2 Materials and methods 2.1 Experimental fields The empiric data was collected from three field experiments established at the Agricultural Re- search Centre ofFinland in Jokioinen (60°49’N; 23°28’E) on undifferentiated Spodosols (Foth 1990) (Appendixes 1-3). The experiment on fine sand soil was set up in all experimental years (1989-1991), but the experiment ofclay soil rich in organic matter was conducted only in 1989. These two fields were situated in the same field area. The third field, a mull soil (mixture of peat and clay), was studied in 1990-1991, and it was situated around 10 km away from the other two fields. The sown area was 1040 m 2 for the ex- periments of sandy and clay soils and 520 m 2 for organic soil. Topsoil (i.e. plough layer) and subsoil samples were taken from all four blocks of the experimen- tal fields. In fine sand, the thickness of topsoil was approximately 25 cm, that ofclay soil 26 cm, and that of mull soil 27 cm. The samples were ana- lysed for particle size distribution by the method ofElonen (1971) and for organic carbon content by a Leco analyzer at 1370°C (Sippola 1982) (Table 1).According to the soil classification used in Finland (Juusela and Ware 1956), the topsoil of the sandy field was loamy fine sand. Even though the topsoil ofclay field had a high clay fraction, it had 9.4% organic matter, as the organic carbon content was multiplied by 1.724 (Allison 1969). Hence, this soil type could be used for carrot pro- ductionand for the purposes of this study. The mull field was rich in clay fraction which made it sus- ceptible to compaction. The experimental soils were analysed for potassium, phosphorus, calcium and magnesium 150 Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND Table 1. Particle size distribution and organic Cof fine sand, clay and mull at experimental site (%). Soil Particle size fractions (|im) depth, cm <2 2_ 20- 60- 200- 600- Org. C 6 20 60 200 600 2000 Fine sand 0-25 13 3 4 9 33 36 2 2.4 25- 31 6 7 16 23 16 1 0.9 Clay 0-26 75 7 5 7 3 2 1 5.8 26- 78 6 6 7 2 I 0 0.7 Mull 0-27 81 7 4 4 1 2 2 18.1 27- 85 7 4 3 1 0 0 6.9 extractable in acid ammonium acetate (pH 4.65) (Vuorinen and Mäkitie 1955, Kurki et al. 1965), and pH and electrical conductivity (EC) in wa- ter suspension, at the Institute of Soils and En- vironment at the Agricultural Research Centre ofFinland. The boron content was determined by the azomethine-H method (Sippola and Erviö 1977). These characteristics of experimental top- soils and subsoils before (April 1989, for mull 1990) and after (November 1991) the study are presented in Table 2. At the beginning of the field experiments the nutrient contents for fine sand Table 2. Chemical characteristics of experimental soils before (a) and after (b) field establishment. Soil pH EC Ca K Mg P B depth, cm lOmScnr 1 mg dnr 3 air-dried soil Fine sand 0-25 a 6.8 0.65 2510 210 112 81.8 0.94 b 6.7 0.68 2240 125 129 70.5 1.42 25- a 6.9 0.55 2000 151 147 21.8 0.73 b 6.9 0.74 2350 145 651 6.2 0.60 Clay 0-26 a 6.3 0.83 4200 452 742 18.9 1.35 b 6.3 0.71 4230 436 765 19.5 1.43 26- a 6.4 0.64 3250 272 1146 2.8 0,70 b 6.4 0.70 3390 276 1265 4.4 0.71 Mull 0-27 a 5.5 0.46 2470 290 232 5.8 0.50 b 5.3 1.33 2450 235 221 7.6 0.94 27- a 5.4 0.59 2530 266 266 4.1 n.d.'> b frö E2O 2650 189 818 3.7 0.40 11 n.d. = not determined 151 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND and clay soils were good on average, even though the magnesium content in sandy soil was only fairly good, according to the common classifi- cation used in Finland (Soil Testing Laboratory ofFinland 1990). In mull, both phosphorus and boron contents were only fairly good. 2.2 Treatments sector 90°). Thus, one sprinkler was needed to irrigate each main plot. Rate of irrigation was 5 mm per hour. The amount of water given to ever ry sprinkling sector was controlledby eight plas- tic flasks eguipped with funnels (Elonen et al. 1967). The dates of irrigation and quantities of water applied are presented in Appendixes 1-2. The water applications were rather uniform in- side the main plots with a variation less than 20% (1989), 15% (1990) or 25% (1991). The field experiments were established as a split- plot design with four replicates (Appendixes 1— 3). The effects of irrigation were studied in the main plots which were divided into five subplots according to different soil mechanical treatments before sowing as follows: A. Sprinkler irrigation A 0 = 0 mm A 1 = 2 x 30 mm B. Mechanical treatment Loosening B = narrow ridges B 2 = rotary harrowing B 3 = no traffic (untreated) Compaction B 4 = one pass of the tractor wheel B 5 = three passes of the tractor wheel 2.2.1 Irrigation Irrigation treatments were studied in sandy and clay soils and consisted of no irrigation (A0 ) and sprinkler irrigation (A ( ). Irrigation was not performed on mull. One irrigation of around 30 mm was given when the water content at a depth of 15 cm had been depleted to around 50% of the plant-available water capacity, as meas- ured by the gypsum block method (see 2.5.1). Irrigation was performed at night by rotary sprinklers (radius 14 m and angle of irrigated 2.1.2 Soil mechanical treatments Spring tillage treatments in the sub-plots were set up on autumn ploughed land, by loosening (B [-B 2) or compacting (B 4-B 5 ) the experimen- tal field area before fertilization and sowing. All tractor operations were performed by the con- trolled traffic concept. The distance of both the rear and front wheels was adjusted to 2 m which was the widthof the sub plots. The tractor wheel was never allowed to compact the experimental area in treatments Bj-B 3 . Soil loosening After harrowing to the depth of 5 cm, four nar- row ridges were prepared for treatment B, by means of coulters used for potato planting. The ridge tops were slightly compacted by rolling (10 kPa) to the state which could carry the sow- ing units. The final height of the loose ridges was around 10cm and the ridge distance 45 cm, which was the carrot row distance in all subplots. In 1991, the ridges were 1-2 cm higher than in 1989-1990 because the design and the compac- tion of ridges were repeated. In the B, plots, the soil was loosened by a rotary harrow to a depth of 20 cm (clay 15 cm). The top surface was af- terwards slightly compacted by a roller. Soil compaction For treatments 84-B4-B 5 , wheel traffic was imposed on ploughed land by a tractor with a harrow which was kept up. In treatment 84,B 4, the subplot 152 Pietola, L.: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND was established by consecutive passes across the plot with the rear wheels compacting the entire subplot. Because the track distance was kept at 2 m, and the width of one rear wheel was 33 cm, six passes compacted a 4-m wide lane. A buffer zone of 2 m was left at the outer sides of the main plots (Appendixes 1-3). In treatment 85,B 5 , the subplots were recompacted twice (18 drives). The rear axle load of the tractor with the harrow was 3030 kg. Because the diameter of the rear wheel was 145 cm, the contact area of the rear wheel was 1290 cm 2 according to the equation of Inns and Kilgour (1978, ref. Soane et al. 1980): S = 0.87b, x 0.31d (2) where S is the contact area, b the section width C and dc the diameter of a wheel. This equation is designed for a hard surface. Because the soil surface of clay and mull sank during the first pass, this equation could not be applied to these soil types in treatment 84,B 4, but it was applicable to all soils after the first pass. According to equa- tion (2), the ground pressure of the rear wheel was 120 kPa, thatof the front wheel 60 kPa. The pressures used in the present study were low as compared to those applied, e.g. in the United States where the ground pressures of agricultur- al vehicles can be as high as 490 kPa (Gupta and Larsson 1985). 2.3 Establishment and manage- ment of field experiments Prior to establishing the experiments in early spring when soil was moist and susceptible to compaction, the entire study area was ploughed to a depth of 25-26 cm each autumn. The first operation was 1-3 passes with a tractor wheel, which was carried out under the same moisture conditions each experimental year. Soil compac- tion and sowing dates are presented in Appen- dixes 1-3. 2.3.1 Fertilization and spring tillage After soil compaction, fertilizers were drilled at the rates of N 80, P 50, K 160 kg per hectare to fine sand and clay, and N 30. P 70, K 170 kg per hectare to mull, in accordance with general rec- ommendations (Soil Testing Laborotory ofFin- land 1990). In fine sandand clay fields, the com- pound fertilizer was applied with a fertilizer drill to the soil surface. In mull, nitrogen was broad- casted. The fertilizers were harrowed in treat- ments B ( and 83-B 3 -B s to the depth of 5 cm, being 1-2 cm shallower in the most compacted treat- ments. After harrowing, the loosening treatments B, and B 2 were performed (see 2.2.2). 2.3.2 Sowing and management during growth For one subplot, seeds were sown in four row beds 10 m long with 45 cm between rows. The carrot (Daucus carota L.) cultivar grown in these experiments was Nantes Duke Notabene 370 Sv which has been used also in other carrot investi- gations in the Nordic countries concerning land preparation (Taksdal 1984) and fertilization (Evers 1988). Seeds (in 1989-1990 coated) were sown 1 cm deep and 6 cm wide with a Nibex sowing machine with four units. The seeding depth ranged 0.5-1.5 cm, with a greater depth on loose organic soil and smaller depth on com- pacted clay soil. In 1991, a pneumatic Gaspor machine was used to avoid gaps in the loose sub- plots. Later, thinning was doneat the same time for all plots of one field to 40 plants per I m. Flowever, in 1989, emergence on clay was une- ven and slow due to the lack of rains following planting although the whole field was irrigated (5 mm). Plant populations were thinned to 20 plants per metre. Weeds were sprayed with linuron (only in 1989) and methoxuron (1990-1991). In addition, setoxidim was used for protection against couch grass (1989-1990). Couch grass was a problem especially in the clay soil field. Also hand weed- 153 Voi 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Table 3. Weather conditions in Jokioinen in 1989-1991 and 30-year averages. Month Mean air temperature (°C) Precipitation (mm) 1989 1990 1991 (1961-91) 1989 1990 1991 (1961-91) April 5.3 5.6 3.4 ( 2.4) 40 35 14 (31) May 10.4 9.3 7.2 ( 9.4) 41 22 29 (35) June 15.4 14.4 12.1 (14.3) 30 20 69 (47) July 16.3 15.2 16.6 (15.8) 85 85 55 (80) August 13.7 15.0 16.2 (14.2) 92 90 92 (83) September 11.0 8.0 9.1 ( 9.4) 51 62 80 (65) October 4.7 4.9 5.4 ( 4.7) 49 48 49 (58) Mean: Sum: 11,1 10.3 10.0 (10,0) 388 362 388 (399) ing was carried out during the growing season as needed. Dimethoate was applied 3-4 times against carrot psyHit, Iriosa apicalis, at the be- ginning of the growing season. Carrot fly, Psila rosae, was controlled by traps. 2.3.3 Sampling Both ends of each subplot were used for soil and plant sampling (Appendix 4). The final carrot yield was measured from an area in the middle ofeach sub-plot (length 6 m) where no sampling had been carried out. Because the sampling areas used for the cylinders in 1989-1990 were no more useful due to mixed soil, an exception was made in 1991.The plastic cylinders forroot sam- pling (see 2.7) were fixed inside the yield area of subplots for treatments B, and 85.B 5. From four carrot rows, only the two inner rows were used for plant sampling and harvest. 2.4 Weather conditions After the dry early season in 1989, the growing season was favourable for growth. Precipitation was slightly higher than normal in July and Au- gust, and September was very warm (Table 3). In 1990, like in 1989, the soils were too dry for good emergence. Dry June was followed by a moderately rainy season, and August was warm. September was, however, cooler than in 1989, with a difference of 3°C in mean temperature. In 1991, the soils were moist enough during emergence which was, however, much slower than in 1989-1990because of low temperatures in May and June. Only late July was dry. During the warm and rainy autumn, carrot growth was very intensive and recovered soon after the cool early season of 1991. 2.5 Soil measurements 2.5.1 Soil water content Electrical resistance The moisture condition of the soil during the growing season was studiedby the gypsum block method (Bouyoucos 1954). Immediately after sowing in 1989,blocks were dug to a depth of 15 cm in subplots B p B, and B 5 (Appendix 4). 154 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot In 1990, the blocks were also placed to 30 cm. In 1991, blocks were placed in all subplots to both depths. Blocks were located in the wall of a hole 8 cm in diameter. After filling the holes, some seeds were resown. In fine sand and clay, electric resistances of gypsum blocks were meas- ured 2-3 times a week, but only once a week in organic field where the moisture condition var- ied least. According to measurements in a pres- sure chamber, the blocks used in 1989registered soil water contents of 93-94% of plant-availa- ble water at field capacity (matric potential of -10 kPa), whereas new blocks used in 1990- 1991 showed only 89-90%. According to Aura (1985), this type of gypsum blocks register the wilting point (-1500 kPa) at 5%. Sampling and drying In addition, soil water content was determined by gravimetric analysis (105°C). On the com- paction day, immediately before mechanical treatments, undisturbed soil cores 5.05 cm in inside diameter were obtained from B 3 plots at 2.5-cm intervals from a depth of 0-20 cm and at 5-cm intervals from a depth of 20-40 cm using a soil core sampler (Heinonen 1960). Samples of top soil were taken with two replicates per plot (2 x5O cm3). Later this sampling method was used together with the measurements of gamma ray transmission (see 2.5.2) and pene- trometer resistance (see 2.5.4). Additionally, in 1991 the effect of soil mechanical treatments on soil moisture before emergence was studied by means of this core sampling. 2.5.2 Soil dry bulk density Gravimetric analyses Dry soil bulk density was determined by gravi- metric analysis (ratio of soil dry weight to its total volume) from core samples which were also used for soil water contents and pore size distri- bution (see 2.5.1 and 2.5.5). Gamma ray transmission In 1990, the dry bulk densities of fine sand and mull were measured to a depth of 35 cm at 2.5-cm intervals according to the principle of gamma emission using a French signal genera- tor (Stengel et al. 1986). Before soil irradiation, two parallel tubes of aluminium (one for a source ofradiation, another for crystal indicator, inside diameter 3.5 cm, length 40 cm, distance from each other 30 cm) were fixed in soil. During the measurement, soil was irradiated with a source of l37 Cs, and the quantity of photons transmitted through the soil, during 30 seconds, was record- ed by a counter. The determined soil length was 30 cm and the maximum radius of effective vol- ume of measurement 2 cm (Stengel et al. 1986). The maximum radius was located in the middle of the irradiated area, where the gravimetric measurement ofsoil water content was made (2 x 50 cm3 ). The relation between the quantity ofphotons and the bulk density of a substance was deter- mined in two substances of which the densi- ties were known, an organic substance named Technyl (p = 1.26 g cm 3) and concrete (p = 2.49 gcm 3). The equation of calibration for the tube distance of 30 cm and transmission time 30 seconds, was determined as: (3)p c = -0.664 In C + 5.96 where p c is the bulk density of a substance in a certain moisture condition and C the number of photons per second penetrating the substance. The effects of soil water content were eliminat- ed by means of the equation of Double-sonde Gammametrique L.P.C.-I.N.R.A. (1985) and Stengel et al. (1986): (4)p b = Pc /(I + 1.1 VV) where p b is the dry bulk density and w is the soil gravimetric water content. For w, the gravimet- ric water content in soil between the aluminium 155 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND tubes was analysed at 2.5-cm intervals by the sampler of Heinonen(1960) after irradiation (see 2.5.1). In all measurements, a tube distance of 30 cm was used. At this distance, high soil wa- ter contents do not interfere with soil density analyses (Stengel et al. 1986). According to the measurements in Avignon (Double-sonde Gam- mametrique L.P.C.-I.N.R.A. 1985), the empiric relation is rectilinear between densities of 1 and 2.5 g curl In the present study, the bulk densi- ties were within these limits, even in loose or- ganic soil where the “moist densities” (p c ) were twice as high as the moisture-corrected dry bulk densities (p b ). 2.5.3 Degree of soil compactness For determination of dry bulk density (p d) for fine sand and mull, dry bulk densities at the soil depth of 7.5-17.5 cm were averaged, based on the 50-cm3 coring and gamma ray transmission (see 2.5.2) in July-August 1990. Dry bulk den- sities from depths of 2.5-7.5 cm and 12.5- 17.5 cm, based on 200-cm3 coring in September 1989 and 1990, were averaged for comparison including clay. For the determination of dry bulk density at thereference state (p dp ), a soil sample from each field experiment replicate was collected from the plough layer (20 x 20 x 25, 26 or 27 cm for fine sand, clay and mull, respectively) in May 1991. For the preparation of a sample representing one field, four moist samples were pooled and 40 dm3 of soil was kept in a plastic bag to retain the moisture until analysed in Sweden at the Agri- cultural University of Uppsala by the method of Håkansson (1990). A soil sample (30 dm3 ) of a moisture of field capacity was compacted by an unaxial pressure of 200 kPa (surface area of soil 0.1 m 2) for about one week, until no water came out from the water outlet. The maximum soil dry bulk density was then analysed with four repli- cates. The degree of soil compactness (D = 100 p d / p dp ) was then expressed for different mechanical treatments. 2.5.4 Soil penetrometer resistance A recording cone penetrometer described by Anderson et al. (1980) was used to measure soil mechanical impedance to a depth of 52 cm at 3.5-cm intervals. The kilograms shown by the penetrometer were converted into pascals with the coefficient 76.2, as the diameterof cone was 1.29 cm (Instruction Manual for Use of Bush Recordings Soil Penetrometer 1979). The cone index in the profile of each subplot was meas- ured four times, from two inner rows at both ends near the area reserved for yield measurements. The medians were calculated from four meas- urements of two inner rows. Medians were used instead of means because of high spatial varia- bility in penetrometer resistances of subsoil. In 1989, the penetrometer resistance was measured only once, at the time of vigorous growth of car- rot tap roots, with only gypsum block record- ings for soil water content. In 1990-1991, the cone indexes were taken three times per season, simultaneously with gravimetric determination of soil water content. 2.5.5 Soil pore size distribution Undisturbed soil cores (inside diameter 7.1 cm, height 5 cm) were collected in metal cylinders of 200 cm3 in autumn before harvest between two inner carrot rows in the sampling area. The sam- pling depths were 2.5-7.5 cm, 12.5-17.5 cm, 22.5-27.5 cm and 32.5-37.5 cm. In 1989, only one sample per each subplot (B 2-, B ? - or B 5) was taken at a given depth, but in 1990 the samples were collected from all subplots with two repli- cates. Samples of treatment Bj were taken only from two upper layers and near the carrot row. Pore size was divided into three categories: >3O pm, 0.2-30 pm, <0.2 pm. These equivalent pore diameters corresponding to the matric po- tentials of -10 kPa to -1500 kPa were calculat- ed on the basis of the corresponding matric suc- tions and effective pore size drainage using the capillary rise formula (e.g. Aura 1975): 156 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot d = 0.3 / h (5) where d is the equivalent pore diameter(cm) and h is the pressure head (cm), i.e. h = 2 y cos a / r p, g (6) where y is the surface tension between the liq- uid (water) and the air, a the contact angle (for pure water and clean surfaces a —> 0), r the radi- us of the capillary, p ( the density of liquid water and g the acceleration of gravity (e.g. Hillel 1971).Later, the pressure head (h) is termed mat- ric potential (v)/m ). For determination of soil moisture at field capacity (\|/ = -10 kPa), samples in metal cyl- inders were saturated by wetting from the bot- tom, the water level being in the middle of the soil core. Cylinders were placed into a pressure chamber, using the equipment describedby Aura (1983). Soil water retention at the matric potential of wilting point (\j/m = -1500 kPa) was deter- mined using a tension caused by osmosis (Aura 1975). After premoistening, samples (9 g) from metal cylinders of 5 cm3 collected in the field near the large cylinders were placed into semi- permeable plastic tubes in polyethylene glygol solution.After two weeks, soil water contents in samples of potentials of-10 and -1500 kPa were analysed gravimetrically. Soil-specific densities were determined by pycnometry (Blake 1965). 2.5.6 Soil air composition Soil oxygen and carbon dioxide contents were measured at the depths of 15 cm (1989-1990) and 30 cm (1990) from both A () and A, plots in three replicates. In 1989, air samples were col- lected from the B 2 and B 5 subplots, whereas in 1990, subplots B 2 (the most loose soil) and B 5 were analysed. After sowing, a porous cup (di- ameter 2.5 cm, height 3.5 cm) was placed at the bottom of a hole 2.5 cm in diamerer in a carrot row in the sampling area, beside a gypsum block (Appendix 4). During growing seasons, air sam- ples of 1.5 cm 3 were taken up from cups by sy- ringes at a constant rate 0.5 cm 3 s' 1 (after one extra air sucking) once a week (1989), or every two weeks (1990). Soil water contents were measuredby means of gypsum blocks on the day of air sample collection. Air composition was analysed by gas chromatography (Porapak Q and Molecular Sieve 5A columns and TC-detector and He-carrier flow) at the University of Hel- sinki, Department of Applied Chemistry and Microbiology (Jaakkola et al. 1990, Simojoki et ai. 1991). Retention times for C02 and 02 were 0.61 and 1.36 minutes, respectively. 2.6 Plant measurements 2.6.1 Sampling Carrot tap roots and shoots were sampled from two inner rows in all subplots twice before har- vesting from the sampling area and at harvest from the yield area (Appendix 4). The first sam- ple was collected when tap roots began to swell, i.e. 7-8 weeks before harvest. The second sam- pling was performed 3-4 weeks before harvest, when tap roots were almost full size, but inner quality was still developing (see 1.2). The third sampling was done at the time of harvest. The fresh weights of shoots and washed tap roots were recorded. After analysis of the exter- nal quality (see 2.6.4) of tap roots, they were grated and frozen (-20°C) in plastic boxes of500 ml (2 boxes per sample) for analysis of internal quality (see 2.6.5). When the sample size was over 1.5 kg, it was not possible to grate the en- tire sample. Only whole representative carrots were used for grating. An additional frozen sam- ple of carrot slices (300-400 g) was prepared for determination of carotene, but only from sam- ples collected at harvest. In 1989, a sample con- sisted of 15 shoots or tap roots. In 1990-1992, 157 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND 20 carrots were sampled due to the great varia- tion among plants. In 1990-1991, only non- branched and non-split, medium-sizeroots (50- 250 g) were used for internal quality analysis. 2.6.2 Yield Harvesting was doneby pulling plants manually on 11,24and 17 September 1989-1991 from fine sand, and on 15 September 1989 from clay soil field. The mull field was harvested on 2 October 1990and on 23 September 1991. The sample area for final tap root yield was 12 m, i.e. 6 m from the two inner rows (Appendix 4). Shoot samples of 15 carrots (1989) or 20 carrots (1990-1991) were taken simultaneously with the tap root har- vest. After harvest, the weights and numbers of non-branched and non-split, branched, split and both branched and split tap roots were recorded. Non-branched and non-split roots were classi- fied together into three size classes: below 50 g, 50-250 g and over 250 g, and weights and quan- tities of carrots were recorded per size class. 2.6.3 Tap root quality External quality The length and diameter (maximum and 0.5 cm below crown) of 50 tap roots were measured from final yield for each treatment. In 1990- 1991, the measurements were performed also from the samples collected before harvest (first and second sampling) of 15-20 non-branched and non-split carrots. Each non-branchedand non-split sample car- rot (n = 15-20) was weighed separately in order to calculate the cylidrical index (C) (Bleasdale and Thompson 1963)by means of tap root weight (W), length (h) and radius of root top (r) as fol- lows: C = W / Jt r 2 h (7) In the present study, tap root diameters at the crown were much smaller than were maximum diameters. Therefore, cylindrical indices were calculated on the basis of maximum diameter(at 2-3 cm below the crown). The index measured the ratio of root fresh weight (W) and volume (V2 =7t r2 h), i. e. the volume of a cylinder given by carrot length and maximum diameter, as pre- sented in Figure 2. Because carrot contains around 90% water, root fresh weight is near root volume Thereby, the ratio for a cylindric carrot is near the value of 1, for a conical root it is near 0.33 (Thompson 1969).A cylindric shape is a characteristic of a Nantes-type carrot and an indication of excellent external quality. According to Rosenfeld et al. (1984), the sample size for measurements of the cylindri- cality index ofcv. Nantes Duke should be at least 50 carrots to minimize the effect of random var- iation. Thus, the sample size of the present study, i.e. 20 carrots per replication, might not have been large enough for statistically significant differences between group means. Nevertheless, an average of four replicates represents a sam- ple size large enough. Fig. 2. Cylindrical index of carrot tap root based on tap root fresh weight (W ~V ) to volume of cylinder (V,) ratio. The radius of cylinder (r) represents maximum diameter of tap root, and cylinder height is taproot length (h). A modi- fication of the cylindrical index by Bleasdale and Thomp- son (1963). 158 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot In addition, carrot shape was measured as root length to maximum diameterratio. The por- tions of small (below 50 g), heavy (over 250 g), branched and split roots of the total yield also illustrated the external quality. Internal quality Frozen samples for determination of root inter- nal quality, relating tap root ripeness and bio- chemical maturity, were analysed after harvest in November-January. Dry matter, crude fibre, juice, juice dry matter, sugar and carotene con- tents were measured in relation to fresh carrot weight. The dry matter and crude fibre content was analysed on the gravimetric basis (60°C 48 h, 105°C 24 h) (Wilde and Voigt 1955, Jo- kinen 1977). Crude fibre content, which is a good indicator of root firmness (Aubert et al. 1979) was determined as dry weight of the carrot sam- ple after juice extraction. Juice was spun from carrots with a Moulinex liquidizer at 9000 g for 10 minutes. Juice content in carrots was also re- corded because it is another good indicator of root firmness (Aubert et al. 1979). The dry mat- ter content for carrot juice, i.e. juice density, was analysed by refractometry (Bellingham + Stanley 60/95) with glucose calibration from 0-10% glu- cose. According to Aura (1985), the dry matter content of carrot juice measured by this method is only slightly higher (4%) than the sugar con- tent ofcarrot juice by colorimetry. Asample size of 100 g (grated and frozen carrot) was used in the above mentioned analyses. Sugar and carotene contents were analysed at the Food Research Institute at the Agricultur- al Research Centre ofFinland. Saccharose, glu- cose and fructose contents were analysed by gas chromatography (Li and Schuhmann 1981) only from samples collected at harvest (100 g grated and frozen carrot). Alfa and beta carotenes were determined from the tap roots from the harvest (300 g sliced and frozen carrot) by high-perform- ance liquid chromatography (Heinonen et al. 1988). Two subsamples were taken for the ana- lyses. 2.7 Studies on carrot fibrous root system 2.7.1 Destructive sampling In 1990-1991, the distribution and morphology of fibrous root system of full-grown carrots were studied by destructive sampling from large cyl- inders (inside diameter 30 cm, wall thickness 1.0 cm) of polyvinyl chloride (PVC) presented in Figure 3. After plant emergence, these cylin- ders were pressed into moist soil at the ends of B 2 and B 5 subplots (Appendix 4). Six carrots were left in each cylinder under field conditions surrounded by carrots growing in the rest of the plots. Cylinders were kept in the field during the growing season and removed manually 1-2 weeks after harvest in 1990 (170 and 180 days after sowing in fine sand and mull, respectively) and in 1991 at harvest (145 days after sowing), in order to measure the root system at full length and close to the time of biochemical maturation (see 1.2). In October 1990 and September 1991, the plastic walls of the cylinders were sawn length- Fig. 3. PVC cylinders for destructive sampling of carrot fibrous root system, placed into the fields for the growing season. Root sampling performed at 5 cm horizontal soil layers of experimental fields. White circles indicate carrot plants. 159 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND wise along two lines in order not to destroy the soil profile inside the cylinder. One half of the cylinder wall was removed and each soil profile inside the cylinder was partitioned horizontally into 10-12 cylinders at 5-cm intervals.For most of the profiles, the two deepest layers below 50 cm broke as they were removed from the field, and the data were collected only from the 10first soil layers from a depth of 0-50 cm. Each soil sample consisted of 3534 cm3 soil. Fibrous roots which were divided into ten strata were then collected with the soil into plas- tic bags where they were frozen (-20°C) until separated from soil. Carrot tap roots (i.e. the or- ange part of the swollen root system, diameter over 1 mm) were sliced at the same time with the soil profile, and tap root pieces in each sam- ple were separated from the soil and fine roots. Fresh and dry weights of tap roots were then measured with shoots. Prior to root separation from soil, the frozen samples were soaked in a solution of 0.015 M NaOH to disperse the clay and to wash the roots. The soil was washed from the root samples with a hydropneumatic elutriatior (Smucker et al. 1982) which separated any organic material which was less dense than the mineral fraction of the soil, retaining up to 99.4% of all roots >0.05 mm. Thus, a lot of debris was included with extracted fine roots especially in the sam- ples from the mull soil. Based on the determination of debris dry weight during video recording (see 2.7.3), the purity of sandy soil samples was 90% on dry weight basis after some manual pre-cleaning. Organic debris associated with the root samples from the fine sand were manually removed so that only 0.6% of the total debrisremained with the root sample during image analysis. In the mull soil samples, the amount ofdebris was very high. Even after some precleaning based on sed- imentation in water, the root content was 35% in 1990, and 50% in 1991. The major non-root residue was straw material from the oat crop in 1989. It took a lot of time (2-4 hours per sam- ple) to clean these samples for image analysis. Yet, in video recording, 19% (1990) and 9.5% (1991) of the debris was left, but only short straw and peat material, which was not included in the final analysis of root dry weight, root length den- sity or root surface area (see 2.7.3). Root samples for image analysis were mailed (5 days) in 50-100 ml of 15-20% ethanol to Mich- igan State University (USA) where theroots were cleaned and videorecorded by the end of the same year 1993, meanwhile storing the samples in the same solution at 4°C. Root samples of 1990 from sandy field were not included in the image analy- sis in the USA because these roots were used for dry weight analysis at the Agricultural Research Centre ofFinland in 1992. Similarly, in 1989, some prestudies for fi- brous root growth were performed on fine sand and clay soil. At that time, the PVC cylinders were lifted up from the field after carrot emer- gence and placed on wet coarse sand. During the warm and dry summer, the soil temperature was as high as 26°C at the depth of 15 cm, and the soil moisture status varied (35-50% v/v). This resulted in abnormal forking of tap roots, even in loose soil, and some fibrous root accumula- tion on the cylinder wall and cracks (Fig. 4). These samples were washed manually by siev- ing in water. Otherwise these root samples were collected as described above. Results for fibrous root dry weight distribution in clay soil are giv- en only for 1989 because no measurements of this soil type were performed after that year. 2.7.2 Dry weight For fibrous roots from mineral soils in 1989- 1990, dry weight determination (48 h at 60°C, 24 h at 105°C) was performed after final clean- ing, 1-3 days storing at4°C (1989 in water, 1990 in 100 ml of 0.1% formaldehyde) and wrapping in a cloth. The decrease of dry weight given by the extra drying at 105°C averaged 3% for fi- brous roots and 5% for shoots. The weight loss was as much as 10%for tap roots because of the large quantities of the material which had not dried completely during the first drying period at 60°C. Fresh and dry weights of tap roots col- 160 Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. lected separately from fine root samples were recorded at the time of sampling. For root samples collected from fine sand in 1991 and mull in 1990-1991, dry weight data on fibrous roots was recorded in Michigan State University (48 h at 70°C). After storing in etha- nol (15-20% v/v) for 2-6 months, some cellular weight may have been lost due to root tissue fix- ation with ethanol. 2.7.3 Root morphology Fibrous root lengths and widths from samples of fine sand of 1991 and mull soil samples were determined by automated image processing of video recorded roots (Smucker 1993) in the Soil Biophysics Laboratory, Department of Crop and Soil Sciences, Michigan State University, East Lansing, USA. Video recording Roots were dyed with Malachite green oxalate for 24-48 hours before video recording by in- jecting 2-10 ml (depending on the sample size) of 4% dye into the plastic storage bag contain- ing 15% ethanol. Stained roots were rinsed with water on an ultra fine (25 pm) nylon screen. Roots were then placed uniformly on a clear glass tray. Roots from sandy soil were recorded twice, in Procedure 1 (see next paragraph) for determination of length, and in Procedure 2 for measurement of width. Procedure 3 represents roots from mull field from which only 1/8 of the sample was video recorded because of the con- siderable amount of non-root residues, most of which were removed from the subsamples be- fore video recording. The results obtained by analysing subsamples were rather representative. As the mean root length and surface area per soil volume for three whole samples, containing large quantities of debris, were 0.773 cm cm 3 and 0.050 cm 2 cm 3 the means for corresponding sub- samples (1/8) were 0.764 cm cm' 3 and 0.048 cm 2 cm 5 , respectively. This resulted in 1.1% and 4% errors in measurement of these two root param- eters. Procedure 1. The samples of fine sand (3534 cm 3) taken from the depth of 10-50 cm were analysed as follows. Samples from soil surface (0-10 cm) were analysed only according to Pro- cedure 3 due to the high quantity of non-root residues (over 30%). At the depth of 10-25 cm the root length was so great that it was impossi- ble to place all the roots into the large tray (42 cm x 37.5 cm) without considerable over- lapping. Therefore, samples were split into 2-4 subsamples. Some manual cleaning was per- formed while placing roots uniformly on the tray containing a thin water film (3 mm) which was attained by adding 500 ml water into the tray. Fig. 4. Carrot fibrous root systems in 1989 inside cracks of compacted clay profile (above) and on the surface of com- pacted fine sand (below). Photos: Liisa Pietola. 161 AGRICULTURAL SCIENCE IN FINLAND The glass tray was illuminated from below by a translucent light table and video recorded by a computer-controlled robotic camera (Smucker 1990). Sixty images, i.e. 60% of the total area of the tray, were recorded on a VHS video tape, 19 trays per tape (117 min). For this given tray size, the images were recorded with a lens size capable of measuring the root length, but a root width less than 0.4 mm was not measured accu- rately (lens 12.5 mm, lens extension tube 10 mm). Therefore, Procedure 2 was conducted for accurate root width analysis. Procedure 2. After recording of roots in Pro- cedure 1, roots were subsampled from the mid- dle of the large sample glass tray for Procedure 2 by means of a 100-ml polyethylene syringe. Subsample size was 200 cm 2 with 60 ml solu- tion volume. The remaining root sample was rinsed and collected on a nylon screen and dried for 48 h at 70°C for dry weight determinations. The subsample was cleaned from some short debris of which dry weight was determined in order to measure the amount of debris per root dry weight. For this given tray size (18.5 cm x 19.5 cm), 70 images per tray (70% of the area) were recorded, 17 trays per tape, with a lens size cabable of recording images for precise width determination of roots >0.15 mm in diameter (lens 12.5 mm, 4 x magnification, lens exten- sion tube 10 mm). After video recording, roots were dried and dry weight was determined. Root length and surface area (see Image analysis) of a subsample were then multiplied by the ratio between root dry weights in the whole sample and the subsample. Procedure 3. After the first cleaning by sed- imentation in water, the whole sample material was placed uniformly on the big glass tray used in Procedure 1. The subsampling was performed as in Procedure 2. The rest of the material, con- sisting of roots and non-root residues, was dried and weighed. The subsamples were cleaned and analysed as in Procedure 2. The debris content was determined on a dry weight basis. The re- maining short debris was evaluated visually by the same person. During the computer image analysis, material which was less than 4 times longer than its width was identified as non-root debris and discarded from the root data (Fergu- son 1994, personal communication). Thus, this kind of non-root residue was not a source of error in the measurements of carrot root morphology. Image analysis The images of stainedroot samples on the video tapes (12 for Experiments 1-2, and II for Ex- periment 3) were transferred to the Vicom im- age analysis system in the MSU Root Image Processing Laboratory (Smucker et al. 1987, Smucker 1990,1993). The binary image was cre- ated from the original grey scale image and dig- itized for measurement of root total length and individual lengths for each root width class (de- termined by the lengths of centre lines of origi- nal roots by using skeletonizing algorithms which contain information about root widthand length) (Fig. 5).For Procedure 1, the width class- es were 0.40, 1.21, 2.21, 3.64 and 6.12 mm as the resolution for each image was 100 pixels per 1 cm. For Procedures 2-3, the classes were 0.15, 0.40, 0.75, 1.25 and 1.95 mm, with an image resolution of 276 pixels per 1 cm in the Vicom pipeline and parallel computer image processor. The processing time for each image was 2-2.5 minutes in Procedure 1 and, due to smaller sam- ple sizes, 1.5-2 minutes in Procedures 2-3. The processing time was dependent also on the quan- tities of non-root residues in each sample. The final root length of a sample was the sum of all image lengths for all width classes. For root surface area (A), the calculated projected area (sum ofall image profile areas in all width classes) was multiplied by n, based on the cy- lindrical nature of roots and the formula: (8)A=2 n R L where R is root radius and L root length. Be- cause root roundness varied, being higher in sur- face soil and lower in subsoil and in compacted fine sand (based on visual observations during video recording), this approach has limitations. The three-dimensional estimation was, howev- 162 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot er, used because the cross-sectional area of roots was primarily round or at least elliptical. The two-dimensional measurement has been convert- ed to three-dimensional estimation also in other recent studies (Kokko et al. 1993). 2.8 Statistical analysis The statistical analyses were performed accord- ing to Steel and Torrie (1981) and Ranta et al. (1991). For analysis of variance, soil parameters and carrot growth in mull soil were studied ac- cording to the randomized complete-block de- sign (4 replicates) where loosening operations and wheel traffic represented the treatments. In mineral soils, data for soil air composition (3 replicates) and carrot growth (4 replicates, ex- cept 3 replicates at the first and second sampling in clay soil) were studied according to the split- plot design where two irrigation schemes repre- sented the main treatments. Group means were compared using Tukey’s HSD (Honestly Signif- icant Difference) test to find statistically signif- icant (P < 0.05) differences. The relationships between different soil pa- rameters and between soil properties and carrot growth were studiedby the linear correlationand regression analyses. The significances of the correlation coefficient (r) at the 95, 99 and 99.9% significance levels were marked *, ** and ***, respectively. In Figures 36-42, the splined lines connect- ing the data points were drawn by the CoPlot procedure (CoPlot Programme for Scientific Graphs 1990) based on the Bézier lines (Barn- hill and Riesenfeld 1974). Fig. 5. Root image on computer screen during image ana- lysis at two stages. This particular image represents roots from the loose fine sand, at 25-30 cm soil depth. Above: First stage ofanalysis where root image is created based on colour intensity of stained roots. Below: Further stage of image analysis where root images are skeletonized. Dis- tance(no. of pixels) between this centre line and root edges (outer line of root image) determines image widths. This image represents 1% of the 360-cm2 tray, from which 70 root images were recorded for a given sample. The width of roots is mostly around 0.15 mm and root branching uni- form. To the left there is apiece of thickerroot with laterals which were more abundant in compacted non-irrigated fine sand. Photos: Liisa Pietola. 163 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Results 3.1 Effects of treatments on soil physical properties 3.1.1 Soil moisture Clay soil was treated and sown as soon as the upper 5-cm layer was dry enough for the field operations. The fields of fine sand and mull were also established as soon as possible in order to take full advantage for carrot growth from the short growing periods in Finland. The volumet- ric water contents at compacting dates and the plant-available water capacity (moisture at mat- ric potentials between -1500 kPa and -10 kPa, untreated B 3 subplots) of experimental soils are presented in Figure 6. These results imply high wetness in soil profiles, especially for the clay and fine sand soils, when treated, which was cru- cial for maximum and more uniform soil com- pressibility. Results for the plant-available water record- ed by gypsum blocks in untreated B 3 subplots during the growing seasons (Figs. 7-9) suggest that irrigation was needed in mineral soils twice in each experimental year, but at different times of the growing season. In 1989, drought occurred at the beginning of July, but the next year earli- er in mid-June. In 1991, there was no lack of water until mid-July. On mull field, there was no need for irrigation, except for one week at the beginning of August in 1990 and 1991. The aim of irrigations, i.e. to maintain the soil mois- ture level above 50% ofplant-available water at the depth of 15 cm, was reached in 1990-1991 in fine sand. In 1989, the first water application was delayed, and the drought at the beginning of July affected all plots in fine sand and clay. The effect of irrigation on soil moisture was most distinct in 1990 during a long dry period when fine sand was irrigated. According to the soil moisture measurements before carrot growth in May 1991, loose soil Fig. 6. Volumetric soil water contents of experimental sites at different soil depths (in untreated B 3 plots) at compact- ing dates in 1989-1991. Plant-available water capacity between broken lines (matric potential from -1500 kPa to -lOkPa). 164 Pietola, L.: Effect ofsoil compactness on the growth and quality of carrot AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. profiles were drierthan compacted soil, but only if moisture was calculated on volume basis. Measurements from soil core samples indicated that the volumetric water contents of both fine sand and mull were higher than that of the more compacted soil, when averages of five samples (at 7.5-17.5 cm, 2.5-cm intervals) were com- pared (Table 4). Fig. 7. Moisture status by gypsum block recordings in fine sand field at the depth of 15 cm in non-compacted B, treatment in A„ plots without irrigation (circles) and in A, plots with irrigation (triangles) in 1989-1991. White columns indicate irrigation and black columns precipitation. 165 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot Fig. 8. Moisture status by gypsum block recordings in clay field at the depth of 15 cm in non-compacted B 3 treatment, in non-irrigated A 0 plots (circles) and in irrigated A, plots (triangles) in 1989. White columns indicate irrigation and black columns precipitation. Fig. 9. Moisture status by gypsum blockrecordings in mull field at the depth of 15 cm (circles) and 30 cm (triangles) in non- compacted B, treatment in 1990-1991.Black columns indicate precipitation. 166 AGRICULTURAL SCIENCE IN FINLAND Table 4. Water content (%) of loosened and compacted soils at the depth of 7.5-17.5 cm duringplant emergence in May 1991. 6 = volumetric water content, w= gravimetric water content. Soil Soil Tractor wheel passes (n) HSD005 loosening" 7 j Fine sand 6 23.527.6 30.331.1 2.7 w 20.821.9 22.021.1 n.s. 2 » Mull 0 42.144.7 50.853.8 3.5 w 73.273.3 72.871.9 n.s. "rotary harrowing 2) n.s. = not significant During vigorous plant growth, the gypsum block values of all other subplots did not differ significantly from those of B 3 plots in fine sand. In July-August,both in clay (P < 0.05, at 15 cm) and mull fields (P < 0.10 in August 1991, at 15 and 30 cm), the plant available water remained at the highest level in B 3 treatment (Fig. 10). This could indicate more efficient water uptake un- der compacted soil conditions and lower water supply in loose soil, which is shown by positive correlations between soil dry bulk density and soil volumetric water content (r = o.49*** for fine sand, r = o.9o*** for mull, based on soil 50-cm3 corings at 7.5-17.5 cm in July 1990 and August 1991, respectively). 3.1.2 Soil dry bulk density Gravimetric analysis The role of soil mechanical treatments on gravi- metric soil bulk densities of all experimental soils at four different soil depths is shown in Figure 11. These drybulk densities are based on samples of200 cm3 , collected in September 1990 (clay 1989) for determination of soil moisture at field capacity (see 2.5.5). As the clay soil dry bulk density was measured only by this 200-cnr3 sampling, the results in Figure 11 are given for comparison between soil types. In all three soils, the mechanical treatments did not affect the dry bulk density below the depth of 27.5 cm. Above this depth, soil dry bulk density increased from treatment to treatment 85.B 5 . In B 2 treatment, at the depth of 12.5-17.5 cm, the dry bulk density of fine sand was lower than in B, treatment. In clay soil, only one sample was takenper subplot which explains the great variation. Thus, differ- ences between treatments at the 95% significance level were found only in the surface layer (2.5- 7.5 cm), even though the increase of dry bulk density by soil compaction (B s) was as much as Fig. 10. Plant-available soil water (monthly averages of gypsumblock recordings) as affected by mechanical treat- ments (8,-B5) in clay and mull at the depthsof 15 cm (cir- cle) and 30 cm (triangle). White circles or triangles indi- cate non-irrigated plots (A (1) and black circlesirrigated plots (A,). B 1 = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B 3 = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD005 for significant differences: in July 11% (at thesame A level 1 6%). in August 15% (21 %) 167 Vol 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Fig. 11.Soil dry bulk density at four depths as affected by soil depth and soil mechanical treatments (B ( -B5 ) based on dry weights of core samples (200 cm 1) taken in September 1990 (for clay 1989). B, = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B 3 = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD (M), for significant differences: Fine sand: 0.06 g cm" 1at 2.5-7.5 cm, 0.09 g cm"3 at 12.5-27.5 cm Clay: 0.13 gcm ' at 2.5-7.5 cm Mull: 0.05 g cm' at 2.5-7.5 cm, 0.06 g cm 1 at 12.5-17.5 cm, 0.12 g cm' at 22.5-27.5cm 168 Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND Table 5. Dry bulk density (g cnr3 ) of loosened and compacted soils based on gravimetric core analyses in July-August 1990. Soil Soil loosening 11 Tractor wheel passes (n) HSD, 0.05 depth, cm I 2 0 1 3 Fine sand 2.5- 7.5- 12.5- 25.0- 30.0- 1.13 1.23 1.30 1.28 1.33 0.14 1.28 0.181.23 1.20 1.31 1.46 1.33 1.24 1.31 1.34 1.49 0.16 n.d.2) 1.49 1.59 1.53 1.49 n.s.3) n.d. 1.56 1.56 1.58 1.56 n.s. Mull 2.5- 7.5- 12.5- 25.0- 30.0- 0.62 0.61 0.61 0.64 0.68 n.s. 0.65 0.62 0.66 0,71 0.75 0.07 0.62 0.66 0.76 0.130.66 0,73 n.d. 0.58 0.62 0.72 0.61 0.16 n.cl. 0.47 0.43 0.61 0.51 n.s. 11 1 = ridge preparation, 2 = rotary harrowing 21 n.d. = not determined 3) n.s. =not significant Table 6. Dry bulk density (g cm'3 ) of loosened and compacted soils based on y-ray transmission in July - August 1990. Soil Soil loosening 0 Tractor wheel passes (n) HSD, 0.05 depth, cm 0 I 3 Fine sand 2.5- 7.5- 12.5- 25.0- 30.0- 1.30 1.33 1.37 1.47 0.07 0.071.25 1.33 1.44 1.54 1.28 1.37 1.45 1.57 0.13 1.58 1.68 1.60 1.63 n.s.” 1.69 1.69 1.66 1.66 n.s. Mull 2.5- 7.5- 12.5- 25.0- 30.0- 0.65 0.67 0,71 0.77 0.07 0.64 0.69 0.77 0.79 0.10 0.63 0.66 0.74 0.80 0.10 0.62 0.61 0.71 0.70 n.s. 0.48 0.50 0.72 0.68 n.s. 11 rotary harrowing 21 n.s. = not significant 0.2 g cm 3 at the depth of 12.5-17.5 cm as com- pared with B 2 treatment. The results were the same at the 50-cm3 cor- ing as the impact of mechanical treatments on dry bulk density was recorded in the plough layer (Table 5). The loosest soil was determined in B, treatment and the most compacted soil by three tractor wheelings (B 5 ). Based on this sampling, the effect of soil compaction went deeper in mull soil than in fine sand. 169 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Pietola, L.: Effect ofsoil compactness on the growth and quality ofcarrot HSDoos and averagedry bulk densities at 5-cm intervals in Table 6. ments as is shown by the core methods above. According to the measurements at 2.5-cm inter- vals, the mechanical treatments affected dry bulk density to a depth of 25 cm in fine sand, but deep- Gamma ray transmission The dry bulk density measured by gamma ray transmission differed by soil mechanical treat- Fig. 12.Soil dry bulk density as affected by soil depth and soil mechanical treatments (B2-B,) based on y-ray transmission in July-August 1990. B, = soil loosening by rotary harrowing, B, = untreated B 4 = one pass of the tractor wheel, Bs = three passes of the tractor wheel. 170 AGRICULTURAL SCIENCE IN FINLAND er in mull (Fig. 12). Table 6 shows these dry bulk densities based on gamma ray transmission from the same sites which represented core sampling in Table 5. Gravimetric analysis vs. gamma ray transmission Tables 5 and 6 indicate that the gravimetric core samples of 2 x 50 cm 3 gave the lowest bulk densities for a given treatment at a given depth. In sandy soil, dry bulk densities by gamma ray transmission were in loose plots on average 0.05 g cm 3 higher than after the core analysis. In compacted soil, the difference was 0.1 gcm 3 . Similarly, in mull the differences were 0.03 and 0.05 g cm 3 , respectively. In mull, the relationship between dry bulk densities (at 2.5-17.5 cm) by core sampling and gamma ray transmission was a little closer (r = o.73***) than in fine sand (r = o.69***). There was a negative relationship between dry bulk density by gamma-ray transmission and soil gravimetric water content (fine sand r = -o.s6***, mull r = -o.3B***). For both soil types, the coefficient of determination (R 2 ) was slightly higher when soil gravimetric water con- tent (w, %) was included in the regression ana- lysis. For sandy soil, the regression equation (P < 0.001) for drybulk density measuredby the gamma source (p bga gcm 3 ) was: p bga = 1.13 +0,49 p bgr-0.02* (R 2 = 0.55) (9) where pbgr is the reference dry bulk density by the gravimetric method. For mull, the corre- sponding equation (P< 0.01) was Pb*a =O4B + 0- 66 Ph.gr" 0- 03VV (R: = 0.59) (10) According to Stengel et al. (1986), the dry bulk densities analysed from the irradiated area by core samples of 4959 cm3 (p bgr ) were very closely correlated with the dry bulk densities (pbga ) given by gamma sounder (n = 28): P Ma =0 %6 P h .8r- 0 003 (R 2 = 0.94) (11) Without elimination of the effect of soil moisture, the coefficients of determination (R 2 ) were 0.47 and 0.53 for fine sand and mull, respectively. These low coefficients of determi- nation of 96 cases imply a heterogeneity of the soil, as the method of gamma ray transmission is compared with two small-core samples of 50 cm 3 located in the middle of the 30 cm wide irradiated area at 2.5-cm depth intervals. In or- der to have absolute values for soil dry bulk den- sity by a gamma sounder, soil moisture should be analysed very accurately. For example, if the gravimetric soil water content of fine sand were 25%, but had been measured as 20%, the correct dry bulk density 1.25 g cnr3 would have been 1.31 g cm 3 , assuming the bulk density without moisture calibration to be 1.60 g cnr 3(see Eq. 4, 2.5.3). These data suggest that increasing soil water contents may obliterate the accuracy of evaluating soil bulk density by gamma ray trans- mission. 3.1.3 Degree of surface soil compactness The dry bulk densities of very compact plough layers at the reference state (p dp, Håkansson 1990) were as follows (volumetric water con- tents after compression in parentheses): Fine sand 1.589 (22.0) Clay 1.156 (45.2) Mull 0.803 (74.7) The different degrees of soil compactness, where dry bulk densities in each treatment were calculated on the basis of both the core method and gamma ray transmission, confirmed that soil compactness increased by one tractor wheeling relatively more on mull than on fine sand (Fig. 13). In addition, these results exemplify the higher correlation between two methods of dry bulk density measurements in organic soil than in sandy soil, particularly in loose soil profiles. Compactness of the plough layer of clay soil, where no 50-cm3 coring or gamma ray analysis 171 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot was carried out, can be compared with that of other soils by means of dry bulk densities based on core samples of 200 cm 3 from the depths of 2.5-7.5 cm and 12.5-17.5 cm (see 2.5.5). The degrees of clay soil compactness averaged (D) 80.1, 81.6 and 92.0 in treatments 8,, B, and 8,, respectively (HSD |||)5 = 8.2). The respective val- ues for fine sand were 76.5, 78.3 and 88.0 (HSD ()()5 = 5.5), and for mull 76.1,80.5 and 92.5 (HSD ()05 = 9.1). These results showed that the 0-20 cm depths of soils are compressed more as the clay content increases (Table 1, p. 151). These degrees ofcompactness are, however, low- er than in Figure 13 for fine sand and mull, be- cause the compact zone at the depth of 7.5- 12.5 cm was not measured by 200-cm3 coring. 3.1.4 Penetrometer resistance Under dry soil conditions, soil compaction af- fected soil penetrometer resistance more than loosening treatments as compared with B, plots in all experimental soils (Fig. 14). The maximum resistances in B 5 treatment were 3.25 MPa (at 30 cm), 3.25 MPa (at 10 cm and 30 cm) and 2 MPa (at 30 cm) in fine sand, clay and mull, respectively. In Figure 14, the results of fine sand and clay soils represent the date of 5 July 1989, and mull soil 17 July 1990. Based on averages of the same measurements from the depths of 7.0-17.5 cm and 24.5-35.0 cm, the penetrome- ter resistance decreased by soil loosening and increased by soil compaction in fine sand also at the depth of24.5-35.0 cm, while the clay sub- soil remained unaffected (Table 7). Based on the measurements in 1990-1991, soil loosening decreased the penetrometer resist- ance to the depth of 35 cm in both fine sand and mull. Also irrigation decreased the resistance in fine sand to this depth (Table 8). According to these results of penetrometer resistance in dif- ferent moisture conditions of fine sand (Fig. 15) and mull (Fig. 16), soil moisture had a greater effect on penetrometer resistance in compacted Fig. 1 3. Soil compactness according to the method of Håkansson (1992), D value = relation between actual and maximal (at 200 kPa) soil bulk density (%) at the depth of 7.5-17.5 cm as affected by soil mechanical treatments (B -B ). B : = soil loosening by ridge preparation. B 2 = soil loosening by rotary harrowing, B =untreated, B 4 = one pass of the tractor wheel, B, = three passes of the tractor wheel. 172 AGRICULTURAL SCIENCE IN FINLAND Vol. 4. 139-237. Fig. 14. Penetrometer resistance of experimental soils under dry conditions as affected by soil depth and soil mechanical treatments (B ,-B,), on 5 July 1989 (for mull 17 July 1991).Plant-available water content by gysum block recordings (%) at the depth of IS cm in parentheses. B ( = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B 3 = untreated, B 4 = one pass of the tractor wheel,B, = three passes of the tractor wheel. HSD (M[ , and average resistances al the depths of 7.0- 1 7.5 cm and 24.5-35 cm in Tables 7 and 8(mull) 173 AGRICULTURAL SCIENCE IN FINLAND Table 7. Penetrometer resistance (MPa) of loosened and compacted soils. Measurements were done under dry conditions on 5 July 1989(plant-available watercontent at the depth of 15 cm by gypsum blocks 10% for fine sand, 30% for clay). Soil Soil loosening 1’ Tractor wheel passes (n) HSD 005 deP th - Cm 12 0 13 Fine sand 7.0- 0.51 0.45 0.63 1.00 1.46 0.41 24.5- 2,01 1.94 2.44 2.80 2.93 0.45 Clay 7.0- 1.24 1.22 1.41 1.93 2.86 0.81 24.5- 2.39 2.44 2.70 2.73 2.92 n.s.2 ’ " 1 =ridge preparation, 2 = rotary harrowing 2) n.s. = not significant Table 8. Penetrometer resistance (MPa) of loosened and compacted soils under different soil moisture conditions. Depth 1 =7.0-15.5 cm, Depth 2 = 24.5-35 cm. A 0 = non-irrigated. A, =irrigated. Soil moisture refers to plant-available water content by gypsum block recordings at depths of 15 cm and 30 cm. Gypsum Depth Soil loosening 1* Tractor wheel passes (n) HSD 005 biodt, ~ r ~ö i r % Fine sand, 13 June 1990 A» 1 55 1 0.76 0.34 0.84 1.76 2.76 0.46 80 2 1.95 2.02 2.44 3.14 2.90 0.69 A,: 90 1 0.28 0.10 0.32 0.53 1.06 0.42 90 2 1.96 2.27 2.19 2.30 2.89 0.75 Fine sand, 7 July 1990 A„: 45 1 0.77 0.20 0.54 1.31 1.60 0.32 30 2 2.85 3.02 3.16 3.83 >4 0.48 A,: 90 1 0,36 0.24 0.39 0.57 0.91 0.18 90 2 2.01 2.16 2.36 2.46 3.22 0.43 Mull, 13 May 1991 A„: 85 1 0.21 0.18 0.25 0.45 0.75 0.13 30 2 0.78 0.80 0.80 1.01 1.19 0.17 Mull, 3 July 1991 A„: 85 1 0.38 0.32 0.37 0.59 0,86 0,11 90 2 0.93 0.90 0.86 1.16 1.25 0.29 Mull. 17 July 1991 A„: 50 1 0.48 0.50 0.52 0.92 1,40 0.20 75 2 1.41 1.28 1.29 1.65 1.79 0.42 " 1 = ridge preparation, 2 = rotary harrowing 174 Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. soil (B s ) than in loose soil (B 2 ). The changes of penetrometer resistance between A 0 and A, plots in fine sand at the first measurement indicate an effect of water application of 10 mm because the penetrometer resistances were measured between two nights of irrigation.At the second measure- ment, resistances of subsoils of non-irrigated compacted fine sand plots were at the upper limit of the pressure which could be measured by the penetrometer (4 MPa). Relation to soil moisture and dry bulk density For sandy soil in 1990 at the depth of 7.5-17.5 cm, penetrometer resistance was related to soil gravimetric water content (r = -o.s9***) and to dry bulk density (r = o.79***). Soil dry bulk density alone explained the penetrometer resist- ance as much as 63% (R2 ). Only a slightly high- er coefficient of determination (R was attained by including soil moisture in theregres- Fig. 15. Penetrometer resistance of fine sand at different recording times in 1990 (I = 13 June, II = 17 July) as affected by soil depth, soil rotary harrowing (B 2 ) and three passes of the tractorwheel (B,) in non-irrigated soil (A () ) and in irrigated soil (A,). HSD II()V soil moisture and average resistances at the depths of 7.0-17.5 cm and 24.5-35 cm in Table 8. 175 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot sion analysis. The regression equation (P< 0.01) for penetrometer resistance (Q p, kPa), deter- mined by soil bulk density (p bgr , gcm 3) and gravimetric water content (w, %), was as follows: Qp=3455 p 100 vv - 1700 (R 2 =0.68) (12) Soil volumetric water content (0, %) had a negligible effect on this regression (P < 0.01): Qp =4915pbgr-0.80-3515 (R2 = 0.69) (13) For mull in 1991, a close positive correla- tion was found between penetrometer resistance and soil dry bulk density (r = o.B6***). The penetrometer resistance was best explained by soil dry bulk density (R 2 = 74%). Like in fine sand, only a slightly higher coefficient of deter- mination (R 2 = 77%) was attained by including soil moisture in the regression analysis. The re- gression equation (P < 0.05) for penetrometer resistance (Qp , kPa), determined by soil bulk density (pbgr , gcm 3) and gravimetric water con- tent (w, %), was as follows: Qp = 5870 pbgr +3O w - 4645 (R 2 = 0.77) (14) Fig. 16. Penetrometer resistance of mull at different recording times in 1991 (I = 13 May. II = 3 July. 11l = 17 July) as affected by soil depth, soil rotary harrowing (B,) and three passes of the tractor wheel (B 5 ). HSD(m 0 soil moisture and average resistances at the depths of 7.0-17.5 cm and 24.5-35 cm in Table 8. 176 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Table 9. Volume fraction (%) of macropores (>3O pm) in loosened and compacted soils. At the sampling time in September 1990 (clay 1989) soil moisture condition was near field capacity (t|/m » -10 kPa). Soil Soil loosening 0 Tractor wheel passes (n) HSD 005 depth, cm 12 0 13 Fine sand 2.5- 31.2 28.4 25.9 22.7 19.6 3.6 12.5- 23.4 26.9 24.3 18.8 17.5 5.8 22.5- n.d.2 ' 18.1 17.8 15.1 16.5 n.s. 31 32.5- n.d. 17.9 19.6 15.3 16.6 n.s. Mull 2.5- 35.2 30.2 24.9 24.6 17.1 4.5 12.5- 26.6 28.4 25.6 19.0 15.1 6,1 22.5- n.d. 27.8 25.0 20.0 17.7 5.6 32.5- n.d. 26.1 28.5 24.1 27.0 n.s. Clay 2.5- n.d. 32.9 34.5 n.d. 24.7 n.s. 12.5- n.d. 20.8 17.0 n.d. 8.4 n.s. 22.5- n.d. 13.1 9.0 n.d. 12.4 n.s. 32.5- iuL M 5A md. IX6 n.s. 0 1 = ridge preparation, 2 = rotary harrowing 2> n.d. = not determined 31 n.s. = not significant Also soil volumetric water content (6, %) had a negligible effect on this regression (P < 0.06): Qp = 3190 p hpr + 0.4 0 - 2840 (R2 = 0.77) (15) The decreasing effect ofhigh soil water con- tent on penetrometer resistance (Figs. 15 and 16, Table 8) was shown by these regression analy- ses only in fine sand with soil gravimetric water content. 3.1.5 Soil porosity and water retention capacity Pore volume Comparison of total porosity and pore size dis- tribution, as affected by soil mechanical treat- ments (Table 9, Fig. 17), indicates a significant role for soil compaction in the volume of ma- cropores (equivalent diameter > 30 pm). Even if the totalporosity decreased by increasing soil compactness, the differences between total pore volume were smaller than between macroporos- ity as a result of the increase of small (<0.2 pm) and, to some extent, of medium-size pores. in fine sand and mull, total porosity de- creased significantly (P < 0.05) by soil compac- tion to the depth of 27.5 cm while the volume of medium-size pores increased in comparison with loose soil profiles. To the depth 17.5 cm, the to- tal volume of small pores was 1-2 percentage units higher in compacted fine sand than in loose fine sand. At mull soil surface (2.5-7.5 cm), the total porosity of B, treatment was significantly higher than in B 3 treatment and in compacted B 4 and B 5 treatments. Here, too, compaction in- creased the volume of small pores to the depth of 17.5 cm. Moreover, at a depth of 12.5- 17.5 cm, even one mull soil wheel traffic in- creased the portion of small pores by 3-4 per- 177 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect of soil compactness on the growth and quality ofcarrot Fig. 17.Total porosity and volumetric pore size distribution of experimental soils as affected by soil depth and soil mechan- ical treatments (B ,-B,), based on core sampling near field capacity (i|/ m <•-10 kPa). B t = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B = untreated, B 4 = one pass of the tractor wheel, B, = three passes of the tractor wheel. HSD ()(), for significant differences: Total porosity: in fine sand 2% at 2.5-7.5 cm, 3% at 12.5-17.5 cm, 3% at 22.5-27.5 cm; in mull 3% at 2.5-7.5 cm, 4% at 12.5-17.5cm, 7% at 22.5-27.5 cm Macropores: in Table 9. Medium-size pores: in finesand 2% at 2.5-17.5 cm; in mull 3% at 2.5-7.5 cm Micropores: in fine sand 1% at 2.5-17.5 cm, 2% at 32.5-37.5 cm; in mull 3% at 2.5-17.5 cm 178 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Table 10. Volumetric water content (%) of loosened and compacted soils at = -10 kPa, collected near field capacity (v m = -10 kPa) in September 1990 (clay 1989). Soil Soil loosening” Tractor wheel passes (n) HSDoo; dePth ' cm I 2 0 I 3 ' Fine sand 2.5- 24.2 25.8 26.9 28.0 28.5 1.8 12.5- 27.0 26.0 26.8 29.1 29.0 2.7 22.5- n.d. 21 27.1 24.1 29.2 24.1 4.1 32.5- n.d. 26.3 22.0 28.7 25.7 n.s. 3 ’ Mull 2.5- 38.3 41.5 45.4 44.9 50.2 3.2 12.5- 44.7 43.6 44.9 48.0 49.6 4.4 22.5- n.d. 45.8 48.7 50.1 49.7 n.s 32.5- n.d, 50.0 47.6 49.5 49.1 n.s. Clay 2.5- n.d. 33.5 33.1 n.d. 37.6 n.s. 12.5- n.d. 41.0 43.1 n.d. 46.5 n.s 22.5- n.d. 45.6 47.6 n.d. 44.9 n.s. 32.5- n.d. 49.9 48.8 n.d. 45.4 n.s. ” 1 = ridge preparation, 2 = rotary harrowing 21 n.d. = not determined 3) n.s. = not significant centage units as compared with loose soil pro- files. For medium-size pores, significant differ- ences between treatments were found only in the surface layer. However, these differences were negligible as compared with the differences around 50% in macroporosity (Table 9). Despite the few statistically significant dif- ferences in clay soil porosity between treatments (because of lack of subsamples of a heterogene- ous soil type), the differences are worth men- tioning because of their magnitude. In the sur- face layer, three tractor passes decreased the to- tal pore volume by 5.3 percentage units as com- pared with B, treatment (HSD()()5 = 5.4). Again, at a depth of 12.5-17.5 cm, total porosity de- creased as much as 7 percentage units by three tractor passes (HSDhos = 9) as compared with the porosity of B 2 treatment. These differences were mainly due to the variation in the total vol- ume of large pores which was relatively most distinct at 12.5-17.5 cm (Table 9). In the sur- face layer, comparison of the macroporosity in compacted and non-compacted soils indicated a decrease in pore volume of 10 percentage units by soil compaction. Deeper, at 12.5-17.5 cm, this decrease was 12 percentage units as compared with macropores in loosened soil (B,). Like in fine sandand mull, the volume of small pores increased by clay soil compaction more than medium-size pores. At a depth of 12.5-17.5 cm, the volume of small pores increased by soil com- paction by as much as 5.4 percentage units (HSD (|(|< . = 6.4) as compared withrotary harrowing (B2 ). No effect of mechanical treatments on pore size dis- tribution was, however, observed below the plough layer ofclay soil. Water retention No remarkable effect of mechanical treatment on the plant-available water capacity was found, as indicated by the distributionsof medium-size 179 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot pores (Fig. 17). However, the effects ofmechan- ical treatments on water retention capacities on volume basis at the matric potential of-10 kPa were more significant, as indicated by the por- tions of pore sizes of 0-30 pm (Fig. 17 and Ta- ble 10). Also soil volumetric water content at a matric potential of -1500 kPa, i.e. the volume of small pores in Figure 17, increased by soil Fig. 18. Soil air02 and CO, contents of fine sand (above) and clay (below) as affected by soil mechanical treatments (8,, B,) and irrigation at the depth of 15 cm during the growing period 1989.A() =no irrigation (left), A! =sprinkler irrigation (right). B, = untreated, B, = three passes of the tractor wheel. HSD ()IW for significant differences: 0 ; : in fine sand 1.1% (at the same A level 1.5%) on 22 May, 1.0% (1.4%) on 29 May and 5 June, 0.9% (1.3%) on 12 June, 0.7% (1.0 %) on 19 June, 0.5% (0.7%) on June 26; inclay 2.0% (2.9%) on 29 May, 4.1% (5.7%) on 5 June, 3.4% (4.8%) on 12 June, 0.5% (0.7%) on 14 August-t September CO,: close to that of02 180 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. compaction, especially in clay and mull. These results suggest higher water retention for the ex- perimental soils after compaction than after loosening below the suction at field capacity (V(fm < -10 kPa). This was observed also by soil water content measurements on volume basis right after sowing in 1991 (Table 4, p. 167). Fig. 19. Soil air O, and CO, contents of fine sand at the depths of 15 cm (above) and 30 cm (below) as affected by soil mechanical treatments (B„ B 5) and irrigation during the growing period 1990. A 0 =no irrigation (left), A, = irrigation (right). B, =rotary harrowing, B 5 = three passes of the tractor wheel. HSDIM), for significant differences: 0,: at 15 cm 0.6% (at the same Alevel 0.8%) on 30 May, 0.2% (0.3%) on 25 June, 0.2% (0.2%) on 9 July and 6 August, 0.8% (1.1%) on 20 August, 0 3'i il) v , ion h September. .11 30 em differences significant hetween B treatments, except on 23 .Ink CO : close to that of O, 181 AGRICULTURAL SCIENCE IN FINLAND 3.1.6 Soil air composition Composition of soil air was measured at a depth of 15 cm in 1989 and at depths of 15 and 30 cm in 1990 during the growing seasons. Generally, the oxygen content was lower than normal in early summer when soil water contents were high (Figs. 18-20). The precipitation and irrigations (Figs. 7-9) decreased the oxygen content while soil compaction decreased the content most in irrigated plots. Air composition in fine sand was not, however, significantly affected by the irri- gations in 1989, when the drought at the begin- ning of July affected all plots. Oxygen content of fine sand decreased and carbon dioxide con- tent increased by soil compaction in June 1989, but only around one percentage unit as compared to untreated soil (B 3 ) or to narrow ridges Later, the measured differences of around 0.5 percentage units for both oxygen and carbon di- oxide contents between untreated (or loosened) and compacted soils were still statistically sig- nificant. Even though soil moisture affected soil air composition in early summer, two applications of water (2 x 30 cm) at the beginning of July did not affect the oxygen or carbon dioxide contents in clay soil. The first amount of water applied at the beginning of July 1989 was too small for the dry clay soil to maintain moist soil conditions for more than a few days. Thus, a second water application was neededafter 5 days. The dry pe- riod continued at the end of July. Thereafter pre- cipitation was enough for plant growth but did not create a difference greater than 1 percentage unit in oxygen or carbon dioxidecontent between loose and compacted soil. At the beginning of the growing period 1989, oxygen content at a depth of 15 cm was significantly lower in com- pacted (B 5 ) than in untreated (B 3 ) clay soil (Fig. 18). In narrow ridges (B ), soil oxygen content was about the same as in treatment B 3 (only 0.2- 0.3 percentage units higher, not presented in Fig. 18). This decrease in oxygen content by clay soil compaction was 2-5 percentage units, and it last- ed for about the first 4 weeks after sowing. Sim- ilarly, the increase of carbon dioxidecontent was 1.5-3 percentage units at the depth of 15 cm. With advancing growing period, the differences in oxygen and carbon dioxide contents dimin- ished between the mechanical treatments. Fig. 20. Soil air O, and CO, contents of mull at the depths of 15 cm (above) and 30 cm (below) as affected by soil mechanical treatments (B„ B 5) during the growing period 1990. B, = rotary harrowing, Bs = three tractor wheelings. HSDiHI, tor significant differences: O,: at 15 cm 2.4% on 30 May, 1.0% on 20 August, 0.6% on 6 September; at 30 cm 4.9% on 30 May, 2.5% on 9 July, 1.4% on 20 August CO,: at 15 cm close to that of O,; at 30 cm differences significant, except on 25 June and 23 July 182 Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND In 1990, soil air composition was measured at depths of 15 and 30 cm, from the most loose (B,) and compacted (B 5 ) fine sand and mull (Figs. 19-20). The effect of irrigation on soil air composition was more significant in 1990 than in 1989, reflecting the remarkable effect of irri- gation on soil water contents in fine sand in 1990 (Fig. 7). In fine sand, the decreasing effect of irrigation on oxygen content was found at the depth of 15 cm in July 1990, after two water applications of 35 mm in June. In August, the effects were significant only at P < 0.10 at the depth of 15 cm (even though the difference was over 3 percentage units between irrigated and non-irrigated plots). At 30 cm, however, the ir- rigations caused significantly lower oxygen and higher carbon dioxide contents than were meas- ured in non-irrigated plots. Comparison of oxygen and carbon dioxide contents in loose and compacted fine sand at the depth of 15 cm indicates differences of around one percentage unit at the beginning of the grow- ing period 1990, but also in early autumn with high soil moisture content (85% plant-available water). Comparison of irrigated and non-irrigat- ed soils indicates that soil air oxygen content decreased and carbon dioxide content increased by irrigation more in compacted than in loose soil. The interaction was significant on 9 July (at both depths) and 20August (at 15 cm only at P < 0.10). In irrigated soil at the depth of 30 cm, the greatest change in fine sand air composition with advancing growing period was found at the end of July when the oxygen content decreased on average to 17-18%, the carbon dioxide con- tent being 3%. This might indicate a high respi- ration of carrot plant, as the tap root growth was most vigorous at this time. In 1989, this effect was not significant because of the low plant den- sity, 20 carrots per row meter. The influence of soil mechanical treatments was not statistically significant due to the exceptionally great varia- tion between replicates. The results from mull field at the depth of 15 cm implied a statistically significant effect of soil compactness on soil air composition only at the first and the last two measuring dates in 1990 (Fig. 20). At the depth of 15 cm, the changes in air composition were to some extent higher than in fine sand, as oxygen content decreased and carbon dioxide content increased by 1.5-3 per- centage units by compaction (B 5 ) as compared with B 2 treatment. At 30 cm, however, a much more negative effect of soil compaction on soil oxygen content than in fine sand was observed in the early summer, when the decrease of oxy- gen content was 3-10 percentage units and the increase of carbon dioxide content 3-5 percent- age units. Later, the smaller differences of 1.5- 2 percentage units were still statistically signif- icant for carbon dioxide content, but only at P < 0.10 for oxygen content. 3.2 Response of carrot growth and yield quality to treatments 3.2.1 Yield and external quality of tap roots Biomass accumulation Visual observation of carrot emergence indi- cated a more even and faster shooting in the com- pacted plots than in loose soil. After plant thin- ning, some seedlings died, especially in loose mull in 1990, when the plant population re- mained lower (35 plants per metre) in narrow ridges than in other treatments (40 plants per metre). Additionally, seedlings grew much more slowly in mull than in fine sand. In fine sand in 1989, tap root fresh weight was on average 10 g higher when irrigated as compared with no water applications at first sam- pling (Fig. 21). Similarly, shoot fresh weight was on average 13 g higher. This difference between two irrigation schemes was found also at the sec- ond sampling, but at harvest it had diminished to 5 g because of the adverse effect of water ap- plications on shoot growth under intensively 183 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot compacted conditions. As late as at the second sampling, shoots in irrigated and three times compacted plots were still significantly greater than in the irrigated loose B 2 plot. Mechanical treatments alone did not have any significant influence on carrot biomass. However, interac- tions (P < 0.05) were found between compaction and irrigation in carrot tap root and shoot devel- opment, as indicated by the second sampling 137 days after sowing. Tap root and shoot fresh weights in B 5 treatments decreased only in irri- gated plots. An opposite effect was caused by one tractor pass (B 4). In clay, irrigation had a significant positive effect on root growth only at the second sam- pling, when tap root mean weight increased by 10 g by water applications (Fig. 21). On aver- age, irrigation did not affect either shoot growth or shoot to root ratio. Mechanical treatments had a very remarkable effect on carrot growth. The highest tap root fresh weight was attained by B 3 treatment, while both very loose and compacted Fig. 21. Mean shoot and tap root fresh weight (n = 15) as affected by soil mechanical treatments in clay and fine sand at three sampling times during the growing period 1989. DAS = days after sowing. A 0 = no irrigation (left), A, = irrigation (right). B t = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD (H)S for significant differences: Tap roots: inclay 12 g (at the same A level 17 g) 81 daysafter sowing (DAS), 23 g (32 g) 108 DAS, 30 g (41 g) 140 DAS Shoots: in clay 4 g (6 g) 81 DAS, 6 g (8 g) 108DAS, 4 g (5 g) 140 DAS 184 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Fig. 22. Mean shoot and tap root fresh weight (n = 20) as affected by soil mechanical treatments ) in fine sand and mull at three sampling times during the growing periods 1990 - 1991. DAS = days after sowing. A 0 = no irrigation, A! = irrigation. B t = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD„,„ lor significant differences: Tap mills 1990: in fine sand 4 g(at the same A level 6 g) 89 daysafter sowing (DAS), 21 g (29 g) 123 DAS, 35 g (50 g) 158 DAS Shoots 1990: in fine sand 2 g (3 g) 89 DAS, 4 g(6 g) 123 DAS; in mull 3 g 95 DAS Tap roots 1991: in fine sand 14 g (20 g) 145 DAS 185 AGRICULTURAL SCIENCE IN FINLAND clay soils had a negative effect on tap root growth at the first and second sampling. Later, at har- vest (third sampling), soil compaction (B s ) had a very adverse effect on mean tap root weight, especially under irrigated conditions. In narrow ridges (B,), tap root growth was favoured by ir- rigation. Shoot growth was affected by clay soil mechanical treatments almost like root growth. Under loose clay soil conditions, shoot biomass increased by irrigation, but in compacted clay soil it decreased by water applications. Interac- tion was statistically significant at the first and second sampling date. In 1990 in fine sand, tap root fresh weight increased significantly by irrigation (Fig. 22), on average 10 g and 20 g after the first and second sampling, respectively. This positive effect of irrigation was found also in shoot weights, dif- ferences averaging 8 g, 9 g and 6 g, after the first, second and third sampling, respectively. Comparison of tap root weights between loose and compacted fine sand indicates a statistically significant negative effect of soil compaction on root fresh weight at all sampling times. Shoot development was also adversely affected by soil compaction, but to a smaller extent and only at the first and second sampling dates. The late applications of water to fine sand in 1991 did not have any influence on carrot root weight, but shoot growth was still affected (Fig. 22). At the second sampling the average shoot fresh mass was 37 g without irrigation and 48 g with water applications. At harvest, the values were 41 g and 47 g, respectively. The soil me- chanical treatments did not affect statistically significantly tap root size until the third sam- pling. Earlier, at the second sampling, the nega- tive effect of soil compaction (B 5 ) was signifi- cant only at P < 0.10 as compared with B 2 treat- ment. In mull field, the modifications of biomass accumulationby mechanical treatments were not statistically significant (Fig. 22). Even the great difference in root fresh weight between two loos- ening treatments (B t = 117 g and B 2 = 141 g) at the second sampling in 1990 was significant only at P < 0.10. However, shoot growth in 1990 showed positive response to soil compaction (B 4 ) at the first sampling, when shoot growth was the most retarded in narrow ridges. In 1991, this tendency was measured in August as well when the shoot fresh mass averaged 28 g and 38 g in B [ and B 4 treatments, respectively (P < 0.10). Final yield Comparison of final carrot root yields per hec- tare in non-irrigated conditions in all experi- mental years (Fig. 23) indicates only slight or negligible influence of water applications on carrot production. In fine sand, the mechanical treatments had a slight effect on total yields. In 1989, the highest total yield was achieved by B 2 treatment. In 1990, the highest yield was produced by B 3 treatment, but the difference be- tween this yield and yields from B ( and B 5 treat- ments was significant only at P < 0.10. The effect ofsoil mechanical treatments on the yield of medium-size roots (50-250 g, non-split and non-branched) was very similar to the effects on total yields. In 1991, with no drought in early summer, the narrow ridges (B,) produced the highest yield of medium-size roots. Similarly to biomass accumulation, the most distinct effect of mechanical treatments on final tap root yield was found in clay soil (Fig. 23). On average, the lowest total yields were meas- ured in both irrigation schemes from loosened (Bj) and compacted (B,) soils. The total yield of B, treatment increased remarkably (27%) by ir- rigation, whereas irrigations decreased the total yield in compacted clay soil very dramatically (21%). This interaction was statistically signifi- cant only when yields ofall four rows were con- sidered. On the basis of the yield of two inner rows, the interaction was significant only at 11% level. Comparison of the yield of medium-size roots produced by three times compacted soil (B s ) and by all other treatments, including Bj plots, shows distinctly the very poor soil condi- tions of compacted clay soil for carrot production. In mull, carrot yields were affected by the mechanical treatments only in 1990, when the total yield was reduced remarkably by the B, 186 Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND Vol. 4. 139-237. treatment (Fig. 23). Also the lowest yield of medium-size roots was achieved by B, treatment, but differences between treatments were not any more statistically significant. Root size Among different soil types, roots sizing over 250 g fresh weight were the most abundant in Fig. 23. Yield of tap roots as affected by mechanical treatments of experimental soils under non-irrigated (A 0 ) and irrigated (A ( ) conditions. White columns = total yield, shaded columns = yield of non-split and non-branched roots of 50- 250 g. B, =soil loosening by ridge preparation, B, = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B, = three passes of the tractor wheel. MSI) nl , for significant differences: Total yield: inline sand 7 Mg (al thesame A level 8 Mg) in 1989; in clay II Mg(ls Mg): in mull 7.5 Mg in 1990 Yield ofnon-split and non-branched roots of 50-250 g: in tine sand 6 Mg (9 Mg) in 1989, 12 Mg(l7Mg)in 1990. 6 Mg (8 Mg) in 1991; in clay 13 Mg(18 Mg) 187 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot fine sand (Fig. 24). Irrigation affected root size in fine sand in 1990, decreasing the amount of small roots. Mechanical treatments had a signif- icant influence on the root size distribution in fine sand during 1990-1991. In 1990,the number of too heavy roots was reduced clearly by soil compaction, whereas the number of medium-size roots was significantly lower in loose soils (B- Fig. 24. Tap root size distribution, i.e. population counts in three weight classes from number of non-split tap roots as affected by mechanical treatments (8,-B,)of experimental soils under non-irrigated (A (| ) and irrigated (A,) conditions. B, = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B, = three passes of the tractor wheel. HSD nO, for significant differences: 250-g size class: in fine sand 5% (al the same A level 7%) in 1990, 3* (5%) in 1991; in clay 3% (4%) 50-250-g size class: in fine sand 7% ( 1 0») in 1990, in clay 1 3% (18%); in mull 8% in 1991 50-g size class: in fine sand 7% (10%) in 1990, 5% (7%) in 1991; in clay 13% (18%); in mull 8% in 1991 188 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Fig. 25. Tap root deformity, i.e. gravimetric contents of split, branched and both split and branched roots from total fresh yield as affected by mechanical treatments of experimental soils under non-irrigated (A (| ) and irrigated (A,) condi- tions. B, = soil loosening by ridge preparation, B, = soil loosening by rotary harrowing, B, = untreated, B 4 =one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD jWfor significant differences: Split lap mots: in line sand M7, (at Ihe same A level h", I m 19X9, s'Jf 17'/, ) in I WO: in clay 7% (10%); in mull 4% in 1990, 6% in 1991 »ranched lap roots: in fine sand 5% (6%) in 1989: clay 11% (16%); Bolh splil and branched lap roots; in fine sand 1% (2%) in 1989; in clay 5% (7%) 189 AGRICULTURAL SCIENCE IN FINLAND Table 11. Tap root length (cm) of carrots at third sampling of loosened and compacted soils in 1989. A 0 = non-irrigated, A, = irrigated. Soil Irrigation Soil loosening 0 Tractor wheel passes (n) HSD00J mean (A) 12 0 13 Fine A 0 17.3 17.0 16.1 15.5 15.4 1.2 16.2 sand A, 17.9 17.4 17.0 16.4 15.3 1.2 16.7 mean 17.6 17.2 16.6 16.0 15.3 0.9 Clay A 0 16.6 16.6 15.8 14.5 13.3 1.7 15.4 A, 16.7 17,0 16.7 14.6 13.5 1,7 15.7 mean 16.7 16.8 16.2 14.5 13.4 1.2 '* 1 =ridge preparation, 2 =rotary harrowing B 2) than in other treatments. The number of small roots (below 50 g), on the contrary, was the high- est in treatment (22.1%) and the lowest (14.3%) in B, treatment. In 1991, the number of too heavy carrots was lower in B than in B 2 treat- ment. For medium-size roots, no statistically sig- nificant differencesbetween mechanical treatments were found that year but, like in 1990, small car- rots were most abundant in B treatment. In 1991, the number of medium-size roots increased by 8 percentage units and the number of small roots decreased by 9 percentage units by the late water applications in three times compacted soil. Inter- action was statistically significant. Irrigation had no effect on root size in clay soil (Fig. 24). Clay soil compaction clearly in- creased the number of small roots. Similarly, the number of medium-size roots decreased to a great extent by three tractor wheelings. Only few heavy roots (> 250 g) were found in clay field, mainly in B ? treatment. In mull, root size was not affected by mechanical treatments in 1990, but in 1991 soil compaction reduced to some extent the number of medium-size roots, as com- pared with B 2 treatment. Root splitting and branching Among different soil types, root splitting and branching was most abundant in clay (Fig. 25). On average, splitting decreased by irrigation in clay soil, from 16.2% to 11.2%. In all soil types, the lowest number of split roots was recorded in loose soils, as soil compaction increased signif- icantly the number of split roots. Only in 1991 did sandy soil compaction not have any signifi- cant influence on root splitting. In the treatment of one tractor pass of clay soil, irrigation in- creased splitting while in three times compact- ed clay the effect was clearly opposite (P < 0.001 for interaction). Irrigation influenced root branching only in clay soil, where 16% of total yield was branched without water applications, but as much as 20% of the yield was branched when irrigated (Fig. 25). Soil compaction did not increase the number of branched roots as it affected root splitting. Both in fine sand and clay in 1989, the number of branched roots was the highest in B 2 treat- ment. In 1990- 1991, when the plant density was higher, mechanical treatments did not affect sig- nificantly root branching, although the highest average amount ofbranched roots was still found in B 2 treatment. The highest number of roots which were both split and branched was found in compacted clay (Fig. 25). Soil compaction increased the amount of split and branched roots also in fine sand in 1991, with irrigation in B 4 treatment and with- out irrigation in B 5 treatment. Some tap roots which were damaged at harvest were found in clay soil (5%), and less than 3% in other fields. 190 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot Vol. 4: 139-237. Fig. 26. Tap root length as affected by mechanical treatments (B -B )of mull and fine sand under non-irrigated (A (J ) and irrigated (A,) conditions at three sampling times in 1990-1991. B, = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD()||s for significant differences: Fine sand: 0.9 cm (at the sameA level 1.2cm) 89 days after sowing (DAS), 1.4cm (1.9 cm) 123 DAS, I.3cm(l.Bcm) 158 DAS in 1990; 1.2 cm (1.7 cm) 90 DAS, 1.1 cm (1.5 cm) 117 DAS, 0.9 cm (1.3 cm) 145 DAS in 1991 Mull: 1.7 cm 129 DAS, 1.5 cm 165 DAS in 1990; 1.5 cm 90 DAS, 2.1 cm 121 DAS, 1.3 cm 146 DAS in 1991 191 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality of carrot Fig. 27. Tap root maximum diameter as affected by mechanical treatments of mull and fine sand under non-irrigated (A () ) and irrigated (A,) conditions at three sampling times in 1990-1991. B, = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B 3 = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD nl), forsignificant differences: Fine sand: 0.2 cm (at the same A level 0.2 cm) 89 days after sowing (DAS), 0.2 cm (0.3 cm) 158 DAS in 1990;0.1 cm (0.1 cm) 145 DAS in 1991 Mull: 0.3 cm 95 DAS, 0.3 cm 129 DAS, 0.3 cm 165 DAS in 1990; 0.2 cm 146 DAS in 1991 192 AGRICULTURAL SCIENCE IN FINLAND Table 1 2.Tap root maximum diameter (cm) ofcarrots at third sampling of loosened and compacted soils in 1989. A 0 = non-irrigated. A, = irrigated. Soil Irrigation Soil loosening 0 Tractor wheel passes (n) HSD00, mean (A) 12 0 13 Fine A 0 4.0 4.0 4.1 4.0 4.3 0.3 4,1 sand A, 4.2 4.0 4.2 4.5 4.1 0.3 4.2 mean 4.1 4.0 4.1 4.3 4.2 0.2 Clay A(| 3.6 3.7 3.8 3.8 3.7 0.3 3.7 A, 3.7 3.8 3.9 4.0 3.8 0.3 3.8 mean 3.6 3.7 3.8 3.9 3.7 0.2 11 1 = ridge preparation, 2 =rotary harrowing Tap root length and diameter Irrigation affected differentlyroot length in each experimental year. In 1989, no statistically sig- nificant effect was found at harvest either in fine sand or clay soil (Table 11) but in 1990, when root length was measured at three growth stages in fine sand, a significant positive effect of wa- ter application was shown at the first sampling in mid-July (Fig. 26). In 1991, on the contrary, tap root length decreased by the late applications of water to fine sand. This negative effect was shown at the first and second sampling, but like in other years, the influenceof irrigation on root length disappeared by the harvest time. Tap root length decreased by compaction in all soil types and years. The measurements in 1990-1991 indicated that during the growing period the maximum diameter of roots was on average slightly higher in irrigated than in non-irrigated plots (Fig. 27). Soil mechanical treatments affected tap root length more than diameter. Only a slight influ- ence of mechanical treatments on root maximum diameter was found in this experiment. The most clear effect was observed in mull where soil com- paction increased root diameterby around 0.5 cm at all three growth stages in 1990. In 1989, roots were thickest at harvest in the treatment of one tractor pass (B 4) when irrigated, both in fine sand and clay (Table 12). In addition, in fine sand, maximum tap root diameterat harvest decreased by irrigation in B 5 treatment, whereas the effect was opposite in B, plots. The interaction was sig- nificant both in 1989 and 1991. Root shape A positive effect of irrigation on the ratio be- tween length and maximum diameterwas shown only in fine sand at the first sampling in 1990 (Fig. 28). At the last two samplings in 1991 the role of irrigation was slightly negative. The ef- fect of soil mechanical treatments on root shape was more clear when it was measured as the ra- tio between length and maximum diameter than as the cylindricality index (Tables 13-14, Figs. 28-29). As compared with compaction, fine sand loosening increased the length to diameter ratio in all three years, while the cylindrical index was increased only at the second sampling in 1991. The cylindrical index was affected by soil com- pactness only in clay soil, being the smaller the more compacted the soil was (Table 14). In mull, soil compaction decreased the length to maxi- mum diameter ratio at all sampligs (Fig. 28). 193 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot Fig. 28. Tap root length to maximum diameter ratio as affected by mechanical treatments (B -B ) of mull and fine sand under non-irrigated (AJ and irrigated (A,) conditions at three sampling times in 1990-1991. B t = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B 3 = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD 0(I, for significant differences: Fine sand: 0.4 (at the same A level 0.5) 89 daysafter sowing (DAS), 0.4 (0.5) 123 DAS, 0.2 (0.3 cm) 158 DAS in 1990; 0.5 (0.7) 90 DAS, 0.3 (0.4) 117 DAS, 0.2 cm (0.2 cm) 145 DAS in 1991 Mull: 0.8 (95 DAS), 0.6 (129 DAS) and 0.2 (165 DAS) in 1990;0.8 (90 DAS), 0.5 (121 DAS) and 0.3 (146 DAS) in 1991 194 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Fig. 29. Tap root cylindrical index as affected by mechanical treatments of mull and fine sand under non-irrigated ( AJ and irrigated (A t ) conditions at three sampling times in 1990-1991.B ; = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD (M(, for significant differences: I in. 5.,,,,1: 0.03 (at the same A level 0.03) 89 days after sowing (DAS), 0.03(0.04) 123 DAS, 0.03 (0.04) 158 DAS in 1990; 0.03 (0.04) 117 DAS in 1991 195 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot Table 13.Tap root length to maximum diameter ratio ofcarrots at thirdsampling of loosened and compacted soils in 1989.A 0 = non-irrigated, A = irrigated. Soil Irrigation Soil loosening 1* Tractor wheel passes (n) HSD00J mean (A) 0 1 3I 2 Fine sand Ao A, 4.34.3 4.3 4,3 4.34.3 4.03.8 3.6 4.13.6 3,8 4.03.7 3.7 0.4 0.4 0.3 4.0 4.0 mean Clay A„ A, 4.64.5 4.64.5 4.64.5 4.23.8 3.7 4.33.7 3.5 4.23.8 3.6 0.4 0.4 0.3 4,2 4.1 mean '* 1 = ridge preparation, 2 = rotary harrowing Table 14. Tap root cyhndricality index of carrots at third sampling of loosened and compacted soils in 1989.A 0 = non-irrigated. A, = irrigated. Soil Irrigation Soil loosening 1* Tractor wheel passes (n) HSD() 0J mean (A) I 2 0 1 3 Fine sand A„ A, 0.820.75 0.770.81 0.780.07 0.79 0.730.79 0.760.71 0.780.07 0.75 0.770.77 0.760.76 0.78 n.s. 2>mean Clay A„ A. 0.730.78 0.750.67 0.590.09 0.70 0.760.78 0.710.64 0.590.09 0.69 0.750.78 0.730.65 0.590.06mean 11 1 = ridge preparation, 2 =rotary harrowing 21 n.s. = not significant 3.2.2 Internal quality of tap roots Dry matter content In fine sand, irrigation decreased dry matter con- tent in tap roots from an average 9.8% to 8.5% by water applications at first sampling in 1989 (Fig. 30). In 1990, irrigation decreased the dry matter content at the first sampling at 10% lev- el. In 1991, the decrease was 0.4 percentage units. As compared with soil loosening, fine sand compaction increased root dry matter content, as was shown at the first sampling in 1989, at the third sampling in 1990and at the second sam- pling in 1991. The interaction between treat- ments was statistically significant in 1989 because the decrease in dry matter content by irrigation was as high as 2 percentage units in B 3 treatment, and much less in other treat- ments. Soil mechanical treatments had the most sig- nificant effect on root dry matter content in clay soil where it increased clearly with increasing soil compactness at the last two samplings (Fig. 30). At harvest, the interaction between irriga- tion and mechanical treatments was significant because irrigation increased dry matter content from 9.5% to 10.1% in B, treatment. In other 196 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. treatments, no influence of irrigation on dry matter content was measured. Compaction of mull increased root dry mat- ter content, but only at the first sampling in 1990 and at the third sampling in 1991. The dry mat- ter content was negatively related to soil mois- ture recorded on plant sampling days in 1991 (gypsum block at 15-cm depth). That year, the growing season in June-July was rainy and au- tumn was dry (r = -o.7B*** for fine sand, r = -0.56 *** for mull). Crudefibre content In fine sand, the only statistically significant ef- fect of irrigation or soil mechanical treatments on crude fibre content (firmness) was measured in 1989 at the first sampling, when the average crude fibre content was 1 percentage unit lower in irrigated than in non-irrigated plots (Fig. 31). The interaction was significant because the de- creasing effect of irrigation was the most pow- erful in B, treatment (1.5 percentage units). Irrigation of clay soil had no effect on the accumulation of crude fibre on average (Fig. 31). Clay soil mechanical treatments had, however, the most remarkable influence on this quality factor. The increasing effect of clay soil com- paction on crude fibre content persisted through- out the growing season. A statistically signifi- cant interaction between treatments was found at the second sampling, as the crude fibre con- tent was 0.6 percentage units higer in three times compacted and irrigated soil than in non-irrigat- ed soil. At the same time, the crude fibre content decreased slightly by irrigations in other treat- ments. In mull, the crude fibre content at the third sampling in 1991 increased with increasing number of tractor passes (Fig. 31). In August 1991, there was some evidence of an increasing effect of mull compaction on crude fibre con- tent (P < 0.10). The crude fibre content was positively de- pendent on soil moisture recorded on plant sam- pling days (gypsum block at 15-cm depth, rep- resenting the approximate soil moisture status of the past ten days) in clay soil (r= 0.46**), as the less close correlation was determined in fine sand in 1991 (r =-o.37***). Juice content Irrigation of fine sand increased (P < 0.001) the juice content of carrots at the first sampling in 1989, while root dry matter content decreased (Fig. 32). Compaction of fine sand did not af- fect the juiciness of tap roots. In clay, the juice content decreased with increasing clay soil com- pactness during the whole growing period of 1989. Irrigation had a strong influence on juice content at the second sampling, as in irrigated and three times compacted plots the roots con- tained around 2 percentage units less juice than without irrigation. The interaction was signifi- cant, as in other mechanical treatments the ef- fect of irrigation was positive. In mull, no statistically significant differences were found until at the third sampling in 1991, when root juiciness decreased with increasing number of tractor wheelings (Fig. 32). The juice content was positively dependent on soil mois- ture on plant sampling days (gypsum block at 15-cm depth, representing the approximate soil moisture status of the past ten days) in mull soil (r = 0.33* for 1990, r = 0.42** for 1991), but only slightly, and at P < 0.10 in clay soil (r = 0.22). In fine sand, no significant correlation was observed. Juice dry matter content In fine sand, the content of juice dry matter (mainly sugar) decreased at the first sampling in 1989 by 1 percentage unit as a result of irri- gation (Fig. 33). The same statistically signifi- cant effect was found also in 1990 at the first sampling, although the difference between two irrigation schemes was only 0.5 percentage units. The same effect continued in 1991, as the juice dry matter content at the third sampling was still slightly lower in irrigated than in non-irrigated plots. At this sampling, sandy soil compaction (B 5 ) increased juice dry matter content. This ef- 197 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot 198 AGRICULTURAL SCIENCE IN FINLAND Fig. 30. Tap root dry matter content as affected by mechanical treatments (B -B ) of mull and fine sand and clay under non- irrigated (A n ) and irrigated conditions at three sampling times in 1989-1991.B : = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel,B, = three passes of the tractor wheel. HSD0((, for significant differences: Fine sand: 0.4% (at the same A level 0.6%) 80 daysafter sowing (DAS) in 1989;0.6% (0.8%) 158DAS in 1990; 0.7% (0.9%) 117 DAS in 1991 Clay: 0.4% (0.6%) 108 DAS, 1.1% (1.6%) 140 DAS Mull: 0.5% 95 DAS in 1990; 0.6% 146 DAS in 1991 199 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot 200 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Fig. 31. Tap root crude fibre content as affected by mechanical treatments ) of mull, fine sand and clay under non- irrigated (A (| ) and irrigated (A,) conditions at three sampling times in 1989-1991. B, = soil loosening by ridge preparation, B, = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B, = three passes of the tractor wheel. HSD(MIS lor significant differences: Clay: 0.7% (al Ihe same A level 0.7%) 81 days after sowing (DAS), 0.3% (0.4%.) 108 DAS, 0.7% (1.0%) 140DAS Mull: o.7'* 146DAS in 1991 201 AGRICULTURAL SCIENCE IN FINLAND AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot 202 Vol. 4: 139-237. Fig. 32. Tap root juice content as affected by mechanical treatments (B -B ) of mull and fine sand and clay under non- irrigated (A (| ) and irrigated conditions at three sampling times in 1989-1991.B, = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD(II), for significant differences: Clay: 3.2%(«I ihc same A level 4.5%) 81 days after sowing (DAS), 2.0% (2.9%) 108 DAS. 3.7% (5.2%) 140 DAS Mull: 3.9* 146 DAS in 1991 203 AGRICULTURAL SCIENCE IN FINLAND Pietola, L.: Effect ofsoil compactness on the growth and quality ofcarrot 204 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Fig. 33. Dry matter content of tap root juice (juice density) as affected by mechanical treatments (B-B s )ofmull, fine sand and clay under non-irrigated ( An) and irrigated ( A, ) conditions at three sampling times in 1989-1991.B, = soil loosening by ridge preparation, B, = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSI) (MI, for significant differences: Fine sand: 0.4% (at the same A level 0.5%) 145 days after sowing (DAS) in 1991 Clay: 0.3% (0.4%) 108 DAS Mull: 0.4% 95 DAS in 1990: 0.5 % 121 DAS in 1991 205 AGRICULTURAL SCIENCE IN FINLAND feet of sandy soil compaction was found also in 1989, but only at P < 0.10. At the first sampling in clay soil, the juice dry matter content decreased only by 0.3 per- centage units by water applications (Fig. 33). Further, soil compaction had only a small posi- tive effect on juice dry matter content at the sec- ond sampling. At this time, there was a signifi- cant interaction between treatments because juice dry matter content decreased by water ap- plications in B 3 treatment (0.5 percentage units), but a slight opposite effect was found in com- pacted soil. In mull field, juice dry matter con- tent showed some tendency to increase by soil compaction (Fig. 33), but statistically signifi- cantly only twice in this experiment, i. e. at the first sampling in 1990 and at the second sam- pling in 1991. The juice dry matter content correlated posi- tively with soil moisture recorded on plant sam- pling days (gypsum block at 15-cm depth, repre- senting the approximate soil moisture status of the past ten days) in clay soil (r = 0.42**) and in fine sand in 1990 only slightly (r = 0.28**). Negative correlations were recorded in 1991 (r = -o.B3*** for fine sand, r = -0.46** for mull) when the rainy early summer was followed by a dry late season. Sugar content In fine sand, irrigation had some small effect on sugar concentration as glucose concentration decreased by water applications in B, treatment at harvest in 1990 (Fig. 34). The effect was in- verse both in loosened and compacted soils. This study did not show any significant effect of soil compactness on total root sugar concentration in any soil type. When statistical analyses were carried out for each sugar concentration sepa- rately, some statistically significant, but still a very negligible effect, was found in sandy soil. In 1989, as compared with B, treatment, the su- crose concentration increased by B s treatment. The difference disappeared in total sugar con- centrations because the glucose and fructose concentrations were the highest in the roots of B 2 treatment. Carotene content Irrigation affected carotene, but in a very differ- ent way in each experimental year (Fig. 35). In 1989, the concentration of beta carotene in clay soil increased on average by 1 mg by water ap- plications. In the same year, this positive effect of irrigation was 1.5 mg (P < 0.06) in fine sand. In 1990, on the contrary, the beta carotene con- centration decreasedby 0.75 mg per 100 g fresh carrot by irrigation of sandy soil. The role of ir- rigation was similar in alfa carotene, but the pos- itive effect of water applications in 1989 re- mained around 0.5 mg (P < 0.05 in fine sand, P < 0.10 in clay), and the negative effect of 1990 was only 0.3 mg. In 1991, no statistically sig- nificant effect was found. In fine sand and clay, neitheralfa or beta car- otenes differedstatistically significantly between mechanical treatments (Fig. 35). In mull, com- paction increased the concentration of beta car- otene significantly (1.1 mg per 100 g fresh car- rot) and alfa carotene at P < 0.10 as compared with B 2 treatment. This was demonstrated in 1990 when carotene contents remained very low after a cool September. Internal quality vs. roots size Some factors of internal quality were dependent on tap root size at harvest, but not in every ex- perimental year and field, as suggested by the coefficients of correlation for dry matter, crude fibre, juice, juice dry matter, sugar and carotene contents. For dry matter, crude fibre and total sugars, a negative correlation was found only in clay soil (r = -o.6o***, -0.37*, and -0.39*, re- spectively). In mull soil in 1991, the correlation between root size and crude fibre (r = -0.39) was significant at P < 0.10. For juice dry matter con- tent, the negative correlation with tap root fresh weight was significant in fine sand in 1991 (r = -0.31 *), in addition to clay soil (r = -0.39*). For juice content, a positive correlation was observed both in clay (r = o.63***) and fine sand in 1990 (r = 0.38*), and at P < 0.10 in mull soil in 1991 (r = 0.40). Carotene content was dependent on root size only in mull in 1990 (r = -0.46*). 206 Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Fig. 34. Tap root sugar content as affected by mechanical treatments (B-B5) ofmull, fine sand and clay under non-irrigated (A(l ) and irrigated (A,) conditions at third sampling in 1989-1991. B, = soil loosening by ridge preparation, B 2 = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B, = three passes of the tractor wheel. HSD()(1S for significant differences: fine -and: for sucrose 0.4% (at the same A level 0.6%). for glucose 0.1 % (0.2%) in 1989 207 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot Fig. 35. Tap root carotene contents as affected by mechanical treatments (B--5)B5 ) of mull, fine sand and clay under non- irrigated (A () ) and irrigated (A : ) conditions at third sampling (harvest) in 1989-1991.B ( = soil loosening by ridge prepara- tion, B, = soil loosening by rotary harrowing, B, = untreated, B 4 = one pass of the tractor wheel, B 5 = three passes of the tractor wheel. HSD 005 for significant differences: Mull: for beta carotene 1.1%in 1990 208 AGRICULTURAL SCIENCE IN FINLAND 3.3 Response of whole-root system and shoot growth to differently treated soil profiles in PVC cylinders 3.3.1 Distribution of weights Tap root and shoot weight According to the preliminary study of clay soil PVC profiles in 1989,soil compaction decreased tap root fresh weight (B, = 155 g, B s = 95 g), while shoot weight remained unaffected. The effects of irrigation and mechanical treatments (B 2 , B s ) on total tap root biomass and shoot growth in PVC cylinders in 1990-1991 are sum- marized in Table 15. Irrigation of fine sand in- creased the tap root weights and shoot dry weights only in 1991. The final shoot weigth was not affected by soil mechanical treatments. Soil compaction decreased the fresh and dry weights of tap roots in fine sand cylinders, but in mull only on fresh weight basis in 1991. The distribution of tap root dry weights in each 5-cm soil layer was not affected by irriga- tion (Fig. 36). In fine sand in 1990 at depths of 5-10 cm and 10-15 cm, soil compaction de- creased tap root biomass significantly. After the rainy early summer in 1991, compaction de- creased tap root biomass less than in 1990, but still significantly at depths of 10-15 cm and 15- 20 cm (at 5-10 cm P < 0.10). Inversely, com- paction increased the biomass accumulation in the top soil layer. Statistically significant inter- actions between treatments were determined at 0-5 cm, where irrigation increased dry weight only in compacted soil, and at 15-20 cm, where irrigation increased dry weight in loose soil but decreased it in compacted soil. In mull soil, the tendency of biomass accumulation of tap roots at surface soil by soil compaction was signifi- cant at P < 0.10 (1991). The tap root dry weight was significantly diminished by soil compaction at the depth of 10-15 cm in 1991 and at 15-20 cm in both years. The clear difference of 1.7 gat the depth of 5-10 cm in mull soil in 1991 was significant at P < 0.06. Fibrous root weight According to the preliminary studies ofPVC pro- files in 1989, soil compaction had no significant effect on fibrous root dry weight accumulation in clay soil. In relation to tap root dry weight, the fi- brous root dry weight tended to increase by soil compaction (Appendix 5). Some roots were observed below the soil depth of 60 cm. Table 16shows the total dry weights per plant for 1990-1991 summarized at the depth range of 0-50 cm. In fine sand, no effects oftreatments were observed in 1990 for a small root system, at sampling 10days after harvest. In 1991,when sampling was performed at harvest, the dry weight for a larger root system increased both by soil irrigation and compaction. In mull, com- paction increased the total dry weight of the fi- brous root system in 1990. In fine sand in 1990, the dry mass distribu- tion of fibrous root weights per plant was increased by irrigation at the depth of 25- 30 cm (Fig. 37). As compared with the profile for the loosened soil, a larger amount of fi- brous roots was found in compacted soil at the depth of 10-15 cm. The weight of fibrous roots was 25% higher in compacted soil at the depth of 15-25 cm than in loose fine sand. In com- parison with irrigated cylinders, soil compact- ness did not affect fibrous root weights signifi- cantly. In 1991, a very remarkable positive effect of irrigation was observed at 15-20 cm, with an increase from 21 to 29 pg cm 3 soil, and at 20-25 cm from 19 to 32 pg cm 1 . That year, the larger fibrous root system was affected more by soil compaction than in 1990. Fi- brous root weight increased by soil compac- tion in all 5-cm layers at the depth of 5-25 cm. In subsoil, a small positive effect of soil com- paction was observed at the depth of 45-50 cm. In 1990, mull compaction increased fibrous 209 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Pietola, L.: Effect ofsoil compactness on the growth and quality ofcarrot Table 15.Total fresh weight (TFW) and dry weight (TDW) of tap roots and shoots per plant (n = 6) in PVC cylinders of rotary harrowed (B 2 ) and three times compacted (B,) soils. A 0 = non-irrigated, A, = irrigated. Soil Year Treatment Tap root, g planr l Shoot, g plant’ l Shoot:root TFW TDW TFW TDW TDW:TDW Fine 1990 A„B, 170 17 20 4.3 0.25 sand A„b' 100 11 17 3.8 0.35 A,B, 140 15 22 4,6 0.29 A,B, 110 12 18 3.6 0.30 HSD 005 (B) 40 4 n.s. l » n.s. n.s. mean (A0) 140 14 18 4.0 0,30 mean (A,) 130 14 20 4.1 0.30 HSD005 n.s. n.s. n.s. n.s. n.s. mean (B 2 ) 150 16 21 4.4 0.27 mean (B,) 110 12 18 3.7 0.33 HSD()05 30 3 n.s. n.s. n.s. Fine 1991 A„B, 110 10 20 3.5 0.33 sand An 90 10 22 3.7 0.39 A,B, 120 12 26 4.3 0.36 A,b‘ 100 11 26 4.2 0.40 HSD 00,(B) 20 2 6 n.s. n.s. mean (A (| ) 100 10 21 3.6 0.38 mean (A,) 110 11 26 4.3 0.36 HSD 1105 10 1 n.s. 0,7 n.s. mean (B2 ) 120 II 23 3.9 0.35 mean (B 5 ) 90 10 24 4.0 0.40 HSD ()05 10 I n.s. n.s. n.s. Mull 1990 A„B, 150 15 20 4.0 0.27 A O B~ 140 14 23 4.6 0.33 HSD005 n.s. n.s. n.s. n.s. n.s. Mull 1991 A„B, 110 12 36 5.4 0.48 A o b‘ 90 II 38 5.7 0.55 HSD005 20 n.s. n.s. n.s. n.s. 11 n.s. = not significant root dry weight at 10-20 cm. In 1991, the in- root biomass was not affected by irrigation of crease was significant at 5-15 cm (at 15-20 cm fine sand, but was higher in compacted fine sand at P < 0.10). than in loose soil in 1990-1991 (Table 16). Mull soil compaction did not affect this ratio. In fine . . ■,, sand in 1990, fibrous root dry weight distribu-Fibrous root weight to tap root weight ratio , tion related to tap root biomass, mg g’ l tap root For the whole root system, at the depth of 0- dry weight, increased by irrigation at the depth 50 cm, the fibrous root dry weight related to tap of 25-30 cm, from 1.4 to 2.3 mg g’1 (Fig. 38). 210 Vol. 4: 139-237. Fig. 36. Tap root dry weight per plant in PVC cylinders in fine sand and mull in 1990-1991 as affected by rotary harrowing (Loosened) and three passes of the tractor wheel (Compacted) under non-irrigated (A () ) and irrigated conditions (A,). HSD (I|), for significant differences: Fine sand in 1990: 0.3 g (at the same A level 0.5 g) at 5-10cm, 0.1 g (0.2 g) at 10-15 cm Pine sand in 1991: 0.6 g (1.5 g)at 0-5 cm, 0.4 g (0.9 g) at 10-15 cm, 0.1 g (0.3 g) at 15-20cm Mull in 1990: 0.7 g at 15-20 cm Mullin 1991: 1.4g at 10-15 cm, 0.9 g at 15-20cm 0.9 g 211 AGRICULTURAL SCIENCE IN FINLAND Pietola, L.: Effect ofsoil compactness on the growth and quality ofcarrot Table 1 6. Total dry weight (TDW), length (TL) and surface area (TSA) of fibrous roots per plant and per tap root dry weight in PVC cylinders of rotary harrowed (B,) and three times compacted (B ) soils at the depth of 0-50 cm. A 0 = non-irrigated, A, = irrigated. Soil Year Treatment TDW TL TSA TDW TL TSA (g) (m) (cm 2) (mg) (m) (cm 2 ) Plant (n = 6)~' Tap root dry weight (g) ' Fine 1990 A()B, 0.30 n.d." n.d. 18 n.d. n.d. sand Atfil 0.34 31 A,8 2 0.34 23 A,8 5 0.38 34 HSD00J(B) n.s. 21 14 mean (A () ) 0.32 25 mean (A,) 0.36 28 HSD|)(|5 n.s. n.s. mean(B,) 0.32 21 mean (B5 ) 0.36 33 HSD„ (IS n.s. 10 Fine 1991 A 0B2 0.37 120 660 37 12 66 sand A()B 5 0.59 170 1010 62 18 104 A,8 2 0.50 180 960 43 15 82 A,B, 0.77 200 1190 73 19 113 HSD 1||)5(B) 0.24 70 400 23 6 38 mean(A„) 0.48 150 830 49 15 85 mean (A,) 0.64 190 1070 58 17 97 HSD0(15 0.09 30 140 n.s. n.s. n.s. mean(B,) 0.44 150 810 40 14 72 mean(B 5 ) 0.68 190 1100 67 18 110 HSD005 0.17 n.s. 280 16 4 25 Mull 1990 A„B, 0.94 210 1400 65 15 95 A O B, 1.15 230 1550 82 17 117 HSD 005 0.13 20 120 n.s. n.s. n.s. Mull 1991 A„B, 0.85 240 1340 76 21 126 A„b; 1.10 300 1900 109 30 190 HSD00ä n.s. n.s. n.s. n.s. n.s. n.s. " n.d. = not determined 21 n.s. = not significant The ratio was higher in compacted soil than in loosened fine sand at the depth of 5-15 cm. In 1991, the ratio increased by irrigation at 15-20 from 7.8 to 9.5 mg g' 1 (P < 0.10) and at 20-25 cm significantly, from 7.1 to 10.3 mg g'. A pos- itive effect of soil compaction was demonstrat- ed at the depths of 5-25 cm and 40-50 cm. In mull in 1990, the increase of the ratio by soil compaction was not significant. In 1991, mull soil compaction increased the ratio at 5-10 cm, but at the depths of 10-15 cm and 35-40 cm only at P < 0.10. 212 AGRICULTURAL SCIENCE IN FINLAND Voi 4: 139-237. 3.3.2 Distribution of fibrous root length R °°t length per plant The total root length at 0-50 cm increased by irrigation, and compaction (at P < 0.10) (Ta- ble 16)- In fine sand’ irrigation increased root length per soil volume (Fig. 39) at 10-15 cm from 0.57 to 0.80 cm cm'3 (P < 0.10), at 15— 2 0 cm from Qsg tQ QJ7 cm cm - 3> at 20-25 cm from 0.52 to 0.78 cm cm 3 (P < 0.06) and at 25- 30 cm from 0.24 to 0.40 cm cm 3 (at P < 0.10), Fig. 37. Fibrous root dry weight per plant in PVC cylinders in fine sand and mull in 1990-1991 as affected by rotary harrowing (Loosened) and three passes of the tractor wheel (Compacted) under non-irrigated (A () ) and irrigated conditions The results are from whole sample material. HSD(I(), forsignificant differences: Fine sand in 1990: 3.4 ugcm 1 (at the same A level 4.6 |!gcm 1 ) at 10-15 cm, 0.1 ug cm° (0.2 fig cm 1) at 30-35 cm Fine sand in 1991: 12.8 ugcm ' (18.0 |j.g cm 1 ) at 10-15 cm, 5.5 u,g cm-' (7.8 ugcm 1 )at 15-20 cm, 14.2 \lgcnr' (20.1 p.g cm 1 ) at 20-25 cm, 0.9 tig cm ' (1.2 ug cm-") at 45-50 cm Mull in 1990: 16.0 ug cm'at 10-15 cm, 20.7 ug cm 1 at 15-20 cm Mullin 1991: 18.7 Hg cm'at 5-10 cm, 13.9 Lgcnf 1 at 10-15 cm 213 AGRICULTURAL SCIENCE IN FINLAND after analysis of the whole sample material (Pro- cedure 1). At 0-10 cm, according to subsampling (Procedure 2), irrigation did not affect root length density. The correlation betweenroot lengths at 10-50 cm according to Procedures I and 2 was similar (r = o.9B***). Fibrous root length density was adversely affected by fine sand loosening particularly un- der non-irrigated conditions. A significant posi- tive effect offine sand compaction on root length density was observed at the depth of 10-25 cm (at 20-25 cm P < 0.10). In subsoil, the positive effect of soil compaction remained statistically significant at the depth of 40-45 cm. In mull soil, the tendency of soil compaction to increase root length density particularly at the depth of Fig. 38. Fibrous root dry weight in relation to tap root dry weight in PVC cylinders in fine sand and mull in 1990-1991 as affected by rotary harrowing (Loosened) and three passes of the tractor wheel (Compacted) under non-irrigated (A () ) and irrigated conditions (A ). The results are from whole sample material. HSD (|(| , for significant differences: Fine sand in 1990: 1.6 mgg' 1 (at the same A level 2.2 mgg 1) at 5-10cm, 0.5 mgg' 1(0.7 mgg"')at 10-15 cm Fine sand in 1991: 4.0 mgg 1 (5.6 mg g') at 5-10cm, 2.7 mg g' (3.8 mg g') at 10-15 cm, 1.9 mg g" 1 (2.8 mg g 1)at 15-20 cm, 3.2 mg g 1 (4.5 mg g') at 20- 25 cm, 0.9 mg g ' (1.2 mg g ') at 40-45 cm, 0.5 mg g ' (0.7 mg g ') at 45-50 cm Mullin 1991: 6.6 mg g'at 5-10 cm 214 Pietola, L.: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. 0-20 cm continued (Fig. 39), even though less statistically significant differences were measured. In 1990,an increase by mull soil compaction was demonstrated at 15-20cm (at 10-15 cm P < 0.06). Similarly, in 1991, a positive effect was found at 0-10 cm. Root length to tap wot weight ratio The total root length at 0-50 cm, in relation to tap root biomass, increased by fine sand com- paction. Irrigation or mull soil compaction had no significant effect on this ratio (Table 16). No Fig. 39. Fibrous root length density per plant and soil volume in PVC cylinders in fine sand and mull in 1990-1991 as affected by rotary harrowing (Loosened) and three passes of the tractor wheel (Compacted) under non-irrigated (A0 ) and irrigated conditions (A.). The results are from whole sample material in Procedure 1 (Fine sand 1991, Proc. 1), and from sub-samples in Procedure 2 (Fine sand 1991, Proc. 2) and Procedure 3 (Mull 1990, 1991). HSD (I(I, for significant differences: i-'ine sand in 1991, Proc. 1: 0.14 cm cm ' (at the same A level 0.20 cm cm'1 ) at 10-15 cm, 0.19 cm cm ' (0.27 cm cm ') at 15-20 cm, 0.05 cm cm ' (0.07 cm cm ') at cm. 0.02 cm cm ' (0.02 cm cm') at 45-50 cm l-'int- sand in 1 9') I. I'roc. 2:0.19 cm cm ' (0.27 cm cm 'l.n 10 IS cm, 0.03 cm cm ' (0.04 cm cm 1)at 45 50 cm Mullin 1990: 0.26 cm cm'at 15-20 cm Mull in 1991: 0.37 cm cm -'at 0-5 cm, 0.43 cm cm' at 5-10 cm 215 AGRICULTURAL SCIENCE IN FINLAND Pietola , L: Effect ofsoil compactness on the growth and quality ofcarrot clear response of soil irrigation was shown to the ratio in different soil layers (Fig. 40). The increase from 2.2 to 2.8 m g’ 1 by irrigation at 15-20 cm was significant only based on Proce- dure 2. Fine sand compaction increased root length related to tap root weight at the depths of 10-25 cm and 40-50 cm. In mull, the ratio increased by soil compaction at the depth of 15-20cm in 1990, and at 0-10 cm in 1991 (at 5-10 cm P < 0.10). Fig. 40. Fibrous root length in relation to tap root dry weight in PVC cylinders in fine sand and mull in 1990-1991 as affected by rotary harrowing (Loosened) and three passes of the tractor wheel (Compacted) under non-irrigated (A(1 ) and irrigated conditions (A,). The results are from whole sample material in Procedure 1 (Fine sand 1991, Proc. 1), and from sub-samples in Procedure 2 (Fine sand 1991, Proc. 2) and Procedure 3 (Mull 1990, 1991). HSD for significant differences: Fine sand in 1991. Proc. 1: 0.53 m g ' (at the same A level 0.75 m g') at 10-15 cm. 0.67 m g ' (0.94 mg ') at 15-20 cm, 0.76 m g ' (1.07 m g ') at 20-25 cm, 0.19 m g" 1 (0.26 mg ') at 40-45 cm, 0.10 m g' (0.14 m g ') at 45-50 cm Fine sand in 1991, Proc. 2: 0.73 m g- 1(1.04 mg' at 10-15 cm, 0.96 m g' 1 (1.36 mg') at 20-25 cm, 0.17 m g" 1(0.24 m g 1) at 40-45 cm, 0.10 m g" 1(0.15 m g ') at 45-50 cm Mull in 1990: 0.87 m g ' at 15-20 cm Mullin 1991: I.7Bmg' ato-scm 216 AGRICULTURAL SCIENCE IN FINLAND 3.3.3 Distribution of fibrous root surface area Root surface area per plant The total root surface area increased by irriga- tion and soil compaction at the depth of 0-50 cm (Table 16). In fine sand, root surface area in re- lation to soil volume increased at the soil depth of 15-25 cm from 0.03 to 0.05 cm 2 cm 3 , and at 25-30 cm from 0.01 to 0.02 cm2 cnr3 (P < 0.06) by irrigation (Fig. 41). The positive effect offine sand compaction on root surface area was dem- onstrated to a depth of 25 cm. The same positive impact was found also in subsoil at the depth of 40-45 cm, with a significant interaction, as the surface area in loose non-irrigated soil was low- er than in other soil conditions. In mull in 1990, root surface area increased by soil compaction at 10-20 cm. In 1991, this positive effect was sig- nificant to a depth of 10cm. Fig. 41. Fibrous root surface area (jr. * projected area of root image) per plant and soil volume in PVC cylinders in fine sand and mull in 1990-1991 as affected by rotary har- rowing (Loosened) and three passes of the tractor wheel (Compacted) under non-irrigated (A n ) and irrigated condi- tions (A,). The results are from sub-samples in Procedure 2 (Fine sand) and Procedure 3 (Mull). HSD ))((1 for significant differences: Fine sand in 1991: 0.011 env cm 1 (at the same A level 0.015 cm 2 cm" 3) at 10-15 cm. 0.013 cm2 cm ' (o.ol9 cm'cm') at 15-20 cm. 0.016 cm 2 cm ' (0.023 cm2 cm') at 20-25 cm. 0.004 cm* cm '(0.005 cm' cm ') at 40- 45 cm, 0.002 cm2 cm ' (0.003 cm2 cm ') at 45-50 cm Mullin 1990: 0.0 12cm 2 cm'at 10-15 cm, 0.021 cm2 cm'at 15-20 cm Mull in 1991: 0.020 cm 2 cm'at o-5 cm. 0.026 cm2 cm'at s-1(1 cm 217 Vol. 4: 139-237. AGRICULTURAL SCIENCE IN FINLAND Root surface area to tap root weight ratio Irrigation increased the surface area to tap root dry weight ratio at 15-20 cm. from 12.7 to 15.9 cm 2 g ', at 20-25 cm from 9.7 to 17.4 cm 2 g’ 1 (P < 0.06), and at 25-30 cm from 5.3 to 7.9 cm2 g 1 (P < 0.10). In fine sand, com- paction had a positive effect on this ratio at the depths of0-5 cm (P < 0.06), 5-10 cm (P < 0.10), and significantly at 10-25 cm and 40-50 cm (Fig. 42). In mull in 1990, the ratio increased by soil compaction at the depths of 15-20 cm (P < 0.06) and 45-50 cm (P< 0.10). In 1991, this ratio increased in the two top layers (P < 0.10). Fig. 42. Fibrous root surface area (tt * projected area of root image) in relation to tap root dry weight in PVC cylin- ders in fine sand and mull in 1990-1991 as affected by ro- tary harrowing (Loosened) and three passes of the tractor wheel (Compacted) under non-irrigated (A0) and irrigated conditions (A : ). The results are from subsamples in Proce- dure 2 (Fine sand) and Procedure 3 (Mull). HSD00J for significant differences: Fine sand in 1991: 4.1 cm2 g’ 1(at the same Alevel 5.9 cm2 g ') at 10- 15 cm, 4.8 cm2 g (6.8 cm2 g')at 15-20 cm, 5.3 cm2 g’ 1 (7.6 cm2 g 1) at 20- 25 cm, 1.3 cm2 g (1.7 cm2 g' 1)at 40-45 cm, 0.9 cm2 g 1 (1.3 cm2 g 1) at 45- 50 cm 218 Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Table 17, Percentage of total root length in two dominant diameter classes. B 2 = rotary harrowing, B 5 = three passes of the tractor wheel, A 0 = non-irrigated. A, = irrigated. Soil Year Treatment Depth, cm 0-10 10-20 20-30 30-40 40-50 Diameter =0.15 mm Fine 1991 A(1 B, 94 94 88 83 84 sand A(l bJ 92 86 82 79 78 A,B, 94 94 88 81 78 A,bJ 91 86 78 71 69 HSD 0 05(B) 2 4 6 4 1 mean (A 0 ) 93 90 85 81 81 mean (A,) 93 90 83 76 74 mean (B 2) 94 94 88 82 81 mean (B,) 92 86 80 75 73 Mull 1990 A( ,B, 81 81 81 81 79 A„bJ 79 74 75 76 73 HSD00J n.s." 4 n.s. n.s. 6 Mull 1991 A„B, 88 86 84 81 81 A„b‘ 89 78 75 79 73 HSD 00, n.s. 6 9 n.s. 8 Diameter =0.40 mm Fine 1991 A(1B 2 6 6 II 15 15 sand A„bJ 8 14 17 19 20 A,B, 7 6 11 17 20 A,b‘, 8 14 20 26 28 HSD0,„(B) 2 4 5 2 1 mean (A0) 7 10 14 17 17 mean (A,) 7 10 15 21 24 mean (B 2) 6 6 II 16 17 mean 8 14 18 22 24 Mull 1990 A„B, 17 17 16 16 18 A„b‘ 19 22 21 22 24 HSD00, 2 2 4 4 5 Mull 1991 A„B, 12 13 14 17 17 A(|Bj 10 20 22 19 25 HSD 00J n.s. 5 7 n.s. 7 "n.s. = not significant -3.3.4 Fibrous root width The fibrous root system consisted mostly of very fine roots. In fine sand, about 90% ofroot length in the plough layer consisted of roots of a diam- eter of around 0.15 mm; in subsoil the corre- spending figure was 80% (Table 17). In mull, carrots had more roots over 0.40 mm in diame- ter than in fine sand. Soil compaction decreased the amount of the most fine 0.15-mm roots in both soil types, and increased the length of larg- er roots. Irrigation of compacted fine sand in- 219 Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot creased further the amount of roots over 0.40 mm in diameter, especially in subsoil. By visual observation during the video re- cording ofroots, some roots from compacted and non-irrigated fine sand increased in diameter and were buckled at the depth of 25-40 cm. Below this depth there were some deformities in all treatments. Generally, thicker roots were flat- tened. Some abnormally vigorous branching was associated with greater root diameters. Other- wise, root branching was uniform. Roots from mull soil had not so clear deformities as sandy soil roots, except for some deadroots at the depth of 30-35 cm, mainly in B, treatment, suggest- ing root death under drier soil conditions of the loosened soil media. Based on visual observations on the carrot root morphology during the video recording (roots on the tray) and during the image analy- ses (root images on the computer screen), the image analysis approach could determine accu- rately the length and width of the fibrous root system. The length of fine carrot roots <0.15 mm in diameter, which were not detected by the image processing system, remained below 5%. 220 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. 4 Discussion 4.1 Effect of soil loosening, compaction and irrigation on soil physical growth factors Soil loosening resulted in a reduced supply of stored soil water for non-irrigated soils. In com- parison with compacted soils, volumetric water contents in loosened soils were lower at matric potentials around -10 kPa, recorded in plough layer before carrot growth. This was supported by the lower volume of small and medium-size pores in loosened soils. Additionally, in both clay (P < 0.05) and mull fields (P < 0.10) the gypsum block recordings were the highest in untreated B, plots, despite a high growth in this treatment. This appeared to be a response to an efficient water uptake by plants under compacted soil conditions (Lipiec et al. 1988, 1992) and low unsaturated hydraulic conductivity (Kemper et al. 1971, Voorhees et al. 1979) as a result of the lower volumetric soil water contents in loosened soils (Mehta et al. 1994). Also Domzal and Hodara (1992) demon- strated this positive effect of soil compactness on volumetric soil water contents for sandy and loamy soils at matric potentials between -1 kPa and -50 kPa, and Reicosky et al. (1981) for a clay loam soil at a matric potential below -10 kPa. Further, the results of Boone et al. (1978) on sandy soil (10% clay, 2.3% organic matter) are well in agreement with the findings of the present study. At a matric potential of - 10kPa, a loose bed held less water on volume basis (31%) than a soil compacted once or four times (axle load 5.6 Mg, moisture slightly be- low field capacity at compaction), with water contents of 34% and 33%, respectively. Accord- ing to Mehta et al. (1994), this decrease of the volumetric water content in fine sand decreased its unsaturated hydraulic conductivity. The low volumetric water content in non-ir- rigated soils, especially in loose media, had adverse consequenses for plant growth. In dry soil, a plant loses its ability to expand the root system, as exemplified by Smucker at el. (1991) who reported root branching of maize to de- crease when soil water potentials dropped be- low -100 kPa. The moisture stress-induced root system maintains turgor less effectively which reduces the root elongation rate (Pritchard et al. 1990 a b, ref. Tardieu 1994). Moreover, as soil moisture is about 2/3 of field capacity, signals from the roots (e.g. abscisic acid) have almost immediate effects upon shoot growth, as leaf conductance and the rate of net photosynthesis decrease (Turner et al. 1985, Zhang and Davies 1989). Soil water contents retained by the smaller pores, which were clearly increased by soil com- paction, appeared to be more available to roots growing in that region of the soil profile. The field experiments were established un- der moist soil conditions for an efficient soil compaction. At the time of compaction, howev- er, some water at a matric potential over 10 kPa was found, decreasing the maximum efficiency of compaction (Akram and Kemper 1979). The impact of tractor wheel traffic on increasing soil dry bulk density and penetrometer resistance was confirmed to depths of 25-30 cm and it is con- sistent with data published by Voorhees et al. (1978), Aura (1983) and Domzal et al. (1992) for different mineral soil types. Unlike fine sand or clay, the effect of one pass with a medium- size tractor was remarkable to soil depths of 30- 35 cm in mull field, i.e 0.15 g cm' 3 as compared with the rotary harrowed (B,) treatment. Obvi- ously, this was the result of a low carrying ca- pacity (p d only 0.5 g cm' 3 for subsoil) and high clay content of the organic field (Horn 1988). In this soil layer of 0-30 cm, the physical soil prop- erties are crucial for plant growth, since root systems of annual plants are mainly located at this soil layer (Ouwerkerk and Noordwijk 1991). 221 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot Penetrometer resistances were different and depended on soil type, water content and bulk density. These data agree with Gerard et al. (1982). The critical limitof penetrometer resist- ance for plant growth of approximately 2 MPa (Martino and Shaykewich 1994) was measured in mull soil only in the dry subsoil after three tractor passes (B 5 ). In fine sand of higher inter- nal friction, this limit was exceeded both in non-compacted and compacted subsoil under various soil moisture conditions, but in the plough layer only after three passes by the tractor wheels. In cohesive clay soil, one pass by the tractor was enough to reach this criti- cal limit in the dry plough layer. The increas- ing effect of soil drying on mechanical imped- ance was shown clearly in fine sand, in agree- ment with Lipiec et al. (1990), as the change in dry bulk density of0.25 g cm'3 increased the pen- etrometer resistance by 2.0 MPa at a moisture of 50% of plant-available water, and by 0.7 MPa near field capacity. The relatively similar change in dry bulk density of 0.15 g cm 3 increased the penetrometer resistance in dry mull soil by 0.9 MPa and by 0.6 MPa in moist mull soil, in- dicating that organic matter in mull lowered the cohesive forces of the clay fraction. This is sup- ported by O’Sullivan (1992) who reported mi- nor effects of soil compaction on the tensile strength (Dexter 1988) in soils of higher organ- ic matter content. Soil penetrometer resistance was better explained by soil dry bulk density than by soil water content since the effect of soil moisture was more marked at high bulk densi- ties, as shown also by Ayers and Perumpral (1982). Based on macroporosity, soil aeration limit- ed plant growth only in compacted clay soil, al- though the volume of macropores (equivalent diameters over 30 pm) was most affected by mechanical treatments in all soil types. This re- sult is well exemplified by Eriksson (1976) and Aura (1983) in clay soil, and by Lipiec and Tarkiewicz (1984) in sandy and loamy soils. In the plough layer ofboth fine sand and mull fields, the volume of macropores was over 15% after three passes with the tractor, suggesting that soil aeration was not severely limited, as when sev- eral passes by a 3 Mg tractor decreased the clay soil macroporosity in the plough layer to below 10%. Air-filled porosity < 10% characterized definitely deficient aeration (Glinski and Step- niewski 1985). In mull soil, at the depth of 2.5-7.5 cm, the pore space occupied by pores below 30 pm in diameter increased by as much as 12 percentage units by three tractor passes, suggesting better water retention for a compacted top layer than for a loosened soil. Similarly, in clay field the change was 5 percentage units, in fine sand even less. This agrees with Jorge et al. (1992) who reported that compaction of wet sandy soil (6 Mg) increased the pore space with pores of a diameter below 50 pm by 5 percentage units as compared with uncompacted soils of a micropo- rosity of 18%. The plant-available water capac- ity, however, remained less affected because the main part of the increase in pores of a diameter below 30 pm was due to the increase in small pores. Soil air composition was affected remarka- bly by soil compaction at the depth of 15 cm only in clay soil where, at the beginning of the growing period and under wet soil conditions, soil oxygen content decreased by almost 5 per- centage units by three tractor passes as compared with non-compacted soil. The increase of car- bon dioxide content was 3 percentage units. This is well in agreement with the changes of soil air composition by soil compaction at the depth of 25 cm (Simojoki et al. 1991) in high axle-load studies on Finnish clay soil. In the present study, clay soil dried up soon, and in mid-June the ef- fect on soil air composition had already disap- peared. Shrinking of clay soil caused large cracks, which explains why air composition was not affected by soil compaction even in irrigat- ed plots. Thus, some hollow spaces were suffi- cient to maintain a good aeration and to com- pensate for the decrease of macropores by soil compaction, as was shown by Frede (1985). In early season, the change in soil air composition caused by soil compaction was smaller in fine sand than in clay soil, which is expected based 222 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. on the relatively high macroporosity of compact- ed fine sand (over 15%). In the present study, the lowest oxygen con- tents were measured in mull soil at the depth of 30 cm. At the beginning of the growing period, the oxygen content was as low as 11% after three tractor passes. The soil carbon dioxide content increased respectively from 1% to 6%. Accord- ing to Simojoki et al. (1991), such a low oxygen content was measured only at the depth of50 cm in clay soil. The lower values for soil oxygen combined with greater carbon dioxide contents in compacted soils appeared to be a response to the higher biological activity with increasing soil dry bulk density (Malicki et al. 1991). In early summer, when plant growth was slow, the effect of mull compaction was much smaller in drier surface soil (depth 15 cm) than in subsoil (30 cm). Later, both soil moisture and soil air composition were similar at both depths. Under these quite dry conditions, not even the vigor- ous carrot growth at the end of July changed the soil air composition, as was the case in the moist fine sand. The changes in soil physical growth factors discussed above are consistent with the theory presented in Figure 1 (p. 147). At the depth of 10-20cm, where soils were compacted the most, the average gravimetric D value (Håkansson 1990) for narrow ridges (B,) and untreated soil was 80 in both fine sand and mull, being only 77 for rotary harrowed soil. These low D values resulted in a low water supply. For all soil types, three tractor passes led to a D value of 92-94, with critical penetrometer resistance in dry mineral soils, or poor aeration in wetter mull soils of high clay contents. Possibly because of a high clay content of mull (over 80%), the D value of this soil type was affected quite sim- ilarly to mechanical treatments as compared with mineral soils. After one pass by the tractor, how- ever, the D value was 87 for mull and only 83 for fine sand, which reflects a weaker applica- bility of D value to organic soils (Håkansson 1990). 4.2 Effect of soil physical growth factors on carrot growth and yield quality Only clay soil was compacted to the extent of inhibiting highly carrot growth. Also in loosened clay, the biomass accumulation in tap roots was impaired in non-irrigated plots, indicating a moisture stress when soil compactness was very low. Only shallow harrowing (to the depth of 5 cm) of autumn ploughed clay soil was enough to create an optimal state of compactness for carrot tap root yield (D = 82). This is less than is needed for small grains (Håkansson 1990). In fine sand and mull, the optimum soil compact- ness for carrot tap root yield was not clearly de- termined, as the growth response varied in dif- ferent years in agreement with Taksdal (1984) on sandy soil. Even if the mineral component of mull soil was mainly clay (over 80%), soil com- paction did not affect adversely tap root growth like in clay field. Consequently, the high organ- ic matter content (around 20%) in the mull soil eliminated significantly the effects of clay frac- tion. This is well in agreement with White (1978) and Millette et al. (1981) according to which different organic soil preparations have no or only a slight effect on carrot yield. During the long dry period of 1990 did fine sand compactness affect the carrot growth rhythm. Then, as fine sand dried up, its pene- trometer resistance exceeded the critical growth limit of 2-2.5 MPa (Martino and Shaykewich 1994), delaying the development of tap root growth in the treatment of three passes. In irri- gated plots with lower penetrometer resistanc- es, this decrease of growth by soil compaction was not so severe. This favourable effect of wa- ter applications was still shown at harvest, but less than earlier in the summer. Thus, in terms of final yields, the positive influence of water applications on carrot growth remained low in all experimental years, regardless of the differ- ent timing of dry periods. This agrees with Aura 223 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot (1985) who showed no effect of irrigation on tap root yield during one week of drought at the end of July. As compared with tap root growth, final shoot yield was affected less by soil compactness, in agreement with Olymbios and Schwabe (1977), White and Strandberg (1979) and Agung and Blair (1989) for other cultivars. Early shoot growth was, however, affected, as the growth was the weakest in loose non-irrigated soils with a moisture stress. Unlike in fine sand, the shoot growth was retarded also in compacted and irri- gated clay because of the poor aeration. The low- er emergenceand seedling development in loose mull, as compared with compacted mull, have been demonstrated also in other organic soils (Olymbios and Schwabe 1977, Strzalka 1990). The low carrot yield in narrow ridges in 1990 can be explained by low plant density caused by drought and very low night temperatures (as low as -B°C) during the most sensitive shooting stage. According to Heikinheimo et al. (unpublished data), the lowest soil temperatures by night were measured from ridges in the mull field. Although early growth was promoted by par- tial soil compaction, increased soil compactness affected the external quality of tap roots adverse- ly during the most vigorous radial growth in dry soil of high penetrometer resistance. This was shown clearly in 1990 in the fine sand experi- ment where carrots were short, conical (low cy- lindricality index) and had a low length to di- ameter ratio in compacted treatments. The chang- es in root shape diminished in late season be- cause of the decreasing effect of precipitation on soil mechanical impedance. However, the al- terations in root length sustained throughout the growing season. Tap root length was affected relatively more than the diameter in the present study in agreement with White (1992). The de- creasing effect of high soil wetness on tap root length (Millette 1983, White 1992) was shown only by the late water applications in 1991. Poor macroporosity clearly affeceted tap root cylindricality. In three times compacted clay soil, where the volume of macropores (>3O pm) de- creased below 10% at the depth below 5 cm, car- rot shape was the most conical. The cylindrical- ity index was only 0.59. This is well in agree- ment with Olymbios and Schwabe (1977) for carrots grown in a soil of macroporosity of 8% and a dry bulk density of 1.45 g cm' 3 . In differ- ent soil types and years, root cylindricality of loose soil carrots was slightly below 0.75, which was consistent with Agung and Blair (1989). Root splitting, branching, and size growth were affected by soil compactness the most in clay and least in mull. This was a response to the most altered physical growth factors in clay soil where the increase of soil compactness in- creased root splitting from 5% (on weight basis) in narrow ridges to over 35% in three times com- pacted clay soil under non-irrigated conditions. Splitting decreased by irrigation in clay in all other mechanical treatments except one tractor pass. This can be explained by the assumption that root growth pressure exceeded soil mechan- ical impedance lowered by irrigation, making diameter growth possible, but at harvest this high growth pressure was shown as root splitting. That year (1989), root splitting was very common at har- vest. In three times compacted clay soil, root di- ameter growth was greatly restricted below the lay- er of secondary tillage (upper 5 cm) even in irri- gated soil. This was shown by many small tap roots (< 50 g) grown in three times compacted and irrigat- ed clay (50% of non-branched and non-split roots). Soil mechanical treatments or irrigation did not have as remarkable an effect on root branch- ing as on splitting. Branched carrots were found mostly in non-compacted or non-loosened soil (B 3 ), except for 1991 when carrots were branched the most in compacted soil. The number of branched carrots was very low in 1990 when plant density was higher than in the other years. This agrees with Dowker et al. (1974) who re- ported that deformed carrots are most abundant at a low plant density. For all soil types on aver- age, the largest number of medium-size(50-250 g) roots with no splitting or branching was pro- duced in untreated plots (B,). The amount of small carrots increased by compaction, but this effect could be diminishedby late irrigation with decreasing penetrometer resistance. 224 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. While the tap root size was affected, the re- sponse of carrot internal quality to soil compact- ness was minor. The concentrations of quality components in the tap roots grown in loose soil were lower or similar to those in carrots grown in partially compacted soil, except juice content which decreased by soil compaction. The con- centrations had a moderate negative relation to final root size, indicating that the restricted veg- etative growth in compacted soil had no harm- ful effects on internal quality. The high sugar and carotene contents even in compacted clay soil can be explained by the strong influence of ge- netic factors on carrot chemical quality (Nils- son 1984) and is also well in agreement with Banga (1963) who reported that the balance be- tween tap root vegetative growth and root rip- ening (diameter growth and increase of colour intensity) is controlled mainly by temperature and plant density, and only to a smaller extent by soil moisture and oxygen content in soil air. Clay soil compaction had a remarkable in- creasing influence on carrot dry matter and crude fibre content. In three times compacted clay soil, crude fibre content was even higher in irrigated plots than in non-irrigated plots. This may be explained by lower cell enlargement in compact- ed soil when irrigated because of poorer aera- tion. According to Aubert et al. (1979), both juice and crude fibre contents are good indicators of root firmness. Thus, a high quantity of juice re- fers to a fragile tap root, as roots with a high crude fibre content are firm, like tap roots in compacted clay soil. Additionally, the high crude fibre or low juice content suggests high maturity for carrots grown in compacted clay, as tap root firmness indicates ripeness of carrots, as do also carotene, sugar and dry matter contents (Fritz and Habben 1974). This is in agreement with Banga and Bruyn (1964), who reported that narrower growth lim- its reduce vegetative growth and favour earlier ripening, including carotenogenesis. Carrots grown in compacted clay soil contained at least as much carotenes (alfa and beta carotenes to- gether) as did carrots grown in mineral soils on average, and much more than reported by Evers (1989a) for cv. Nantes Duke 120 days after sow- ing. This appears to be a response to the smaller cell size in tap root tissue accumulated in com- pacted clay, as determined by the increase of crude fibre content (cell walls). As carotene con- centration increases in the individual cell (Ban- ga and Bruyn 1964), the carotene content in tap root tissue reached these high levels in compact- ed soil, even if the roots were small and their shape was very peculiar. Also the sugar analysis showed higher bio- chemical maturity in carrots grown in compact- ed soils. As the total sugars are mainly sucrose in biochemically mature carrots (Phan and Hsu 1973, Fritz and Weichmann 1979), the higher sucrose level in compacted fine sand suggested earlierripening for carrots in compacted soil. The slightly faster ripening was confirmed by juice dry matter content, closely related to carrot sug- ar content (Aura 1985), which increased most along the season in compacted soil. Weather conditions, however, affected the internal quality more than did soil compactness. As dry matter content in tap root tissue and in juice increased with advancing season, soil mois- ture at sampling time was closely related to dry matter content and juiciness. In mull, the dry matter content did not increase much during the cool late summer in 1990, and the low dry mat- ter content at harvest showed that tap roots were not yet ripe (Fritz and Habben 1974). The low- est carotene content was recorded from these carrots, also indicating incomplete carrot growth (Banga et al. 1955, Phan and Hsu 1973). The highest carotene content in compacted mull showed that carotene synthesis was a little more advanced under compacted soil conditions. This appears to be a response to the larger size of the foliage with higher photosynthetic capacity (Banga and Bruyn 1964). In the same year of 1990, carrots grown in fine sand contained al- most three times more carotenes than those grown in mull. In 1991, after a very cool May and July, the season was warm and carotenes in mull soil carrots almost reached the level report- ed for fine sand. Shooting and plant development were more vigorous in fine sand, which explains 225 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot the differences between these soil types report- ed also by Gormley et al. (1971). During three years, the water applications had both a positive, a negative and no effect on caro- tene concentrations in sandy soil. This indicates the great importance of other weather conditions in carotene synthesis, as temperature as well as timing and length of drought period varied a lot between seasons. This agrees with Simon et al. (1982) who reported that of the genotype, soil and climate effects, climate had a major influ- ence on carotene levels. However, neither soil type nor experimental year affected the sugar contents, indicating less importance of soil type and weather conditions in sugar synthesis than was measured for carotenes. This is supported also by Simon et al. (1982) who showed greater genetic and less environmental variation for su- crose and total sugars. However, the sucrose con- tents of the present study were slightly lower than reported by Evers (1989b) for the same cultivar. 4.3 Role of fibrous root system in carrot response to soil compactness The high biochemical maturity (carotene and sugar contents) of tap roots in partially compact- ed soils is in accordance with high length and surface area of fibrous root system observed under compacted conditions. The differences of soil compactness had a clear effect on the distri- bution of the whole root system. Tap root bio- mass in PVC cylinders was accumulated mainly in the upper 10 cm of the compacted soil, but significant growth was observed also at the depth of 10-20 cm in loose soil. In compacted media the biomass tended (P < 0.10) to accumulate more in the soil surface if irrigation was per- formed. This is in agreement with irrigation stud- ies on carrot (Dragland 1978) where the drought of early summer increased tap root length and improved the shape. Fibrous root growth was, on the other hand, stimulated by partial soil compaction to a depth of 30 cm, especially under non-irrigated condi- tions. The total root length per plant was lower in fine sand than in mull, the maximum root length being 200 m for fine sand in agreement with Agung and Blair (1989), and 300 m for mull. This high root length in partially compacted soils was mainly due to better soil water supply, as the root length in loose and irrigated soil was rather similar to that in compacted and non-irri- gated soil. This agrees with Prathapar et al. (1989) who showed decreased root length den- sity in the weak unsaturated hydraulic conduc- tivity. The vigorous branching caused by high mechanical impedance (Pietola 1991) could be another reason for the high fibrous root length in moderately compacted fine sand. Additional- ly, the lower soil water content in loosened soil during plant emergence retarded the early shoot- ing, too. According to Aguirrezabal et al. (1994), this affected the root elongation rate because it is closely related to intercepted light under field conditions, i.e. shoot growth. To a depth of 30 cm the root length density of carrot fibrous roots (0.5-1 cm cm 3 per plant) was in the range of other dicotyledonous spe- cies (Noordwijk and Brouwer 1991) which also have more roots penetrating to depth in both compact (p d = 1.50 g cm 1 ) and loosened soil (p d = 1.33 g cnr 1 at 10-30 cm) than monoco- tydelons (Materechera et al. 1993). Like barley root length (Xu and Juma 1994), the carrot fi- brous root length followed a similar trend as root dry weight. In loose fine sand at the depth of 15-25 cm, however, the root dry weights were relatively slightly lower than the root length. This was caused by the smallerroot diameter in loose, non-irrigated soil than in compacted pro- files. The increasing percentage of thicker roots caused by increasing dry bulk density with no water limitation is in accordance with Huang (1990) who studied Phaseolus vulgaris L. The diameter increase caused by soil compactness was not recorded in mull ofmuch lower dry bulk density. Since the major part of roots (75% in mull, 85% in fine sand) were still included in the 226 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. lowest width class of around 0.15 mm in compact- ed soil (as compared with 80% in loose mull soil and 90% in loose fine sand), the relative differ- ences between fibrous root dry weight, length and root surface area were minimal. The high length of the smallest roots (diam- eter 0.15 mm) is beneficial for carrot plant, as fine roots perform better mechanically in soil than larger roots. A model of Richards and Grea- cen (1986) predicts lower mechanical impedance for finer roots. Further, Barley et al. (1965) showed that wheatroots of a diameter of 0.3 mm grew better in identical soil than pea roots of a diameterof 1.0 mm. Plant species with root sys- tems consisting ofvery fine roots also have larger specific root surface areas (surface area divided by root weight). In the present study, the specif- ic surface area of a carrot fibrous root system averaged 1500-2000cm 2 g 1. Because of the root storing in ethanol for 2-6 months, the weight loss makes this specific surface area slightly overes- timated.The high specific root surface area sug- gests a low requirement of carbon per unit of root surface area and efficient water and nutri- ent intake for the very thin fibrous root system of carrot (Barber and Silverbush 1984, Eissen- stat 1992). A root length to weight ratio of 250- 350 m g 1 of the carrot root systems produced by this study effectively compares with the length to weight ratio of other species (Jones et al. 1991). As the carrot fibrous root system extends to a depth of 60 cm (Schuurman and Shäffner 1974) or even deeper (the present study), it was shown to be the most abundant in the upper 30 cm. Thus, this relatively large root system consisting of mostly very fine roots in the most fertile soil horizon may be a reason for the insensitivity of carrot yield to water and nitrogen supply in ear- lier Finnish studies (Lehtinen 1984,Aura 1985, Vuorinen and Takala 1987, Evers 1988) and the ability of carrot plant to grow on soils low in available phosphorus (Itoh and Barber 1983a). This is not related to root hairs, because carrot root hairs are short (0.04 mm) and have a very limited role in nutrient uptake (Itoh and Barber 1983a, b. Barber and Silberbush 1984). Further, the ability to obtain water and nutri- ents from soil increases with increasing root to soil contact (Veen et al. 1992) which, in turn, increases by reducing soil porosity (Kooistra et al. 1992). As soil porosity, especially macropo- rosity, decreased and root length density and surface area increased with greater soil compac- tion, the root system of the present study had greater absorption potentials under compacted soil conditions than in loose soil. As more fi- brous roots of larger diameters were found in compacted soil with no water deficit than in loose or non-irrigated soil, soil strength was low enough for optimum root length growth but high enough for root diameter growth to attain better root to soil contact. This could explain why soil compaction affected slightly positively the in- ternal quality components of tap root tissue. Moreover, based on the high sugar and carotene contents, fibrous root permeability or conduct- ance appears not to be adveresely affected by the partial soil compaction applied in the present study. Also Agung and Blair (1989) found in a pot experiment that moderate compaction (p b in- crease from 1.25 to 1.40 g cm 3 ) increased sig- nificantly fibrous root length of ripe carrots in loamy sand in a low water regime with increas- ing shoot biomass. More intensive compaction (p b = 1.55 cm 3) did not, however, affect carrot root length, whereas the most intensive compac- tion (p b = 1.70 g cm' 3) decreased root length. Under the field conditions of the present study, root length was not decreased by soil compac- tion in any soil, not even in clay soil. This was due to lower maximum soil compactness (rela- tive dry bulk density) than in the pots in Agung and Blair (1989). Obviously, the existence of biopores and cracks in the natural soil profiles of our PVC cylinder studies favoured the pene- tration of fine roots. This was most distinct in clay soil where the root dry weights remained the same in loosened and compacted soils. In PVC cylinders, the soil volume was 6.5 times higher than in the pots (diameter 15 cm, height 31 cm) reported by Agung and Blair (1989) for three carrots in loamy sand. In these 227 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot two soil volumes, the maximum total root length of a carrot fibrous root system was the same, 180-200 m per plant. In the pot experiment 153 days after sowing, the maximum root length per plant was 180min a high water regime in a dry bulk density of 1.40 g cm'3 , while a lower root length, 80 m per plant, was observed in low wa- ter treatment in a dry bulk density of 1.25 g cm 3 . In the present study, the maximum root length per plant was 200 m in fine sand, also in a high waterregime (irrigated soil) with a dry bulk den- sity of 1.43 g cm' 1 in the plough layer. The low- est length, 120 m, was observed in non-irrigated soil with a dry bulk density of 1.22 g cm'3 (gravi- metric analyses). Agung and Blair (1989) did not measure the root length distribution inside artificial soil pro- files 31 cm high, and reported total root lengths only. For small grains, root length density was negatively affected by soil compaction (Willat 1986), but it could also be increased by soil com- paction to depths of 10 cm (Lipiec et al. 1992). This emphasizes the importance of root length analysis at various depths within the soil pro- file, specifically in natural field soil profiles used for field crops. This study showed that root length distribution of carrots was close to the root distribution pattern of wheat, sugarbeet and on- ion where differences in root length density be- tween traffic intensities were generally small (Ouwerkerk and Noordwijk 1991). Carrot fi- brous roots showed a fairly regular root distri- bution in the subsoil, but in the surface soil (0- 30 cm) where the roots were the most abundant, the root length, dry weight and distribution of surface area varied between seasons and among soil types. The lower root quantity in fine sand in 1990 (170 days after sowing) as compared with 1991 (150 days after sowing) could be due to the different weather conditions but also to delayed sampling and root turnover (death). As the dry weight, length and surface area of fibrous roots were determined in relation to tap root biomass on dry weight basis, the differences in these values between soil compactness were same or slightly increased. This is of great im- portance for the maturation of the tap root tis- sue, as the fibrous roots of a carrot have a much more significant role in carrot growth than the tap root (Benjamin and Wren 1978, 1980). Con- sequently, the importance of carrot effective fi- brous roots in explaining the minor differences ofcarrot internal quality caused by soil compact- ness is obvious. However, absorption ofions and water shouldalso be investigated, as the root ab- sorption potential also depends on root conduct- ance. Since large quantities of carbon are locat- ed in the tap roots of carrots, this supply could serve as a reservoir for the growth and regrowth of fibrous root systems. Whether the quantities of fibrous roots which are produced under com- pacted soil conditions are the result of reallo- cated carbon from the tap root, warrants further investigation. 5 Conclusions Selected soil physical properties and carrot ex- ternal quality were affected by soil compactness (Fig. 43). Among fine sand, clay and mull, opti- mum soil compactness for carrot yield was ob- served most clearly in clay field. The internal quality of carrots appeared to be unaffected, or was slightly improved by soil compaction. The growth stage, weather condi- tions and soil type had a greater impact on inter- nal quality factors than had soil compactness. The carrot root system consisted mostly of very fine roots (diameter around 0.15 mm) and had a total length of 150-200 m to a depth of 50 cm. Partial soil compaction and irrigation in- creased the fibrous root length and surface area in natural soil profiles to a depth of 30 cm. The 228 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. large absorption potential offibrous root system explains the minor differences of carrot internal quality caused by soil compactness. A shallow harrowing of ploughed land is sug- gested as the maximum loosening operation for carrot cultivation in order to avoid the negative consequences associated with too low soil com- pactness, such as lower water supply, delayed ripeness of carrots and sub-optimal absorption potentials by the modified fibrous root system. Additional studies relating optimal soil phys- ical conditions which promote the absorption of ions and water to carbon utilization by carrot fi- brous root systems are needed before the best management practice can be established for spe- cific cultivars of carrots. Fig. 43. Schematic summary of changes in physical soil parame- ters and morphological and phys- iological carrot parameters with changing soil compactness (D) (Håkansson 1990) at a soil depth of 0-30 cm. 229 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality of carrot References Abdalla, A. M., Hettiaratchi, D. R. P. & Reece, A. R. 1969. The mechanics of root growth in granular media. Journal of Agricultural Engineering Research 14: 236- 248. Agrawal, R. P. 1991. Water and nutrient management in sandy soils by compaction. Soil and Tillage Research: 19: 121-139. Aguirrezabal, L. A. N., Deleens, E. & Tardieu, F. 1994. Root elongation rate is accounted for by intercepted pho- tosyntheticphoton flux density and source-sink relations in field and laboratory-grown sunflower. Plant, Cell and Environment 17: 443-450. Agung, I. G. A. M. Sri & Blair, G. J. 1989. Effects of soil bulk density and water regime on carrot yield harvested at different growth stages. Journal of Horticultural Sci- ence 64: 17-25. Aiken, R. M. 1992. Functional relations of root distribu- tions with the flux and uptake of water and nitrate. PhD thesis. East Lansing, Michigan State University. USA. 159 p. Akker, J. J. H. van den, Arb, W. B. M., Koolen, A. J. & Stuiver, H. J. 1994. Comparison of stresses, compac- tions and increase of penetration resistances caused by a low ground pressure tyre and normal tyre. Soil and Till- age Research 29: 125-134. Akram, M. & Kemper, W. D. 1979. Infiltration of soils as affected by the pressure and water contentat the time of compaction. Soil Science Society of America Journal 43: 1080-1086. Allison, L. E. 1969. Organic carbon. Agronomy 9:1367- 1378. Anderson, G. Pidgeon, J. D., Spencer, H. B. & Parks, R. 1980.A new hand-held recording penetrometer for soil studies. Journal of Soil Science 31: 279-296. Ankeny, M. D., Kaspar, T. C. & Horton, R. 1990. Char- acterization of tillage and traffic effects on unconfined infiltration measurements. Soil Science Society of Amer- ica Journal 54: 837-840. Asady, G. H., Smucker, A. J. M. & Adams, M. W. 1985. Seedling test for the quantitative measurement of root tolerance to compacted soil. Crop Science 25: 802-806. Atwell, B. J. 1990a. The effect of soil compaction on wheat during early tillering. II Concentration of cell con- stituents. New Phytology 115: 37-41. -1990b. The effect of soil compaction on wheat during early tillering. 11l Fate of carbon transported to the roots. New Phytology 115: 43-49. Aubert, S., Bonnet, A. & Szot, B. 1979. Mise au point d’indices rhéologigues de texture en relations avec quelques caractéres biochimiques chez la carotte (Dau- cus carota L.). Annales de Technologie et Agriculture 28: 349-422. Aura, E. 1975. Effects of soil moisture on the germina- tion and emergence of sugar beet (Beta vulgarisL). Jour- nal of the Scientific Agricultural Society of Finland 47:1- 69. - 1983. Soil compaction by tractor in spring and its effect on soil porosity. Journal of Agricultural Science in Fin- land 55: 91-107. - 1985. Avomaan vihannesten veden ja typen tarve. Sum- mary: Nitrogen and water reguiments for carrot, beetroot, onion and cabbage. Maatalouden tutkimuskeskus. Tie- dote 7/85. 61 p. Ayers, P. D. & Perumpral, J. V. 1982. Moisture and den- sity effect on cone index. Transactions of the American Society of Agricultural Engineers 25: 1169-1172. Banga, O. 1963. Breeding for quality in some vegeta- bles. Genetica Agraria 16: 27-37. - & Bruyn, J. W. de 1964. Carotenogenesis in carrot roots. Netherlands Journal of Agricultural Science 12: 204-220. Bruyn, J. W. de & Smeets, L. 1955. Selection of car- rots for carotene content. II Sub-normal content at low temperature. Euphytica 4: 183-189. Barber, S. A. & Silberbush, M. 1984. Plant root mor- phology and nutrient uptake. In: Barber, S. A. & Boulen, D. R. (eds.). Roots, nutrient and water influx, and plant growth. Madison, American Society of Agronomy, Spe- cial Publication 49: 65-87. Barley, K. P., Farrell, D. A. & Greacen, E. L. 1965. The influence ofsoil strength on the penetration of a loam by plant roots. Australian Journal of Soil Research 3: 69- 79. Barnes, W. C. 1936. Effects of some environmental fac- tors on growth and color of carrots. Cornell University Agricultural Experimental Station Memoir 186: 1-36. Barnhill, R. E. & Riesenfeld, F. 1974. Computer Aided Geometric Design. New York, Academic Press. 326 p. Bengough, A. G. & Mullins, C. E. 1990. Mechanical impedance to root growth - A review of experimental tech- niques and root growth responses. Journal of Soil Sci- ence 41: 341-358. - & Young, I. M. 1993. Root elongation of seedling peas through layered soil of different penetration resistances. Plant and Soil 149: 129-139. Benjamin, L. R. & Wren, M. J. 1978. Root development and source-sink relations in carrot, Daucus carota L. Jour- nal of Experimental Botany 29: 425-433. - & Wren, M. J. 1980. Root development and source- sink relations in carrot, Daucus carota L. Journal of Ex- perimental Botany 31:1139-1146. Blake, G. R. 1965. Particle density. Agronomy 9: 371- 373. -, Nelson, W. W. & Allmaras, R. R. 1976. Persistence of subsoil compaction in a Mollisol. Soil Science Society of America Journal 40: 943-948. Bland, W. L. 1989. Estimating root length density by the core break method. Soil Science Society of America Jour- nal 53: 1595-1597, Bleasdale, J. K. & Thompson, R. 1963. An objective method of recording and comparing the shapes of carrot roots. Journal of Horticultural Science 38: 232-241. Boone, F. R., Bouma, J. & Smet, A. H. de. 1978. A case study on the effect of soil compaction and potato growth in a loamy sand soil. I Physical measurements and root- ing patterns. Netherlands Journal of Agricultural Science 26: 405-420. Bouyoucos, G. J. 1954. New type electrode for plaster 230 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. of paris moisture blocks. Soil Science 78; 339-342. Bradley, G. & Smittle, D. 1965. Carrot quality as affect- ed by variety, planting and harvest dates. Proceedings of American Society for Horticultural Science 86: 397- 405. Carter, M. R. 1988. Penetration resistance to character- ize the depth and persistence of soil loosening in tillage studies. Canadian Journal of Soil Science 68: 657-668. Clarkson, D. T. & Hanson, J. B. 1980.The mineral nu- trition of higher plants. Annual Reviews of Plant Physiol- ogy 31: 239-298. CoPlot Programme for Scientific Graphs. 1990. Ber- keley, Cohort Software. 228 p. Dexter, A. R. 1988. Strength of soil aggregates and of aggregate beds. Catena Supplement 11: 35-52. - & Hewitt, J. S. 1 978. The deflection of plant roots. Jour- nal of Agricultural Engineering Research 23: 17-22. Domzal, H. & Hodara, J. 1992. Agrophysical effects of different soils compaction. Proceedings of International Conference of Soil Compaction and Management, Tallinn, Estonia, p. 124-128. Hodara, J. & Skrynicki, J. 1992. Deformation depth of soil compacted by multiple passes of wheels. Proceed- ings of International Conference of Soil Compaction and Management, Tallinn, Estonia, p. 31-35. Double-sonde GammametriqueL.P.C. - I.N.R.A. 1985. Montfavet, Centre de Recherches Agronomique d'Avignon. 13 p. Dowker, B. D., Fennell, J. F. M. & Jackson, J. C. 1974. Variation studies in carrots as an aid to breeding. II Ef- fects of sites, years and densities on some quality char- acters. Journal of Horticultural Science 49: 311-321. Dragland, S. 1978. Nitrogen- og vassbehov hos gulrot. Summary: Nitrogen and water requirements for carrots. Forskning og Försök i Landbruket 29: 139-159. Eavis, B. W. 1972. Soil physical conditions affecting seed- ling root growth. I Mechanical impedance, aeration, and moisture availability as influenced by bulk density and moisture levels in a sandy loam soil. Plant and Soil 36: 613-622. Eissenstat, D. M. 1 992. Costs and benefits of construct- ing roots of small diameter. Journalof Plant Nutrition 15: 763-782. - & Caldwell, M. M. 1988a. Competitive ability is linked to rates of water extraction: a field study of two arid land tussock grasses, Oecologia 75: 1-7. - & Caldwell, M. M. 1988b, Seasonal timing of root growth in favorable microsites. Ecology 69: 870-873. Elonen, P. 1971. Particle-size analysis of soil. Acta Agralia Fennica 122: 1-122. Nieminen, L. & Kara, O. 1967. Sprinkler irrigation on clay soils in southern Finland I. Sprinkler irrigation, its techniqueand effect on soil moisture. Journal of the Sci- entific Agricultural Society of Finland 39: 67-77. Erickson, A. E. 1982. Tillage effects on soil aeration. In: Doren, D. M. van (ed.). Predicting Tillage Effects on Soil Physical Properties and Processes. Madison, American Society of Agronomy, Special Publication 44: 91-105. Eriksson, J. 1976. Influence of extremely heavy traffic on clay soil. Grundförbättring 27: 33-51. Evers, A.-M. 1988. Effects of different ferilization prac- tices on the growth, yield and dry matter content of car- rot. Journal of Agricultural Science in Finland 60: 135- 152. -1989a. Effects of different fertilization practices on the carotene of carrot. Journal of Agricultural Science in Fin- land 61: 7-14. -1989b. Effects of different fertilization practices on the glucose, fructose, sucrose, taste and texture of carrot. Journal of Agricultural Science in Finland 61: 113-122. - 1994. The influence of fertilization and environment on some nutritionally important quality criteria in vegetables - a review of research in the Nordic countries. Agricul- tural Science in Finland 3: 177-188. Ewing, R. P. & Kaspar, T. C. 1993. An accurate method for measuring lengths of washed roots. Agronomy Ab- stracts 1993: 315. Farrell, D. A. & Greacen, E. L. 1966. Resistance to pen- etration of fine probes in compressible soil. Australian Journal of Soil Research 4: 1-17. Fiscus, E. L. & Markhart, A. H., 11l 1979. Relationships between root system water transport properties and plant size in Phaseolus. Plant Physiology 64: 770-773. Foth, H. D. 1990. Fundamentals of Soil Science. Bth edition. New York, John Wiley and Sons. 360 p. Frede, H. G. 1985. The importance of pore volume and pore geometry to soil aeration. In: Monnier, G. & Goss, M. J. (eds.). Soil Compaction and Regeneration. Bos- ton, Comission of European Communities, p. 25-29. Fritz, D. & Habben, J. 1974. Determination of ripeness of carrots. Acta Horticulturae 52: 113-238. - & Weichmann, J. 1979. Influence of harvesting date of carrots on quality and quality preservation. Acta Horti- culturae 93: 91-97. Gerard, C. J., Sexton, P. & Shaw, G. 1982. Physical factors influencing soil strength and root growth. Agron- omy Journal 74: 875-881. Gill, W. R. & Bolt, G. H. 1955. Pfeffer’s studies of the root growth pressures exerted by plants. Agronomy Jour- nal 47: 166-168. Glinski, J. & Stepniewski, W. 1985. Soil Aeration and its Role for Plants. Boca Raton, CRC Press. 173 p. Gormley, T. R., Oriordain, F. & Prendville, M. D. 1971. Some aspects of the quality of carrots on different soil types. Journal of Food Technology 6: 393-402. Grath,T. & Håkansson, I. 1992. Effects of soil compac- tion on developmentand nutrient uptake of peas. Swed- ish Journal of Agricultural Research 22: 13-17. Gupta, S. C. & Larsson, W. E. 1985. Modeling soil me- chanical behaviour durig tillage. In: Unger, P. W & Doren, D. M. van (eds.). Predicting Tillage Effects on Soil Phys- ical Properties and Prosesses. Madison, American Soci- ety of Agronomy, Special Publication 44: 151-178. Harris, G. A. & Campbell, G. S. 1991.Automated quan- tification of roots using a simple image analyzer. Agron- omy Journal 81: 935-938. Hartge, K. H. 1992. Compaction or homogenization. Pro- ceedings of International Conference of Soil Compaction and Management, Tallinn, Estonia, p. 63-74. Heinonen, M., Ollilainen, V., Linkola, E., Varo, P. & Koivistoinen, P. 1988.Carotenoids and retinoids in Finn- ish foods: Ready-to-eat foods. Journal of Food Compo- sition and Analysis 1: 221-230. Heinonen, R. 1960. A soil core sampler with provision 231 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot for cutting successive layers. Journal of Scientific Agri- cultural Society of Finland 32: 176-178. Hillel, D. 1971. Soil and Water. Physical Principles and Processes. 1 st edition. New York, Academic Press. 288 p. Hole, C. C. & McKee, J. M.T. 1988. Changes in soluble carbohydrate levels and associated enzymes of field- grown carrots. Journal of Horticultural Science 63: 87- 93. Horn, R. 1988. Compressibility of arable land. Catena Supplement 11: 53-71. Huang, B. 1990. Growth, carbon partitioning, water and ion uptake by the root systems of Phaseolus seedlings subjected to soil mechanical impedance and water defi- cit stresses. M. S. Thesis. East Lansing. Michigan State University. USA. 75 p. Håkansson, I. 1966. Experiments with different degrees of compaction in the topsoil and upper part of the sub- soil. Grundförbättring 19: 281-332. - 1990. A method for characterizing the state of com- pactness of the plough layer. Soil and Tillage Research 16: 105-120. - 1992. The degree of compactness as a link between technical, physical and biological aspects of soil com- paction. Proceedings of International Conference of Soil Compaction and Management, Tallinn, Estonia, p. 75- 78. Inns, F. M. & Kilgour, J. 1978. Agricultural Tyres. Lon- don, Dunlop Ltd. 70 p. Instruction Manual for Use of Bush Recordings Soil Penetrometer. 1979. Midlothian, Irwine Limited. 35 p. Itoh, S. & Barber, S. A. 1983a. Phosphorus uptake by six plant speciesas related to root hairs. Agronomy Jour- nal 75: 457-461. - & Barber, S. A. 1983b. A numerical solution of whole plant nutrient uptake for soil-root systems with root hairs. Plant and Soil 70: 403-413. Jaakkola, A., Auramo, J. & Simojoki, A. 1990. Effect of air composition on plant growth and nutrient uptake. 14th International Congress of Soil Science, Transactions 4: 62-65, Johnson, J. F., Voorhees, W. 8., Nelson, W. W. & Ran- dal, G. W. 1990. Soybean growth and yield as affected by surface and subsoil compaction. Agronomy Journal 82: 973-979. Jokinen, R. 1977. Effects of added magnesium, potas- sium, lime and nitrogen on oats I. Yields. Journal of the Scientific Agricultural Society of Finland 49: 283-295. Jones, C. A., Bland, W. L., Ritchie, J.T. &Williams, J. R. 1991. Simulation of root growth. In: Hanks, J. & Ritchie, J. T. (eds ). Modeling Plant and Soil Systems. Madison, American Society of Agronomy, Agronomy 31; 92-120. Jones, H. A., Tomos, A. D., Leigh, R. A. & Jones, G. W. 1983.Water-relation parameters of epidermal and cor- tical cells in the primary root of Triticum aestivum L. Planta 158: 230-236. Jorge, J. A., Mansell, R. S., Rhoads, F. M, Bloom. S. A. & Hammond, L. C. 1992. Compaction of a fallow sandy loam soil by tractor tires. Soil Science 92: 322- 330. Juusela, T. & Wäre, M. 1956. Suomen peltojen kuiva- tuspa. Draining condition of the cultivated fields in Fin- land. Soil and Hydrotechnics Research 8: 1-89. Kemper, W. D., Stewart, B. A. & Porter, L. K. 1971. Effect of compaction on soil nutrient status. In: Barnes, K. K. et al. (eds.). Compaction ofAgricultural Soils. Amer- ican Society of Agricultural Engineers Monographe p. 178-189. Kesik,T. 1990. Influence of loess soil compaction on the crop and some morphological and physical features of carrot roots. Zeszyty Problemowe Postepow Nauk Rol- niczych 385: 81-96. Kokko, E. G., Volkmar, K. M., Gowen, B. E. & Entz, T. 1993. Determination of total root surface area in soil core samples by image analysis. Soil and Tillage Research 26: 33-43. Koolstra, M. J., Schoonderbeek, D., Boone, F. R., Veen, B. W. & Noordwijk, M. van 1992. Root-soil con- tact of maize, as measured by a thin-section technique. II Effects of soil compaction. Plant and Soil 139: 119- 129. Kurki, M., Lakanen, E. Mäkitie, O. & Sillanpää, M. 1965. Viljavuusanalyysien ilmoitustapa ja tulkinta. Summary: Interpretation of soil testing results. Annales Agricultur- es Fenniae 4: 145-153. Lehtinen, S. 1984. Avomaavihannesten lannoitus- ja kastelukokeet 1978-1983. (Experiments of fertilization and irrigation for field vegetables, in Finnish). Maata- louden tutkimuskeskus. Tiedote 21/84. 62 p. Li, B. W. & Schuhmann, P. J. 1981 . Gaschromatographic analysis of sugars in granola cereals. Journal of Food Science 46; 425-427. Lipiec, J, Håkansson, I. Tarklewicz, S. & Kossowski, J. 1991. Soil physical properties and growth of spring barley as related to the degree of compactness of two soils. Soil and Tillage Research 19: 307-317. -, Kania, W. & Tarkiewicz, S. 1990. Effect of wheeling on physical characteristics of soils and rooting of some cereals. Zeszyty Problemowe Postepow Nauk Rolniczych 385: 97-105, -, Kossowski, J. & Tarkiewicz, S. 1992. Spring barley yield and root growth in response to the degree of com- pactness of soil and weather conditions. Proceedings of International Conference of Soil Compaction and Man- agement, Tallinn, Estonia, p. 86-89. -, Kubota, T. Iwama, H. & Hirose, J. 1988. Measure- ment of plant water use under controlled soil moisture conditions by the negative pressure water circulation technique. Soil Science and Plant Nutrition 34: 417-428. - & Tarkiewicz, S. 1984. The effect of soil compaction on the coefficient of hydraulic conductivity. Polish Jour- nal of Soil Science 17: 9-14. Logsdon, S. D. & Allmaras, R. R. 1991. Maize and soy- bean root clustering as indicated by root mapping. Plant and Soil 131: 169-176. McCully, M. E. & Canny, M. J. 1989. Pathways and proc- esses of water and nutrient movement in roots. In: Lough- man, B. C. et al. Kolek, J. (eds.). Structural and Func- tional Aspects of Transport in Roots, London, Academic Publishers, p. 3-14. McKyes, E., Negi, S., Douglas, E., Taylor, F. & Ragha- van,V. 1979. The effects of machinery traffic and tillage operation on the physical properties of a clay and on yield of silage corn. Journal of Agricultural Engineering Re- search 24: 143-148. 232 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. Mahrer, Y. & Avissar, R. 1985. A numerical study of the effects of soil surface shape upon the soil temperature and moisture regimes. Soil Science 139: 483-490. Malicki, J„ Bieganowski, A. & Dabek-Szreniawska, M. 1991. Mathematical modeling of biological activity in dif- ferently compacted soils. Soil and Tillage Research 19: 357-362. Martino, D. L. & Shaykewich, C. F. 1994.Root penetra- tion profiles of wheat and barley as affected by soil pen- etration resistance in field conditions. Canadian Journal of Soil Science 74: 193-200. Massle, J. & Passioura, J. B. 1987. The effect of soil strength on the growth of young wheat plants. Australian Journal of Plant Physiology 4: 643-656, Materechera, S. A., Dexter, A. R. & Alston, A. M. 1991. Penetration of a very strong soils by seedling roots of different plant species. Plant and Soil 135: 31-41. Alston, A. M., Kirby, J. M. & Dexter, A. R. 1993. Field evaluation of laboratory techniques for predicting the ability of roots to penetrate strong soil and the influence of roots on water sorptivity. Plant and Soil 149: 149-158. Medvedev, V. W. 1992. Compaction of chernozems and plant production. Proceedings of International Conference of Soil Compaction and Management, Tallinn, Estonia, p. 90-94. Mehta, B. K., Shiozawa, S. & Nakano, M. 1994. Hy- draulic properties of sandy soil at low water contents. Soil Science 157: 208-214. Miedzobrodzka, A., Sikora, E., Cieslik, E. & Leszczyn- ska,T. 1993, Nitrate and nitrite levels in carrot roots. Die Nahrung 37: 41-45. Millette, J. A. 1983. Effect of water table depths on the growth of carrots and onions on an organic soil in vitro. Canadian Journal of Plant Science 63: 739-746. -, Vigier, B. & Hogue, E. J. 1981. Seedbed preparation for carrot production on organicsoil. Journal of the Amer- ican Society for Horticultural Science 106: 491-493. Misra, R. K., Dexter, A. R. & Alston, A. M. 1986. Maxi- mum axial and radial growth pressures of plant roots. Plant and Soil 95: 315-326. Moltz, F. J. 1981. Models of water transport in the soil- plant system: a review. Water Resources Research 17: 1245-1260. Monroe, G. E. & Taylor, J. H. 1989. Traffic lanes tor con- trolled traffic cropping systems. Journal of Agricultural Engineering Research 44: 23-31. Newman, E. I. 1966. A method of estimating the total length of root in a sample. Journal of Applied Ecology 3: 139-145. Nilsson,T. 1984. Towards an understanding of the envi- ronmental and genetic influences on vegetable quality. Retrospectsand prospects. Acta Horticulturae 163:199- 212. Noordwijk, M. van & Brouwer, G. 1991. Review of quan- titative root length data in agriculture. In: McMichael, B. L. & Persson, H. (eds.). Plant Roots and their Environ- ment. New York, Elsevier, p. 515-527. - & Schoonderbeck, D. & Kooistra, M. J. 1993. Root- soil contact of field-grown winter wheat. Geoderma 56: 277-286. - & Willigen, P. de 1991. Root functions in agricultural systems. In: McMichael, B. L. & Persson, H. (eds.). Plant Roots and their Environment. New York, Elsevier, p. 381- 395, Nye, P. H. & Tinker, P. B. 1977. Solute Movement in Soil-root System. Ist edition. Oxford, Blackwell Scientif- ic Publications. 342 p. Olymbios, C. M. & Schwabe, W. W. 1977. Effects of aeration and soil compaction on growth of carrot. Jour- nal of Horticultural Science 52: 489-500. O’Sullivan, M. F. 1992. Unaxial compaction effects on soil physical properties in relation to soil type and culti- vation, Soil and Tillage Research 24: 257-269. Ouwerkerk, C. van & Noordwijk, M. van 1991. Effect of traffic intensity on soil structure and root development in a field experiment on a sandy clay loam soil in the Netherlands. Proceedings of the 12th International Con- ference of the International Soil Tillage Research Organ- ization, Ibadan, Nigeria, p. 253-262. Pan, W. L. & Bolton, R. P. 1991. Root quantification by edge discrimination using a desktop scanner. Agronomy Journal 83: 1047-1052. Pfeffer, W. 1893. Druck und Arbeitsleistung durch Wach- sende Pflanzen. Abhandlungen der Königlich Säch- sischen Gesellschaft der Wissenschaften 33: 235-474. Phan, C. T. & Hsu, H. 1973. Physical and chemical changes occuring in the carrot root during growth. Cana- dian Journal of Plant Science 53: 629-634. Pietola, L. 1991. Effect of clay soil strength and struc- ture on root penetration and crop yield. Annales Agricul- turae Fenniae 30: 345-358. Platenius, H. 1934. Chemical changes in carrots during growth. Plant Physiology 9: 671-680. Pohjanheimo, O. & Heinonen, R. 1960. The effect of irrigation on root development, water use, nitrogen up- take and yield characteristics of several barley varietes. Acta Agralia Fenniae 95: 1-20. Prathapar, S. A., Meyer, W. S. & Cook, F. J. 1989. Ef- fect of cultivation on the relationship between root length density and unsaturated hydraulic conductivity in a mod- erately swelling clay soil. Australian Journal of Soil Re- search 27: 645-650. Pritchard, J., Barlow, P. W., Adams, J. S. & Tomos, A. D. 1990a. Biophysics of the inhibition of the growth of maize roots by lowered temperature. Plant Physiology 93: 222-230. - , Wyn Jones, R. D. & Tomos, A. D. 1990b. Measure- ment of yield treshold and cell wall extensibility of intact wheat roots under different ionic, osmotic and tempera- ture treatments. Journal of Experimental Botany 41: 669- 675. Ranta, E., Rita H. & Kouki, J. 1991. Biometria. Tilasto- tiedettä ekologeille. (Biometry. Statistics for Ecologists, in Finnish). 3rd. ed. Helsinki, Yliopistopaino. 569 p, Reicosky, D. C., Voorhees, W. B. & Radke, J. K. 1981. Unsaturated water flow through a simulated wheel track. Soil Science Society of America Journal 45: 3-8. Richards, B. G. & Greacen, E. L. 1986. Mechanical stresses on an expandingcylindrical root analogue-gran- ular media. Australian Journal of Soil Research 24: 393- 404. Riley H. 1988. Cereal yields and soil physical properties in relation to the degree of compactness of some Norwe- gian soils. Proceedings of the 11th International Confer- 233 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot ence of the International Soil Tillage Research Organi- zation, Edinburgh, Scotland, p. 109-114. Rosenfeld, H. J., Martens, M. & Lea, P. 1984. Varia- tions in sensory and physical characteristics in carrots (Daucus carota L.). Acta Horticulturae 163: 63-70. Schumacher, T. E. & Smucker, A. J. M. 1981. Mechan- ical impedance effects on oxygen uptake and porosity of dry-bean roots. Agronomy Journal 73: 51-55. - 1984. Effect of localized anoxia on Phaseolus vulgaris L. root growth. Journal of Experimental Botany 35:1039- 1047. - 1987. lon uptake and respiration of dry bean roots sub- jected to localized anoxia. Plant and Soil 99: 411-422. Schuurman, J. J. & Schäffner, B. E. 1974. De worte- lontwikkeling van enige tuinbougewassenop zandgrond. Instituut voor Bodemvruchtbaarheid, Rapport 11/74.62 p. Simojoki, A.. Jaakkola, A. & Alakukku, L. 1991. Effect of compaction on soil air in a pot experiment and in the field. Soil and Tillage Research 19: 175-186. Simon, P. W., Peterson, C. E. & Lindsay, R. C. 1982. Genotype, soil and climate effects on sensory and ob- jective components of carrot flavor. Journal of the Amer- ican Society for Horticultural Science 107: 644-648. Sippola, J. 1982. A comparison between a dry combus- tion method and rapid wet combustion method for deter- mining soil organic carbon. Annales Agriculturae Fenni- ae 21: 146-148. - & Erviö, R. 1977. Determination of boron in soils and plants by the azomethine-H method. Finnish Chemical Letters 1977: 138-140. Smucker, A. J. M. 1990. Quantification of root dynamics in agroecological systems. In: Goel, V. S. & Normal, J. M. (eds.). Instrumentation for Studying Vegetation Canopies for Remote Sensing in Optical and Thermal Infrared Re- gions. Remote Sensing Reviews 5: 237-248. - 1993, Soil environmental modifications of root dynam- ics and measurement. Annual Reviews of Phytopatholo- gy 31: 191-216. - & Erickson, A. E. 1989.Tillage and compactive mod- ifications of gaseous flow and soil aeration. In: Larson, W. E. et al. (eds.). Mechanics and Related Processes in Structured Agricultural Soils. Boston, Kluwer Academic Publishers, p. 205-221. -, Ferguson, J. C., Debruyn, W. P., Belford, R. L. & Ritchie, J. T. 1987. Image analysis of video recorded plant root systems. In: Taylor, H. M. (ed.). Minirhizotron Observation Tubes: Methods and Applications for Meas- uring Rhizosphere Dynamics. Madison, American Soci- ety of Agronomy, Special Publication 50: 67-80. -, Nunez-Barios, A. & Ritchie, J.T. 1991. Root dynam- ics in drying soil environment. Belowground Ecology, Spring/91: 4-5. - Mcßurney, S. L. & Srivastava, A. K. 1982. Quantita- tive separation of roots from compacted soil profiles by the hydropneumatic elutralion system. Agronomy Jour- nal 74: 500-503. Soane, B. D. 1992. Crop yield responses to zero- and reduced ground pressure- traffic systems in relation to weather conditions in Scotland. Proceedings of Interna- tional Conference of Soil Compaction and Management, Tallinn, Estonia, p. 54-57. •, Blackwell, P. S., Dickson, J. W. & Painter, D. J. 1980. Compaction by agricultural vehicles: A review I. Soil and wheel characteristics. Soil and Tillage Research 1: 207- 237. Soil Testing Laboratory of Finland 1990, Viljavuustut- kimuksen tulkinta peltoviljelyssä. (Soil Analysis and Ap- plication, in Finnish). Helsinki. 70 p. Steel, R. G. D. &Torrie, J. H. 1981. Principles and Pro- cedures of Statistics - A Biometrical Approach. 2nd ed. Singapore, McGraw-Hill International Book Company. 633 p. Stengel, P., Gabilly, Y., Bartuzzi, P. & Gaudu, J. C. & Brifault, A. 1986. La double sonde gamma LPC-INRA. Definition, expérimentation et utilisation en agronomie. Bulletin de Liaison des Laboratoires des Fonts et Chaussées 141: 109-120. Strandberg, J. O. & White, J. M. 1979. Effect of soil compaction on carrot roots. Journal of the American So- ciety for Horticultural Science 104: 344-349. Strzalka, J. 1990. Effects of zone tillage and compac- tion on growth of carrots and onions in organic soils. M. S. Thesis, East Lansing, Michigan State University. USA. 122 p. Taksdal, G. 1984. Effects of tractor wheelings on carrot quality. Acta Horticulturae 163: 255-260. Tardieu, F. 1994. Growth and functioning of roots and of root systems to soil compaction. Towards a system with multiple signalling? Soil and Tillage Research 30:217-243. Taylor, H. M., Upchurch, D. R. & McMichael, B. L. 1990. Applications and limitations of rhizotrons and minirhizo- trons for root studies. Plant and Soil 129: 29-35. Thompson, R. 1969. Some factors affecting carrot root shape and size. Euphytica 18: 277-285. Turner, N. C., Schulze, E.-D. & Collan, T. 1985. The responses of stomata and leaf gas exchange to vapour pressure deficits and soil water content. II In the meso- phytic herbaceous species Helianthus annuus. Oecolo- gia 65: 348-355. Veen, B. W., Noordwijk, M. van, Willigen, P. de, Boone, F. R. & Kooistra, M. J. 1992. Root-soil contact of maize, as measured by a thin-section technique. 11l Effects on shoot growth, nitrate, and water uptake efficiency. Plant and Soil 139: 131-138. Voorhees, W. 8., Carlson, V. A. & Hallauer, E. A. 1980. Root length measurement with a computer-controlled dig- ital scanning microdensitometer. Agronomy Journal 72: 847-851. Farrell, D. A. & Larson, W. E. 1975. Soil strength and aeration effects on root elongation. Soil Science Society of America Proceedings 39: 948-953. -, Senst, C. G. & Nelson, W. W. 1978. Compaction and soil structure modification by wheel traffic in the nothern corn belt. Soil Science Society of America Journal 42: 344-349. -,Young, R. A. & Lyles, L. 1979. Wheel traffic consider- ations in erosion research. Transactions of the American Society of Agricultural Engineers 22: 789-790. Vuorinen, J. & Mäkitie, O. 1955. The method of soil test- ing in use in Finland. Agrogeological Publication 63: 1- 44. Vuorinen, M. & Takala, M. 1987. Porkkanan ja punajuu- rikkaan sadetus, typpilannoitus ja kalkitus poutivalla hiekkamaalla. (Irrigation, N-fertilization and liming for 234 AGRICULTURAL SCIENCE IN FINLAND Voi 4: 139-237. carrot and beetroot, in Finnish). Maatalouden tut- kimuskeskus. Tiedote 10/87. 30 p. White, J. M. 1978. Soil preparation effects on compac- tion, carrot yield and root characteristics in organic soil. Journal of the American Society for Horticultural Science 103: 433-435. - 1992. Carrot yield when grown under three soil water concentrations. HortScience 27: 105-106. - & Strandberg, J. O. 1979. Physical factors affecting carrot root growth: water saturation of soil. Journal of the American Society for Horticultural Science 104:414-416. Whiteley, G. M„ Utomo, W. H. & Dexter, A. R. 1981. A comparison of penetrometer pressures and the pressures exerted by roots. Plant and Soil 61: 351-364. Wiklert, P. 1960. Studier av rotutveckling hos några nyt- toväxter med särskild hänsyn till markstrukturen. Grund- förbättring 3: 113-148. Wilde, S. A. & Voigt, G. K. 1955. Analysis of Soils and Plants for Foresters and Horticulturists. Ist edition. Ann Arbor, J. W. Edwards Publisher. 117 p. Willat, S.T. 1986. Root growth of winter barley in a soil compacted by the passage of tractors. Soil and Tillage Research 7: 41-50. Xu, J. G. & Juma, N. G. 1994. Relations of shoot C, root C and root length with root-released C of two barley cul- tivars and the decomposition of root-released C in soil. Canadian Journal of Soil Science 74: 17-22. Zhang, J. & Davies, W. J. 1989, Abscisic acid produced in dehydrating roots may enable the plant to measure the water status of the soil. Plant, Cell and Environment 12: 73-81. Zoon, F. C. & Tienderen, P. H. van 1990, A rapid quan- titative measurement of root length and root branching by microcomputer image analysis. Plant and Soil 126: 301-308. 235 AGRICULTURAL SCIENCE IN FINLAND Pietola, L: Effect ofsoil compactness on the growth and quality ofcarrot SELOSTUS Maan tiiviyden vaikutus porkkanan kasvuun ja sadon laatuun Liisa Pietola Maatalouden tutkimuskeskus, Kasvintuotannon tutkimuslaitos Helsingin yliopisto, Soveltavan kemian ja mikrobiologian laitos (nykyinen työpaikka) Vuosina 1989-1991 perustettiin porkkanakoekenttiä Etelä-Suomessa hieta-, savi- ja multamaalle. Tutki- muksessa mitattiin kynnöksen kuohkeuttamisen ja tii- vistämisen sekä sadetuksen vaikutuksia maan fysikaa- lisiin ominaisuuksiin ja kasvukuntoon. Tutkimuskas- vina käytettiin porkkanaa (lajikkeena Nantes Duke), jonka varastojuureksen ulkoisen laadun tiedettiin rea- goivan herkästi maan tiiviyteen. Keskeisenä tutkimus- kohteena oli myös sadon sisäinen laatu ja sen suhde juuriston pinta-alaan maan tiiviyden muuttuessa. Kynnöstä kuohkeutettiin kahdella eri tavalla: 1. Perunamaan multauslaitteella muotoiltiin 2 m levei- siin koeruutuihin 45 cm välein 4 harjua. 2. Kelajyr- simellä kuohkeutettiin 2 m leveitä penkkejä 20 cm syvyyteen, savimaalla 15 cm syvyyteen. Lisäksi kyn- nöstä tiivistettiin ajamalla traktorilla pyörä pyörän jälkeen kiinni joko kerran tai kolmesti. Traktorin taka-akselikuormitus oli 3 tonnia. Kuohkeutus- ja tii- vistysajot tehtiin välittömästi ennen kylvöä maan ol- lessa kosteata. Viidentenä oli käsittelemätön koejä- sen. Kaikissa maan mekaanisissa käsittelyissä käytet- tiin traktorissa 2 metrin raideväliä, jolloin vältyttiin ylimääräiseltä maan tallaamiselta. Sadetuskäsittelyssä hieta- ja savimaan kentät sadetettiin maan kuivuessa alle 50 %:iin hyötykapasiteetista. Ruutuihin oli ke- väällä upotettu 60 cm korkeita ja 30 cm läpimittaisia PVC-lieriöitä. Sadonkorjuuaikana ne nostettiin maas- ta porkkanan varasto- ja sivujuurien yksityiskohtai- sia analyysejä varten. Sivujuuriston pituus ja leveys eri maakerroksissa mitattiin kuva-analysointimenetel- mällä. Kenttämittausten mukaan viljelymaan tiivistämi- nen kylvömuokkauksen yhteydessä kevyellä trakto- rilla lisäsi karkean hiedan, runsasmultaisen saven ja multamaan kyntökerroksen tilavuuspainoa ja mekaa- nista vastusta. Kivennäismaalajit tiivistyivät vain ylimmässä 25 cm:n kerroksessa, mutta multamaalla tiivistysten vaikutukset ulottuivat kyntökerroksen ala- puolelle. Traktorilla ajo lisäsi kyntökerroksen tila- vuuspainoa 8 % hiedalla, 10 % multamaalla ja 13 % savella. Sama tilavuuspainon suhteellinen muutos li- säsi mekaanista vastusta hietamaalla 100 % enemmän kuin multamaalla, kun maiden kosteus oli 50 % hyö- tykapasiteetista. Tiivistetyt maat pidättivät vettä enemmän kuin kuohkeutetut (kapeat harjut tai jyrsitty penkki), kun vesipitoisuus laskettiin suhteessa maan tilavuuteen. Tämä johtui lähinnä pienten (< 0,2 pm) huokosten li- sääntymisestä maata tiivistettäessä. Koekenttämaiden kokonaishuokostilavuus ja erityisesti suurten huokos- ten osuus laskivat olennaisesti tiiviysasteen lisäänty- essä. Kolmesti tiivistetyllä savimaalla suurten huo- kosten (> 30 pm) osuus maan tilavuudesta painui alle 10 % rajan. Tiivistysajot vaikuttivat selkeästi maan ilman koostumukseen savimaalla ja varsinkin multa- maalla, jossa alin mitattu happipitoisuus oli 10 %. Myös varastojuureksen nopea kasvu laski maan ilman happipitoisuutta. Hietamaalla maan fysikaalisista ominaisuuksista mekaaninen vastus muodostui kas- vua rajoittavaksi tekijäksi, kun kuivassa tiivistetyssä kyntökerroksessa vastus kohosi yli 2,5 MPa;n rajan penetrometrillä mitattuna. Kaikilla tutkimusmaalajeilla kyntökerroksen tii- viysaste oli noin 93 % kolmen tiivistysajon jälkeen. Maan tiiviysasteella tarkoitetaan maan tilavuuspainon suhdetta maksimaaliseen tilavuuspainoon. Vain savi- maassa tämä tiiviys heikensi ratkaisevasti porkkanan kasvuedellytyksiä. Savimaassa kynnöksen äestys va- kioraiteita käyttäen osoittautui parhaimmaksi muok- kausmenetelmäksi, kun porkkanan eri kehitysvaiheet otetaan huomioon. Liiallinen kuohkeutus johti epä- tasaiseen ja viivästyneeseen taimettumiseen, ja sadon määrä laski sadettamattomissa kuohkeutetuissa har- juissa tai penkissä yhtä paljon kuin tiivistetyissä koeruuduissa. Hietamaalla kuivissa olosuhteissa kuohkeutus viivästytti taimien kehitystä ja tiivistys rajoitti porkkanan kasvua, mutta lopullisen sadon määrää mekaaniset käsittelyt säätelivät vähän. Mul- tamaalla maan tiivistäminen edisti niin ikään taimien kehitystä, mutta vaikutti muuten vähän porkkanan kasvurytmiin ja lopulliseen satoon. Viljelymaan mekaaniset käsittelyt ja sadetus vai- kuttivat huomattavasti selvemmin porkkanan muo- toon kuin sisäiseen laatuun. Maan tiivistäminen hei- 236 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 139-237. kensi porkkanan pituuskasvua, lieriömäisyyttä ja li- säsi haljenneiden porkkanoiden määrää. Porkkanan kehitysvaihe, sääolosuhteet ja maalaji säätelivät pork- kanan kemiallista laatua huomattavasti enemmän kuin maan tiiviysaste. Jopa tiiviillä savimaalla saavutettiin kasvukaudelle ominaiset porkkanan karoteeni- (10 mg 100 grammassa tuoretta porkkanaa) ja sokeripi- toisuudet (5 %). Samanaikaisesti tiiviin savimaan porkkanat sisälsivät vähemmän mehua sekä enemmän kuiva-ainetta ja raakaa kuitua kuin kuohkean savi- maan tai muiden maalajien porkkanat. Alin karotee- nipitoisuus (3,5 mg/100 g) mitattiin multamaan jyr- sitystä penkistä vuonna 1990, jolloin syyskesä oli poikkeuksellisen viileä. Tiivistäminen lisäsi karotee- nipitoisuuden tasolle 5 mg/100 g. Sadetus vaikutti porkkanan karoteenipitoisuuksiin eri tavoin eri vuo- sina, kun taas sokeripitoisuuteen sadetuksella ei ol- lut osuutta. Vakaana pysynyt kemiallinen laatu eri käsittely- jen välillä samassa kasvuvaiheessa ja samana vuon- na voidaan selittää porkkanan ohuiden juurenhaaro- jen avulla, joita yksi porkkana kasvatti noin 200 m tiivistetyssä ja 120 m kuohkeassa sadettamattomas- sa hietamaassa 0-50 cm:n syvyydessä. Tiivistetyssä multamaassa mitattiin vielä suurempia pituuksia. Suurin osa (80-90 %) tästä pituudesta koostui juu- rista, joiden halkaisija oli noin 0,15 mm. Juuristo, joka muodostuu näin ohuista juurista, mahdollistaa porkkanan tehokkaan veden ja ravinteiden oton. Tii- vistys ja sadetus lisäsivät juuriston kuiva-ainepainoa, kokonaispituutta tai pinta-alaa lähinnä vain 0-30 cm:n kerroksessa, jossa porkkanan juuristo pääosin sijaitsi. Tutkimus antaa viitteitä siitä, että porkkanan varastojuureksen karoteeni- ja sokeripitoisuuden li- sääntyminen kasvukauden loppua kohti on varminta kohtalaisesti tiivistetyssä kyntökerroksessa, jossa ve- denpidätyskapasiteetti ja juuriston pinta-ala oli suu- rempi kuin mekaanisesti kuohkeutetussa maassa. Tutkimuksen tulokset mekaanisten käsittelyjen vaikutuksista maan fysikaalisiin ominaisuuksiin ja porkkanan ulkoiseen laatuun saavat tukea aikaisem- mista tutkimuksista. Porkkanan sisäinen laatu, jota perimä ja kehitysvaihe hallitsevat, ei ollut yhtä sel- keästi maan fysikaalisten kasvutekijöiden säädeltävis- sä kuin ulkoinen laatu. Uudet tulokset porkkanan si- säisestä laadusta sekä juuriston pituuden ja pinta-alan jakautumisesta maaprofiilissa korostavat, että pork- kanan sivujuuriston koko on riittävä takaamaan tii- viydestä kärsineen varastojuureksen biokemiallisen kypsymisen. Tätä tukee se, että tiivistetyssä maassa porkkanan nopeasti kehittynyt lehtimassa pystyi hyö- dyntämään kasvukauden säteilyn. Tutkimuksen pe- rusteella porkkanaa ei kannata kylvää hyvin kuoh- kealle kasvualustalle, sillä lievä tiiviys varmistaa porkkanan varastojuureksen nopean kehityksen ja hyvän sisäisen laadun Suomen lyhyenä kasvukautena. 237 AGRICULTURAL SCIENCE IN FINLAND Appendix 1 Arrangement of field plots on fine sand (1989-1991). Dates of mechanical treatments and irrigations. Main plots: Subplots: A. Sprinkler irrigation B. Mechanical treatments A 0 = 0 mm Loosening A =2 x 30mm B, = narrow ridges B 2 = rotary harrowing (dotted line indicates irrigated sector) B 3 = no traffic (untreated) Compaction B 4 = one pass of the tractor wheel B 5 = three passes of the tractor wheel Dates of mechanical treatments (sowing date in parentheses) 26 April (27 April) 1989 18 April (20 April) 1990 24 April (25 April) 1991 Dates of irrigation (amount of water, mm, in parentheses) 3 July 1989(33) 13 June 1990 (35) 12 July 1991 (18) 9 July 1989(34) 27 June 1990 (35) 1 August 1991 (31) AGRICULTURAL SCIENCE IN FINLAND Appendix 2 Arrangement of field plots on clay soil (1989). Dates of mechanical treatments and irrigations Main plots: A. Sprinkler irrigation A () = O mm A, = 2 x 30 mm (dotted line indicates irrigated sector) Subplots: B. Mechanical treatments Loosening B, = narrow ridges B 2 = rotary harrowing Bj = no traffic (untreated) Compaction B 4 = one pass of the tractor wheel Bj = three passes of the tractor wheel Dates of mechanical treatments (sowing date in parentheses) 11 May (12 May) 1989 Dates of irrigation (amount of water, mm, in parentheses) 4 July 1989(31) 9 July 1989(28) AGRICULTURAL SCIENCE IN FINLAND Appendix 3 Arrangement of field plots on mull soil (1990-1991). Dates of mechanical treatments Plots (with no irrigation) B. Mechanical treatments Loosening B l = narrow ridges B 2 - rotary harrowing Bj = no traffic (untreated) Compaction B 4 = one pass of the tractor wheel B 5 = three passes of the tractor wheel Dates of mechanical treatments (sowing date in parentheses) 18 April (21 April) 1990 26 April (30 April) 1991 AGRICULTURAL SCIENCE IN FINLAND Appendix 4 Sampling arrangement for field plots. 0 gypsum block (l9B9-991) * air sampling (1989-1990) penetrometer resistance measurements (1989-1991) final tap root yield area between broken lines (1989-1990) 0 PVC cylinders (black = 1989, shaded = 1990, white = 1991) 1 " first plant sampling (fine sand 1989-1991: 80, 89 and 90 DAS, respectively) (clay 1989: 81 DAS) (mull 1990-1991: 95 and 90 DAS, respectively) 200-cm1 core sampling for soil porosity and dry bulk density (1989-1990) 2" second plant sampling (fine sand 1989-1991: 109, 123 and 117 DAS, respectively) (clay 1989: 108 DAS) (mull 1990-1991: 129 and 121 DAS, respectively) 50-cm' core sampling for soil moisture and dry bulk density (1990-1991) 3 third plant sampling (fine sand 1989-1991: 137, 158 and 145 DAS, respectively) (clay 1989: 140 DAS) (mull 1990-1991): 165 and 146 DAS, respectively) 4 50-cm 1 core sampling for soil moisture and dry bulk density (1990-1991) gamma ray transmission for soil dry bulk density (1990) plant sampling at both ends of the plot (1989-1990) AGRICULTURAL SCIENCE IN FINLAND 4 Appendix 5 Dry weight distribution of carrot fibrous roots in rotary harrowed (B,) and three times compacted (B s ) clay soil in relation to soil volume (above) and tap root dry weight (below). AGRICULTURAL SCIENCE IN FINLAND