The Finnish Forest Research Institute, Research Papers 777 METSÄNTUTKIMUSLAITOKSEN TIEDONANTOJA 777, 2000 Control of growth and frost hardening of silver birch container seedlings: growth retardants, short day treatment and summer planting Jaana Luoranen SUONENJOEN TUTKIMUSASEMA SUONENJOKI RESEARCH STATION The Finnish Forest Research Institute, Research Papers 777 - METSÄNTUTKIMUSLAITOKSEN TIEDONANTOJA 777,2000 Control of growth and frost hardening of silver birch container seedlings: growth retardants, short day treatment and summer planting Jaana Luoranen ACADEMIC DISSERTATION To be presented, with the permission of the Faculty of Agriculture and Forestry of the University of Helsinki, for public criticism in Auditorium XII of the University Main Building, Aleksanterinkatu 5, on 22 Septem ber, 2000, at 12 o'clock noon. SUONENJOEN TUTKIMUSASEMA SUONENJOKI RESEARCH STATION Supervisors: Dr. Risto Rikala Suonenjoki Research Station Finnish Forest Research Institute, Finland Dr. Heikki Smolander Suonenjoki Research Station Finnish Forest Research Institute, Finland Reviewers: Dr. Steve Colombo Ontario Forest Research Institute, Canada Professor Heikki Hänninen Department of Ecology and Systematics University of Helsinki, Finland Opponent: Professor Olavi Junttila Department of Plant Production University of Helsinki, Finland Cover photos: Seven years old plantation of silver birch in Suonenjoki (photographed by Leo Tervo). Inset: Two months old silver birch contaner seedlinq for summer plantinq (photoqraphed by Risto Rikala). Available at: The Finnish Forest Research Institute Library P.O. Box 18 Fin-01301 Vantaa phone +358 9 857 051 fax +358 9 8570 5582 email: kirjasto@metla.fi Layout: Seppo Oja ISBN 951-40-1737-4 Gummerus Kirjapaino Oy Saarijärvi 2000 3 Jaana Luoranen Abstract Luoranen, J. 2000. Control of growth and frost hardening of silver birch con tainer seedlings: growth retardants, short-day treatment and summer planting. Abstrakti: Rauduskoivun paakkutaimien kasvun ja karaistumisen hallinta: kas vunsääteet, lyhytpäiväkäsittely ja kesäistutus. The Finnish Forest Research In stitute, Research Papers 777. 167 p + 5 appendices. ISBN 951-40-1737-4 The aim of this study was to find methods to control or avoid problems concern ing growth and frost hardening of silver birch container seedlings in nurseries. The objective was to find method(s) by which it would be possible: 1) to de crease the risk of frost damage during autumn frosts, 2) to control height growth of seedlings in nurseries and 3) to improve field performance of birch container seedlings. For characterizing the annual development of birch seedlings, the growth rhythm and frost hardening of seedlings was monitored during the first growing season in nursery conditions. After cessation of height growth, diameter growth continued for a few days, and accumulation of dry mass to shoot (stem and branches) and roots continued for about a month. The ability of roots to egress from peat plugs was high until mid-August. After that root growth decreased linearly toward autumn. One purpose of this study was to seek appropriate assessment tests of the frost hardiness of birch seedlings for evaluting cultural practices and monitor ing methods of frost hardening for operational use in nurseries. The tests com pared were electrolyte-leakage, stem-browning and whole-seedling viability as assessed after the freezing exposure. Frost hardening can be observed by using combined measurements of height growth and water content of seedling apices. The first stage of frost hardening starts when height growth ceases. During the first stage water content decreases until stabilization at the time when the second stage of hardening, i.e. rapid increase in frost hardiness, starts. The yellowing and abscission of leaves corre late to some extent with frost hardening, but the previous growing conditions affect this relationship. Frost hardening can also be determined with regular exposure to -10° C. Damage can be evaluated after two weeks incubation by stem-browning test. Effects of growth retardants on height growth were studied using daminozide application sprayed once (on 28 June at concentrations of 1.0-6.0 g 1 _1 ) and CCC (chlormequat chloride) application sprayed twice (on 29 June and 27 July at concentrations of 0.5-3.0 g 1~'). The most suitable applications for retardation of height growth, without negative effects on other morphological variables, were daminozide concentration of 4 g I~' and CCC concentration of 2 g I_l.1 _1 . Nei ther of these compounds, however, affected the frost hardening of seedlings. In the first crop (seeds sown on May), three-week short-day (SD) treatment (16 hours night length) begun at the end of June stopped height growth only temporarily. Therefore, it did not affect the frost hardening of seedlings. In the second crop (seeds sown in mid-June), the three-week SD treatment started at the end of July stopped height growth during the treatment period. This treat ment also hastened frost hardening of seedlings and during the first stage of hardening improved frost hardiness by 2-6° C, depending on the date. Neither early or late SD treatment affected diameter- or root growth. SD treatments had no influence on growth after planting, but SD-treated seedlings grown in small 4 Control of Growth and Frost Hardening of Silver Birch Container Seedlings containers were more susceptible to hare damage and stem lesions than untreated seedlings were. For avoiding problems related to excessive height growth and costs of large birch seedlings, it was investigated the possibilities to expand planting window to summer by planting small leaf-bearing container-grown seedlings during their first growing season. When seedlings were planted in July or before mid-August, the following seasons' growth improved and survival was the same compared to seedlings planted at the traditional time in early September or the following May. The studied methods for controlling growth and frost hardening (growth retard ants, short-day treatment) did not fulfill the objectives or uncertain results were obtained. The most promising results were obtained with summer planting. Leaf bearing seedlings with actively growing roots, which are planted into warm soil during the growing season, have rapid root egress and, thus, increased field performance potential. Keywords: annual cycle, Be tula pendula , CCC, chlormequat chloride, dami nozide, diameter, electrolyte leakage, frost hardiness, growth cessation, height, leaf abscission, leaf yellowing, root egress, root growth potential (RGP), shoot, stem browning, water content, whole-seedling viability Publisher: Finnish Forest Research Institute, Suonenjoki Research Station. Ac cepted by Kari Mielikäinen, Research Director, 28.6.2000. Author's address: Jaana Luoranen, Finnish Forest Research Institute, Suonen joki Research Station, Juntintie 40, FIN-77600 Suonenjoki, Finland, tel +358- 17-513 811, email: jaana.luoranen@metla.fi 5 Jaana Luoranen Abstrakti Luoranen, J. 2000. Control of growth and frost hardening of silver birch con tainer seedlings: growth retardants, short-day treatment and summer planting. Abstrakti: Rauduskoivun paakkutaimien kasvun ja karaistumisen hallinta: kas vunsääteet, lyhytpäiväkäsittely ja kesäistutus. Metsäntutkimuslaitoksen tiedon antoja 777. 167 s + 5 liitettä. Tutkimuksen tarkoituksena oli löytää menetelmiä rauduskoivun paakkutaimien kasvussa ja karaistumisessa taimitarhalla ilmenneiden ongelmien hallitsemiseen tai välttämiseen. Työssä tutkittiin menetelmiä, joiden avulla olisi mahdollista 1) vähentää hallavaurioriskiä syyshalloissa, 2) säädellä taimien pituuskasvua tai mitarhalla ja 3) parantaa rauduskoivun paakkutaimien maastomenestymistä. Taimien kehitysrytmin selvittämiseksi niiden kasvua ja karaistumista seurat tiin ensimmäisen kasvukauden aikana taimitarhalla. Pituuskasvun päättymisen jälkeen läpimittakasvu jatkui muutamia päiviä. Sen sijaan verson (ranka ja ok sat) ja juurten kuivamassa lisääntyi vielä noin kuukauden ajan. Istutettujen tai mien kyky kasvattaa juuria paakusta ympäröivään maahan oli korkea elokuun puoliväliin saakka ja hidastui sitten lineaarisesti syksyn edetessä. Työssä vertailtiin myös koivun taimien pakkaskestävyyden määritykseen so veltuvia testejä kasvatusmenetelmien ja taimitarhoille soveltuvien karaistumi sen seurantamenetelmien arvioimiseksi. Vertailtavina menetelminä olivat pak kasaltistuksen jälkeen käytettävät elektrolyyttivuoto-, ranganruskettumis- ja tai mien kasvatustestit. Karaistumista voidaan seurata pituuskasvu- ja latvan vesipitoisuusmittauksil la. Karaistumisen ensimmäinen vaihe alkaa pituuskasvun päättyessä ja sen ai kana vesipitoisuus alenee kunnes se tasoittuu toisen karaistumisvaiheen, nope an pakkaskestävyyden lisääntymisen, alkaessa. Lehtien kellastuminen ja varise minen korreloi jonkin verran karaistumisen kanssa, mutta aikaisemmat kasvu olosuhteet vaikuttavat näiden muuttujien suhteeseen. Karaistumista voidaan seu rata myös altistamalla taimia säännöllisesti -10°C:een ja arvioimalla vauriot kahden viikon jälkeen ranganruskettumistestillä. Kasvunsääteiden vaikutusta pituuskasvuun tutkittiin käsittelemällä taimet ker ran (28.6.) daminozidilla (pitoisuudet 1,0-6,0 g 1 _1 ) tai kahdesti (29.6. ja 27.7.) CCC:llä (klormekvaattikloridi) (0,5-3,0 g 1 _1 ). Soveltuvimmat käsittelypitoisuu det taimien pituuskasvun hidastamiseen ilman negatiivisia vaikutuksia taimien muuhun rakenteeseen olivat daminozidilla 4 g l _l ja CCC:llä 2 g I_l.1 _1 . Kumpikaan yhdisteistä ei vaikuttanut taimien karaistumiseen. Toukokuun kylvöerällä kolmen viikon lyhytpäivä- (LP) käsittely (16 tunnin yön pituus) kesäkuun lopussa aloitettuna lopetti koivun taimien pituuskasvun vain väliaikaisesti. Käsittely ei vaikuttanut taimien karaistumiseen. Kesäkuun puolivälin kylvöerällä kolmen viikon LP-käsittely heinäkuun lopussa aloitettu na lopetti pituuskasvun jo käsittelyn aikana. Tämä käsittely aikaisti myös taimien karaistumista ja lisäsi taimien pakkaskestävyyttä karaistumisen ensimmäisen vaiheen aikana 2-6 °C:lla riippuen määritysajankohdasta. Kumpikaan LP käsittely ei vaikuttanut taimien läpimitan tai juurten kasvuun. LP-käsittelyt ei vät myöskään vaikuttaneet istutuksen jälkeiseen kasvuun, mutta pienissä paa kuissa kasvatetut, LP-käsitellyt taimet olivat alttiimpia jänistuhoille ja versolai kuille käsittelemättömiin taimiin verrattuna. Koivun paakkutaimien liiallisen pituuskasvun ja korkeiden taimi- ja istutus kustannusten vähentämiseksi tutkittiin myös mahdollisuutta laajentaa istutus- 6 Control of Growth and Frost Hardening of Silver Birch Container Seedlings kautta kesään käyttäen pienempiä, lehdellisiä, ensimmäisen kasvukauden taimia. Kun taimet istutettiin heinäkuussa tai ennen elokuun puolta väliä, kasvu parani seuraavina kasvukausina ja elävyys oli sama kuin perinteisesti syyskuun alussa tai seuraavana keväänä istutetuilla taimilla. Tutkituilla taimien kasvun- ja karaistumisen hallintamenetelmillä (kasvun sääteet, LP-käsittely) ei saavutettu toivottuja tuloksia tai menetelmät osoittau tuivat epävarmoiksi. Lupaavimmat tulokset saavutettiin kesäistutuksessa. Leh delliset, kasvukauden aikana lämpimään maahan istutetut taimet, joiden juur ten kasvu on aktiivista, juurtuvat nopeasti nopeuttaen taimien alkukehitystä istutuksen jälkeen. Avainsanat: Betula pendula, CCC, daminozidi, elektrolyyttivuototesti, juurten kasvupotentiaali (RGP), juurtuminen, kasvun päättyminen, kasvatustesti, klor mekvaattikloridi, lehtien kellastuminen, lehtien variseminen, läpimitta, pakkas kestävyys, pituus, ranganruskettumistesti, verso, vesipitoisuus, vuosirytmi Julkaisija: Metsäntutkimuslaitos, Suonenjoen tutkimusasema. Hyväksynyt tutkimusjohtaja Kari Mielikäinen, 28.6.2000. Kirjoittajan yhteystiedot: Jaana Luoranen, Metsäntutkimuslaitos, Suonenjoen tutkimusasema, Juntintie 40, 77600 Suonenjoki, puh. 017-513 811, email: jaana.luoranen@metla.fi 7 Jaana Luoranen Preface This study was part of the nursery and forest regeneration investigations carried out at Suonenjoki Research Station, Finnish Forest Research In situte. Experiments on summer planting of silver birch container seed lings were made at the suggestion of Mr. Matti Suihkonen, Director of the Pohjois-Savo Forestry Center; these experiments were made in coop eration with Itä-Suomen Taimi Oy. The study was supported by the Finn ish Forest Research Institute, the Graduate School of Forest Ecology, the European Agricultural Guidance and Guarantee Fund, the Finnish Cultural Fundation, Metsämiesten Säätiö, and the Niemi Fundation. These are all gratefully acknowledged. In addition, I am grateful for the excellent working facilities provided by Suonenjoki Research Station. I want to thank several people who have contributed to this work in one way or another. My warmest thanks to Dr. Risto Rikala and Dr. Heikki Smolander, Head of the Research Station, for their cooperation and for supervising of my work. Risto always had time to guide and advise me during the course of the study, and Heikki contributed by organizing the financing for my work. I am also grateful to Dr. Juha Lappi for statistical advise and especially for teaching me nonlinear re gression analysis and for cooperating in deriving the equations for repa rametrization of logistic functions in estimation of growth and frost har diness. I wish express my gratitude to Professor Pentti K. Räsänen from the University of Helsinki for encouraging my research work. I express my gratitude to Dr. Pedro Aphalo, Mr. Hannu Kukkonen, Dr. Maija Pote ri. Professor Pasi Puttonen and Dr. Aija Ryyppö for encouraging me in my work and for their valuable comments on and suggestions about the manuscript. Professor Heikki Hänninen and Dr. Steve Colombo are ac knowledged for their careful work in reviewing the manuscript and Dr. Joann von Weissenberg for revising the English language. I also thank Mr. Seppo Oja for the text layout. I thank my colleagues and friends at Suonenjoki Research Station for their help and support. In particular, I want to thank Ms. Tuija Koleh mainen, Ms. Ritva Pitkänen, Ms. Piijo Pöyhönen, Mr. Pekka Savola, Ms. Eeva Vehviläinen and many other people for technical assistance. I would also like to acknowledge Mr. Esko Ikäheimonen, Mr. Sakari Kal linen, Mr. Keijo Kareinen and Mr. Juhani Mäkelä, the managers of Syr jälä, Tuusjärvi, Onkamo and Lapinlahti nurseries of Itä-Suomen Taimi Oy, for organizing establishment and measurements of summer plant ing experiments in their nurseries. Finally, I extend my warmest thanks to my parents, my sisters and my brother-in-law, for the support and encouragement they have always pro vided. Suonenjoki, July 2000 Jaana Luoranen 8 Control of Growth and Frost Hardening of Silver Birch Container Seedlings Contents 1 INTRODUCTION 11 1.1 Growing and planting of silver birch container seedlings 11 1.2 Theoretical framework 15 1.2.1 The annual growth cycle and frost hardening 15 1.2.1.1 Active period 15 1.2.1.2 Frost hardening 16 1.2.1.3 Response mechanisms to photoperiodic conditions 19 1.2.2 Alternatives for controlling growth and frost hardening 20 1.2.2.1 Growing methods 20 1.2.2.2 Monitoring of seedling quality in the nursery 21 1.2.4 Relation of root growth and root activity to planting windows 24 1.3 Methods for controlling the development of silver birch container seedlings ... 27 1.4 Aims of the study 29 2 SEEDLING MATERIAL, STUDY CONDITIONS AND COMMON METHODS 30 2.1 Introduction 30 2.2 Growing conditions 30 2.3 Morphology and nutrient analysis 30 2.4 Water content 31 2.5 Root growth potential and root-egress test 32 2.6 Statistical analysis 33 2.6.1 Statistical tests 33 2.6.2 Estimation of growth curves 33 3 DEVELOPMENT OF FIRST-YEAR SILVER BIRCH CONTAINER SEEDLINGS 37 3.1 Shoot and root growth 37 3.1.1 Background and study objectives 37 3.1.2 Material and methods 38 3.1.3 Results 42 3.1.4 Discussion 45 3.1.4.1 Timing of growth and dry-mass partitioning 45 3.1.4.2 Ability of birch-seedling roots to egress during summer 47 3.1.5 Conclusion 48 3.2 Frost hardening and assessment of frost hardiness 49 3.2.1 Background and study objectives 49 3.2.2 Material and methods 52 3.2.2.1 Freezing tests 52 3.2.2.2 Estimation of frost hardiness 55 3.2.3 Results 57 3.2.4 Discussion 59 3.2.5 Conclusion 64 4 ARTIFICIAL CONTROL OF GROWTH AND FROST HARDENING OF SILVER BIRCH SEEDLINGS 65 4.1 Growth retardants 65 4.1.1 Background and study objectives 65 4.1.2 Material and methods 66 4.1.3 Results 68 4.1.4 Discussion 70 4.1.5 Conclusion 74 Jaana Luoranen 9 4.2 Short-day treatment 74 4.2.1 Background and study objectives 74 4.2.2 Material and methods 76 4.2.2.1 Seedling material and SD treatments 76 4.2.2.2 Measurement of morphology 77 4.2.2.3 Estimation of frost hardening 77 4.2.2.4 Performance test for summer planting 79 4.2.2.5 Performance test for spring planting 80 4.2.2.6 Statistical analysis 80 4.2.3 Results 81 4.2.3.1 Early SD treatment and summer planting 81 4.2.3.2 SD treatment of late sown seedlings 86 4.2.4 Discussion 91 4.2.4.1 Morphology and growth 91 4.2.4.2 Frost hardening 94 4.2.4.3 Field performance and planting windows 94 4.2.5 Conclusion 97 5 SUMMER PLANTING OF SILVER BIRCH CONTAINER SEEDLINGS 98 5.1 Background and study objectives 98 5.2 Material and methods 99 5.2.1 Field-performance experiments 99 5.2.2 Assessment of frost hardiness 101 5.2.3 Statistical analysis 102 5.3 Results 103 5.3.1 Field performance 103 5.3.2 Frost hardiness 106 5.4 Discussion 108 5.4.1 Field performance of summer-planted seedlings 108 5.4.2 Frost hardiness of summer-planted seedlings in autumn 111 5.5 Conclusion 112 6 CONCLUDING DISCUSSION 113 6.1 Development of first-year silver birch container seedlings 113 6.1.1 Growth rhythm 113 6.1.2 Control of frost hardening 119 6.1.2.1 Development of frost hardiness 119 6.1.2.2 Assessment of frost hardening in a nursery 122 6.2 Evaluation of seedling vigor with the root-growth potential test 126 6.3 Root growth and summer planting 129 6.4 Comparison of control methods 131 REFERENCES 135 Appendix I Estimated parameters in functions of estimation of frost hardiness of silver birch seedlings (Table) 153 Appendix II Nutrients in stems of birch seedlings to which daminozide and CCC were applied (Table) 154 Appendix 111 Nutrients in stems and roots of short-day-treated birch seedlings (Table) 155 Appendix IV Results of summer planting in experiments established in 1995 and 1996 (figures for height and diameter development) 156 Appendix V ANOVA tables 158 10 Control of Growth and Frost Hardening of Silver Birch Container Seedlings List of symbols and abbreviations Symbol Explanation Unit Section RGP Root growth potential g or mg 3.1,4,6 SV Seedling vitality 6 FW Fresh weight mg 2 DW Dry weight mg 2 WC Water content % 2,4 y, Height, diameter, dry mass or injuries in sigmoid function used in estimation of frost hardiness cm, mm, g, % 2, 3, 5 m Estimated function 2,3 £ Error 2,3 X i Age or temperature in the sigmoid function used to estimate growth curves or frost hardiness days or °C 2,3 m Parameter defining the range between upper and lower asymptote of the sigmoid function when estimated the growth responses cm, mm, g 2, 3, 4.2 c Point of inflection of the sigmoid function day, °C 2,3,4.2,5 b Slope of sigmoid function at inflection point relative to parameter m ora day', °C-' 2, 3.2 l Lower asymptote of the sigmoid function when estimated responses in diameter growth mm 2. 3, 4.2 x , Age of transplantation day 2 w Statistical weight coefficient used to homogenize variances 2,4 f Estimate of sigmoid function 4 V Fixed value used to calculate weight w 2 X p Parameter expressing the proportion p from the maximal value of parameter m or a day, °C 2,4 h Point of inflection of the sigmoid function when estimated the shedding of leaves day 2, 3,4 k Slope of sigmoid function at inflection point relative to parameter h day 1 2 r Change point defining the age when growth curve change to abscission curve day 2, 3.1,4.2 RB Relative browning mm/mm 3.2 L. and L, b t Length of brown tissue and total length of cutting in browning test mm 3.2 REL Relative electrolyte leakage 3.2 et and e2 Electrolyte leakage after freezing and killing |iS cm -1 3.2 £T,„ and ETW Temperature causing 50% and 10 % electrolyte leakage °C 3.2,4.2 dBT w Temperature causing 50% and 10 % stem injuries °C 3.2,4.2 D Proportion of damaged seedlings in whole-seedling viability test % 3.2 L Proportion of seedlings dead in whole-seedling viability test % 3.2 DT 50 and DTW Temperature causing damage of 50% and 10 % of seedlings °C 3.2 LT vl *nd LT w Lethal temperature for 50 and 10 % of seedlings °c 3.2 d Parameter used to define the lower asymptote of the sigmoid function 3.2 a Parameter defining the injury range between lower and upper asymptote in sigmoid function 3.2 11 Jaana Luoranen 1 Introduction 1.1 Growing and planting of silver birch container seedlings In Finland, regeneration with planted silver birch seedlings (Betula pendula Roth) started in the late 1960's as a result of large-scale studies of birch breeding, seedling production and planting techniques by Dr Jyrki Raulo (results have been summarized in Raulo 1978, 1981). At present, 15-20 million birch seedlings are planted annual ly (Figure 1.1 a), of which most are silver birch grown in containers (Västilä and Herrala-Ylinen 1999). In the early 19905, the annual production of birch seedlings reached its maximum (28 mill, seed lings delivered for planting) (Västilä and Herrala-Ylinen 1999). Be cause of changes in the annual afforestation area (Figure 1.1b), the demand for seedlings fluctuates greatly, making it difficult to predict the annual need for seedlings. At present, seeds of silver birch are sown in spring directly into containers or broadcast sown into peat-filled trays, and after a few weeks the germinants are transplanted to container cells. Seedlings are grown in greenhouses until they are moved outside in mid- or late June. Seedlings are then raised in outdoor growing areas until leaf abscission, when they are lifted and stored either outside or in frozen storage until the following spring. A schematic diagram of the present practices in the production of silver birch container seed lings in nurseries is presented in Figure 1.2. Growing birch seedlings in containers requires a lot of space. Dur ing their first growing season, seedlings grow 50-100 cm tall, de pending on sowing time and growing conditions. This means that, to give enough growing space for shoot and root growth, seedlings have to be raised in large containers. After the early growth phase, there is also a need to enlarge the distance between containers in order to increase aeration and reduce the risk of disease (Lilja et ai. 1997). So far, it has been difficult to control the growth and frost harden ing of silver birch seedlings in nurseries. If weather conditions dur ing the growing season favor growth, seedlings may grow too tall in relation to root plug volume. During a warm autumn, the frost har dening of seedlings may also be delayed. Compared to conifer seed lings, birch seedlings are difficult to protect against frost. When au tumn frosts are predicted, conifer seedlings are covered with thin 12 Control of Growth and Frost Hardening of Silver Birch Container Seedlings Figure 1.1 a) Annual pro duction of silver birch con tainer and bareroot seed lings (1 000 000 seedlings per year) in Finnish nurs eries in 1966-1998 and b) total annual afforestation area (1000 ha per year) in Finland in 1970-1998. Figure based on numbers presented in Finnish Statis tical Yearbooks of Forestry for the years 1976-1998 (Anonymous 1977-1989, Västilä and Herrala-Ylinen 1999). cloth or are irrigated in order to prevent damage (Landis 1989). In birch, these methods are not usually used because the seedlings are tall and the areas extensive. The use of irrigation to protect seedlings against frost damage is based on the fact that the temperature of the plant will not fall below the freezing point during sprinkling as long as a change of state from water to ice is taking place (Rose and Haa se 1996). The prerequisite for preventing damage to seedlings is the uniform film of water on the surface of seedlings. In birch, the uni form film of water may be difficult to form because of tall seedlings, large leaves and dense crops. Frost means subzero temperatures at ground level. Hard frost, on the other hand, is defined as frost where the temperature is -0.1 °C or lower 2 meters off the ground, which means the ground level tem perature is lower than -3° C (Solantie 1987). In southern Finland (60- 64°N) in regions with few lakes, hard frosts occur about every third year until the end of August. In the lake region, hard frost do not occur in August, but its probability increases to 25 % until mid-Sep tember (Solantie 1987). In late August and early September, the frost hardiness of seedlings may still be poor. Thus there is a need to de velop methods to prevent frost damage of silver birch seedlings dur ing autumn. Jaana Luoranen 13 Figure 1.2 Schematic presentation of annual cycle, current production and planting schedule of silver birch container seedlings and their relation to seasonality of environmental factors. Problems concerning produc tion, planting and field performance of birch container seedlings. The need for methods to hasten frost hardening or at least to deter mine frost hardiness and frost damage was noticed in eastern Fin land nurseries in 1993-1994. The growing season 1993 was cool (Helminen 1993), which delayed the frost hardening of seedlings (Kivivuori 1994b). On a few nights in late September 1993, mini mum air temperatures were between -2 and -B°C, and at ground level the temperatures were as low as -10° C over large areas of cent ral Finland (Helminen 1993). These exceptionally severe early au tumn frosts caused extensive shoot damage to silver birch seedlings in nurseries in eastern Finland (Kivivuori 1994a,b). The resulting damage incurred losses of millions of Finnish marks for nurseries (Kivivuori 1994a,b). Problems and economic losses increased when some of the damaged seedlings were planted the following spring before the damage was discovered. 14 Control of Growth and Frost Hardening of Silver Birch Container Seedlings At present, it is recommended that silver birch seedlings be planted in May and, in addition, in southern Finland in September or in north ern Finland in August (Vastamäki 1993). Most tree seedlings are planted during a few weeks in May and early June (Figure 1.2). Dur ing this short period much labor is needed for lifting, transporting and planting both in nurseries and on planting sites. However, the number of workers employed in silvicultural work decreased, for example, from 5000 persons in 1982 to 2300 persons in 1997 (Ylitalo 1998). In addition, it is predicted that in the near future there will be even fewer laborers to do this work. Thus, there is a need to enlarge the planting window, so that fewer laborers have time to plant all the seedlings. This means there is a need to investigate the possibility to plant seedlings in summer. The field survival of silver birch container seedlings has been low er than that of bareroot seedlings (Parviainen et ai. 1989, Hytönen 1995). Mortality as high as 40% of the planted seedlings may occur on afforestation sites during the first 5 years after planting (Hytönen 1995). At present, most of the planted birch seedlings are container grown (Figure 1.1 a), and thus alternative methods for forestation using container seedlings to increase the survival of planted seedlings need to be investigated. It would be difficult to return to planting bareroot seedlings on a large scale because their production requires about a year longer than container seedlings. The longer time from sowing to planting bareroot seedlings makes it more difficult to forecast the number of seedlings required for planting. In addition, when bare root seedlings are used, the total costs of forestation are higher than for container seedlings, because the seedlings are more expensive and the costs of planting are higher. As described above and in Figure 1.2, there is a need to develop methods for growing seedlings that allow nursery managers to re spond more rapidly and more effectively to changes in demand for seedlings. At the same time, nursery managers need better ways to control the height growth and frost hardening of birch container seed lings. Moreover, the field performance of container seedlings should be improved. There is also a need to enlarge the planting window so that 1) seedlings could be planted over a longer period during the growing season and 2) all silvicultural work required to ensure satis factory regeneration on clear-cut areas could be made with less la bor. Thus, methods should be found to reduce or avoid these prob lems. 15 Jaana Luoranen 1.2 Theoretical framework 1.2.1 The annual growth cycle and frost hardening 1.2.1.1 Active period The growth and development of seedlings of woody plants are regu lated by both internal and environmental factors. Genotype sets the limit for growth, i.e., growth potential; but environmental factors determine the amount of growth potential realized. Environmental factors are highly correlated in nurseries where many growth-limit ing factors are kept at optimum levels. Major factors affecting growth and photosynthesis in nursery conditions are light, temperature, hu midity, carbon dioxide, water and nutrients in the growing medium (Kozlowski and Pallardy 1997). Landis et al. (1999) divided the annual growth cycle of container seedlings in nurseries into three development phases: establishment, rapid growth and hardening (Figure 1.2). Sowing of the seeds starts the establishment phase. It continues until the terminal shoot in the middle of the cotyledons begins to grow quickly, which starts the rapid-growth phase. This phase continues until the final height is achieved and the hardening phase starts. The effects of environmen tal factors and possibilities to affect them vary in each phase. In ad dition, the duration of each phase depends on the species, seed sour ce, environment, and cultural practices (Landis et al. 1999). At northern temperate and boreal latitudes, woody ecotypes and species are adapted to the local seasonal photoperiod and tempera ture conditions (Sakai and Weiser 1973). Due to seasonality, the ac tive growth and dormancy phases are clearly separated (Sarvas 1972, 1974, Fuchigami et al. 1982). During the most intensive phase of growth, most woody species are not able to frost harden and are thus very sensitive to low temperatures (Larcher 1995). Koski (1990) divided the active period into three phases with res pect to response to photoperiod: during the first phase, shoots do not respond to photoperiod; in the second phase, height growth will stop in response to very long nights (>l2 h); and in the third phase, shoots become gradually more responsive to night length. According to this hypothesis, in natural photoperiod conditions, the earlier the seed lings have been sown or the warmer the growing season, the sooner height growth should stop. According to the photoperiod theory, height growth stops when the photoperiod shortens to less than the critical day length (or more correctly, the night is longer than the critical night length) (e.g. Vaar taja 1954, 1957, 1959, Dormling et al. 1968). On the other hand, annual changes in timing of cessation of height growth are explained by the interaction between photoperiod and temperature sum (Koski 16 Control of Growth and Frost Hardening of Silver Birch Container Seedlings and Selkäinaho 1982, Koski 1985, Koski and Sievänen 1985). The warmer the growing season the higher the accumulated temperature sum (>+s°C) and the shorter the night length that can induce cessa tion of growth (Koski and Selkäinaho 1982). In addition to photoperiodicity, soil moisture, nutrients, and day and night temperatures affect cessation of growth. In growth cham ber conditions, fluctuating day/night temperatures in downy birch (.B . pubescens Ehrh.) and willow (Salixpentandra L.) seedlings (Junt tila 1980) and high temperatures (+lB-24°C) in Norway spruce (Pi cea abies (L.) Karst.) (Heide 1974 a) hasten the response on long nights. On the other hand, warm nights may delay the cessation of growth in birch seedlings, which would normally be induced by long nights (Habjorg 1972 a, Downs and Bevington 1981). However, cool nights (<+4°C) may compensate for the influence of photoperiod and induce cessation of growth even under continuous light (Howell and Weiser 1970). However, temperatures may not greatly influence cessation of height growth under conditions of natural photoperiod (Habjorg 1972 a, Heide 1974 a). Both diameter growth of the shoot (Heide 1974 a, 1977, Bjornseth 1985) and root growth stop later than height growth of the shoot. In many conifers and certain hardwood species, under nonfrozen con ditions, root growth may occur all year round (Ritchie and Dunlop 1980 and references therein). Photoperiod does not affect root growth (Heide 1977, Krasowski and Owens 1991, Hawkins, C. et al. 1996), but results concerning photoperiodic effects on cessation of diame ter growth have been contradictory. According to Heide (1974 a, 1977), photoperiod affects the activity of the cambium in Norway spruce; and the night length that will stop this activity is 2-6 hours longer than that which induces cessation of height growth. In the experi ment of Habjorg (1972 a) both photoperiod and temperature affected the diameter growth of downy birch seedlings. Diameter growth may also be regulated by induction of bud dormancy (Ogren 1999 a). 1.2.1.2 Frost hardening Resistance to early autumn frost is closely related to timing of the cessation of height growth (Sakai and Weiser 1973); and in most species, cessation of growth is a prerequisite for frost hardening (Weiser 1970). Many environmental factors can temporarily cause growth to slow or even stop; but after the return of favorable condi tions, growth usually continues (Weiser 1970). In some cases, the cessation of height growth induced by a stress factor might negative ly affect the frost hardening later in the autumn. For example, when nights are long, low temperatures may prevent induction of frost har dening and reduce frost hardiness later in the fall (Fuchigami et al. 17 Jaana Luoranen 1971). According to Fuchigami et al. (1971), this is because the ear ly stages of frost hardening are active processes, which are possible only above a certain threshold temperature. During the early part of dormancy (autumn dormancy or dormancy I, Sarvas 1972, 1974), called rest (Weiser 1970, Sarvas 1974, Powell 1987), plants have no growth competence. Due to the chilling temperatures of autumn and winter, growth competence increases and during the later stage of dormancy in early and mid-winter, which is called quiescence (Wei ser 1970, Hänninen 1990), winter dormancy or dormancy II (Sarvas 1972, 1974), plants have full growth competence. During quiesc ence, however, bud burst does not take place before the buds are exposed for a prolonged period to high enough temperatures (e.g. temperatures >O°C) (Hänninen 1990). Freezing injury to plants is caused by the formation of ice in their extracellular or intracellular spaces (Levitt 1980). During active growth, cells of woody plants are killed at the moment of ice forma tion; but in the frost hardy stage, they are able to tolerate extracellu lar ice formation in their tissues (Levitt 1980, Chen et al. 1995). Int racellular ice formation always kills the cells (Levitt 1980). In many deciduous species, at winter temperatures, woody tissues and buds survive by supercooling (Palta and Weiss 1993), i.e., the tissues are cooled to a temperature lower than the freezing point without freez ing (Levitt 1980). During frost hardening, the tissues and cells in a seedling change to tolerate low winter temperatures (Levitt 1980). Frost hardening is divided into two (or three) stages (Weiser 1970). The first hardening stage starts with cessation of height growth and is triggered by an increase in night length (Fuchigami et al. 1982). During this harden ing stage, several physiological events occur (Weiser 1970). Seed lings accumulate organic compounds such as sugars, proteins, lipids and other cryoprotectants, which results in significant alterations in cellular metabolism (Gusta and Weiser 1972, Yelenosky 1978, Le vitt 1980, Grossnickle 1992, Kacperska 1993, Ögren 1999b). During frost hardening, the water content of the tissues decreases, which is associated with increased concentration of cell solutes and consequent freezing point depression (Pellett and White 1969, Bur ke et al. 1974, Chen et al. 1977, Gusta et al. 1979, Levitt 1980, Simi novitch 1982, Grossnickle 1992, Chen et al. 1995). This stress- (or photoperiod) induced modification in water relations in the harden ing cells during the first stages of frost hardening is a prerequisite for the adjustment of plant growth and metabolism to low tempera ture (Kacperska 1993). The water content in cells is reduced by dis placement of water in the cells during accumulation of sugars, in creased water loss and decreased uptake of water (Levitt 1980). The second stage of frost hardening begins within a few days after the first autumn frost and is characterized by a rapid increase in frost 18 Control of Growth and Frost Hardening of Silver Birch Container Seedlings hardiness (Van Huystee et al. 1967, Howell and Weiser 1970, Gle rum 1973). In woody plants the most effective temperature for in duction of this second stage of frost hardening is about -3° C (Kac percka 1993). In this stage, the cell membranes and enzymes are modified in such a way that the cells become tolerant of the loss of water that occurs during formation of extracellular ice (Larcher 1995). The third stage of frost hardening, which leads to the highest level of frost hardiness, is achieved by uninterrupted exposure to temperatu res of-30°C ... -50° C (Glerum 1973). Only very frost hardy species are able to attain this stage (Glerum 1973), and this type of frost hardiness can be lost very rapidly (Weiser 1970). Frost hardiness of shoots is always greater than that of roots (Cal me et al. 1994, Colombo et al. 1995, Ryyppö 1998). In container grown red oak (Quercus rubra L.), sugar maple (Acer saccharum Marsh.) and yellow birch (Betula alleghaniensis Britton) the roots harden more slowly in autumn and deharden faster in spring than the stems do (Calme et al. 1994). McEvoy and McKay (1997) studied frost hardiness in several deciduous species and concluded that the frost hardiness of roots depends mainly on soil temperature; air tem perature and precipitation have no direct influence. In conifer spe cies the frost hardiness of roots is also increased by decreasing tem perature and slightly affected by photoperiod (Bigras and D'Aoust 1992). According to Colombo et al. (1995), the frost hardiness of first-year black spruce (Picea mariana (Mill.) B.S.P) seedlings was greatest at the terminal bud and least at the root tip; and the differ ences in hardiness along the stem and roots were gradual. Photoperiod and temperature are the major environmental factors that trigger and influence frost hardening of the shoot. Nutrients (Ma cey and Arnott 1986, Edwards 1989, Hawkins, B. et al. 1995, 1996) and soil moisture (Chen et al. 1977, Chen and Li 1978, Macey and Arnott 1986, DeHayes et al. 1989, Welling et al. 1997) have also been observed to modify frost hardiness. The effects of nutrients are well studied in conifer species. Unbalanced nutrient concentrations and fertilization continued too late in the growing season have usu ally been associated with poor frost hardiness (Aronsson 1980, Gle rum 1985, Stimart et al. 1985). On the other hand, fertilization after bud formation may increase frost hardiness (Glerum 1985). In birch seedlings, high K concentrations in foliage have reflected low frost hardiness (Jozefek 1989). However, fertilization seems not to be a very effective method for controlling frost hardening of seedlings. Although drought stress did not induce cessation of height growth in downy birch seedlings in long-day conditions, it increased the frost hardiness of buds after the one week of exposure (Welling et al. 1997). In seedlings of red osier dogwood (Cornus stolinifera Michx.), low temperature, short day or drought stress were capable of increasing frost hardiness, and the final level of hardiness was the sum of the 19 Jaana Luoranen hardiness induced by the above-mentioned factors (Chen and Li 1978). As stress factors, nutrition and soil moisture may also disturb development of frost hardiness in shoots (Glerum 1985). In conclusion, the most promising way to control frost hardening is to stop height growth by artificially increasing night length before the critical length, thereby inducing the first stage of frost harden ing. Since decreasing temperature is the main factor promoting the later stages of frost hardening, this stage is less easily controlled in nursery conditions. 1.2.1.3 Response mechanisms to photoperiodic conditions Responses to photoperiod are mainly sensed through the red/far-red light-absorbing photochromes in the leaves (Thomas 1998). Several phytochromes sense the changes in the wavelength of light, but only phytochromes A and B have been assessed for their involvement in photoperiodism (Thomas 1998). During the light period, the red light (X max 660 nm) absorbing form of a phytochrome (Pr) changes to the far-red light (A, max 730 nm) absorbing form (Pfr). During the dark period, the phytochrome reverse to the Pr-form (Chory 1997, Tho mas and Vince-Prue 1997). When the duration of uninterrupted dark ness is long enough, bud dormancy is induced (Thomas and Vince- Prue 1997). The photoperiod signal is translocated from leaves to other plant parts in the phloem by plant growth regulators (also called plant hor mones or plant growth substances) (Fuchigami et al. 1971, Timmis and Worral 1975, Larcher 1995). Phytochromes and plant growth regulators are associated with both cessation of growth and induc tion of frost hardening (Thomas and Vince-Prue 1997). Previously it was thought that the reason for cessation of height growth and the onset of dormancy was the increase in the concentra tion of ABA in buds induced by short photoperiod (Vince-Prue 1985). Recent studies have showed that the alteration in ABA levels is not directly involved in the cessation of growth in short days (SDs) (Wel ling et al. 1997). Welling et al. (1997) have shown that drought stress and ABA treatment improved the frost hardiness of downy birch slightly, but could not compensate for SD-induced improvement of frost hardiness. Thus ABA could be a component of the endogenous system which controls dormancy. At present, it is well established that gibberellins (GAs) play an important role in cessation of height growth. Application of GA can delay or prevent the induction of dormancy in several tree species even under SD-conditions (Junttila and Nilsen 1993). According to a hypothesis presented by Thomas and Vince-Prue (1997), the con tents and/or composition of GAs in the buds change when tree seed- 20 Control of Growth and Frost Hardening of Silver Birch Container Seedlings lings are exposed to SDs. Consequently, the cessation of height growth and the induction of frost hardening can be regulated either by short ening the photoperiod or affecting the composition of GAs. 1.2.2 Alternatives for controlling growth and frost hardening 1.2.2.1 Growing methods Several methods are used to induce the cessation of height growth and frost hardening in nurseries. When seedlings are grown, it is pos sible to manipulate photoperiod, water supply (drought stress), growth regulators, fertilization and temperature. For several woody species, the use of short-day (SD) treatment, drought stress, a combination of SD and drought stress, timing of fertilization and control of tempe rature has been widely studied in order to control height growth, ces sation of height growth and frost hardening (Jinks 1994, Landis et al. 1999). The effects of treatments differ between species and be tween environmental conditions. In birch species, as a method for nurseries, SD treatment has been studied by Kelly and Mecklenburg (1978) and by Luoranen and Ri kala (1997) in container-grown silver birch and by Calme et al. (1995) in container-grown yellow birch seedlings. According to Luoranen and Rikala (1997), the effectiveness of SD treatment to stop height growth and induce frost hardening of silver birch seedlings is depen dent on the time treatment is started: if started too young, treatment caused only temporary cessation of height growth. In conifer seed lings, SD treatment has also been shown to hasten frost hardening by at least two weeks compared to seedlings grown under natural photoperiod (Aronsson 1975, Timmis and Worrall 1975, Christers son 1978, D'Aoust and Hubac 1986, Bigras and D'Aoust 1992, Folk et al. 1994, Hawkins, C. et al. 1996). SD treatment seems to be a promising method for controlling the development of birch seedlings in nurseries. Plant growth retardants are organic compounds that retard cell di vision and cell elongation in shoot tissues and thus regulate plant height physiologically (Arteca 1996). The activity of most growth retardants is based on their ability to inhibit some step in GA bio synthesis. Which step is blocked depends on the chemical formula of the retardant (Sponsel 1995). Growth retardants that block GA biosynthesis are, for example, onium compounds (chlormequat chlo ride (CCC), mepiquate chloride, AMO-1618, phosphon D), pyrimi dines (ancymidol, fluoprimidol) and triazoles (paclobutrazol, unico nazole, triapenthenol). Another widely used retardant is daminozide; its precise mechanism of action is unknown, although it might affect 21 Jaana Luoranen gibberellin biosynthesis (Arteca 1996). In general, growth retardants are used for producing shorter, and often sturdier, seedlings (Arteca 1996). Plant growth retardants have been used for several years to modi fy plant development, especially in agriculture, floriculture and hor ticulture. The effects of growth retardants have been studied in the production of forest tree seedlings such as container Norway spruce (Dunberg and Eliasson 1972), white spruce (P. glauca (Moench) Voss) (Weston et al. 1980, van den Driessche 1996), lodgepole pine (Pinus contorta var. latifolia Engelm.) (Weston et al. 1980, van den Dries sche 1996) and Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) (van den Driessche 1990, 1996) seedlings. Aphalo et al. (1997) studied the effects of growth retardants on the production of silver birch container seedlings. The effects of drought stress on cessation of height growth and on frost hardening differ between species. The duration and the level of drought are also important. However, drought stress promotes cessa tion of height growth and frost hardening only in certain species. For example, it can hasten cessation of height growth in western red cedar ( Thujaplicata Donn) (Grossnickle et al. 1988) but not in Douglas-fir (van den Driessche 1969, Mac Donald and Owens 1988) or western hemlock (Tsuga heterophylla (Raf.) Sarg.) seedlings (Grossnickle et al. 1988, Grossnickle et al. 1991 a). Drought stress also reduces diam eter growth (Mac Donald and Owens 1988). In conifers, combined SD and drought stress usually gives worse results than one of these meth ods alone (Grossnickle et al. 1988). In red osier dogwood seedlings, however, their interaction is stronger than either one alone (Chen and Li 1978). When western hemlock seedlings are planted in spring, the combination of SD treatment and drought stress produces seedlings that perform well even in harsh field conditions (Grossnickle et al. 1991 a); but in autumn the frost hardiness of western hemlock or Douglas-fir seedlings treated in this way may be reduced (Grossnickle et al. 1991 a, Mac Donald and Owens 1988). According to the comparison of methods used to control height growth, induce cessation of height growth or induce frost hardening in several species, the most promising alternatives for silver birch container seedlings seem to be SD treatment and chemical growth retardants. 1.2.2.2 Monitoring of seedling quality in the nursery High quality seedlings are usually defined as those that will survive well and grow vigorously after planting (Ritchie 1984). Testing of seedling quality is needed to avoid planting damaged stock, to con vince costumers that seedlings are high quality, to ensure the correct 22 Control of Growth and Frost Hardening of Silver Birch Container Seedlings timing of nursery practices and to improve nursery practices (Duns worth 1997). The production of high quality seedlings demands knowledge of the timing of growth and the development of seedlings in the nur sery. Ritchie (1984) categorized the characteristics of seedlings that reflect quality (i.e., performance potential) as either performance at tributes or material attributes. Material attributes (seedling morpho logy, mineral nutrient status etc.) are rapidly measurable, either di rectly or indirecdy, whereas performance attributes are physiologi cal tests (root growth potential (RGP), frost hardiness) that evaluate the seedling's response to certain environmental variables (Ritchie 1984, Mexal and Landis 1990, Mattsson 1997). Puttonen (1997) clas sified seedling quality tests for the purposes of evaluating general applicability, readiness for lifting and prediction of planting success. He also rated these tests using the scoring criteria of Hawkins and Binder (1990) and concluded that morphological characteristics would be the best overall test method for grading seedling stock and predicting field performance. Until now, performance attributes have not been necessary for mea suring seedling quality routinely in Finnish nurseries. In some other countries with a large number of species, a range of site conditions, high costs of seedling establishment or history of plantation failures, operational testing of seedling physiology is used (Dunsworth 1997). Nowadays, the need for testing has also increased in Finland, as grow ing methods are changing and, more often, a nursery needs to grow several crops per season and produce stock from more seed origins. This means that there is a need to estimate the frost hardening of seedlings and thus ensure the correct timing of nursery practices (e.g., readiness for frozen storage). Several methods have been used to evaluate whether seedlings are ready for cold or frozen storage or, in the case of "extended green house culture", to determine the time for moving seedlings out of the greenhouse in autumn. In the Pacific Northwest of the United States frost hardening of conifer seedlings is determined by a whole seedling viability test (Tanaka et al. 1997). In British Columbia, seed ling storability is determined visually by scoring injuries after ex posure to -18° C (Binder et al. 1996, Dunsworth 1997). In Ontario, the electrolyte leakage method has also been used after seedlings have been exposed to-15°C (Colombo 1997). The usefulness of chlo rophyll fluorescence has also been studied (Binder et al. 1996). In evergreen conifer species, both electrolyte-leakage and chlorophyll fluorescence methods seem to be more sensitive and quantitative than visual scoring. According to Binder et al. (1996), chlorophyll fluores cence is also suitable for deciduous species, if the seedlings have enough chlorophyll in their stems. In Sweden, several methods are used routinely. For example, frost hardening is monitored by deter- 23 Jaana Luoranen mining the dry matter content (water content) in the apices of seed lings (without exposure to freezing), root growth potential or electro lyte leakage after exposure to freezing (Dunsworth 1997). In Ontario, visual assessment, measurement of the chlorophyll fluo rescence of the foliage and the RGP test are used in the spring to assess the physiological quality of nursery stock prior to planting (Sampson et al. 1997). In the United States, the RGP and seedling viability tests are also used for minimizing the planting of poor quality stock (Tanaka et al. 1997). Which methods are selected depends on the need for evaluation, the tree species and the resources available for assessment. It is pos sible to determine the time of growth cessation, and thus the trigger ing of frost hardening of shoot by regular measurement of height. In breeding studies, yellowing and abscission of leaves in silver birch seedlings is used to compare provenances (e.g. Veiling 1979). How ever, to my knowledge, there are no studies in which the actual rela tionship between frost hardiness and visual scored development in leaf yellowing has been investigated. Blackburn and Brown (1988) observed that birch trees with some green leaves were not as frost tolerant as trees with yellow and/or abscised leaves. In that study, however, frost hardiness was assessed only once and the actual rela tionship between yellowing and abscission of leaves and frost har diness was not determined. During frost hardening, the water content in seedling tissues de creases. This information is utilized when frost hardening of seed lings has been monitored indirectly by regular determination of wa ter content (dry matter content) in the apices of unfrozen seedlings (e.g. Rosvall-Ähnebrink 1977, Colombo 1990, Calme et al. 1993). This method is commonly used in Swedish nurseries. When water content reaches a certain species-specific value, seedlings are con sidered to be frost hardened enough to withstand frozen storage or light frosts (Hulten 1980, Calme et al. 1993, 1995). It has been pro posed that the critical value would be, for example, 68-66 % (32-34 % as dry matter content) in Norway spruce, 69% (31 %) in Scots pine (Pinus sylvestris L.) (Hulten 1980), 60% (40%) in black spru ce, white spruce and Jack pine (P. banksiana) (Calme et al. 1993), and 55% (45%) in deciduous species (Calme et al. 1995). The rela tionship between water content and frost hardiness is dependent on seedling origin and on previous growing conditions, as has been shown in Scots pine seedlings (Toivonen et al. 1991). The electrolyte-leakage (conductivity, ion leakage) method is wide ly used, and the electrical-impedance spectroscopy used less, to de scribe freezing damage. Both methods are based on the principle that plasma membranes in tissues injured by freezing allow electro lytes to leak from the symplast into the apoplastic water (Ritchie 1991). In the electrical impedance method, changes in extracellular 24 Control of Growth and Frost Hardening of Silver Birch Container Seedlings resistance caused by freezing are measured (Glerum 1985, Repo 1994). This method is quite rapid (Ritchie 1991), but the results are affected by type of tissue and by some physical factors, such as tis sue size (stem diameter) and temperature (Glerum 1985, Repo 1994). 1.2.3 Relation of root growth and root activity to planting windows The field performance of planted seedlings is strongly affected by the seedling's ability to grow roots into the surrounding soil and to resume water and nutrient uptake (Kozlowski and Davis 1975). Af ter planting, root growth is controlled by the growth phase, soil mois ture, nutrition, light intensity and temperature (Lopushinsky and Max 1990, Noland et al. 1997). Photoperiod, on the other hand, does not directly affect root growth (Kelly and Mecklenburg 1978, Bigras and D'Aoust 1993); but the indirect effect of photoperiod on root growth is mediated by net daily carbon assimilation of a seedling. The timing of root growth during the first growing season has been investigated mainly with conifer species. In northern Finland, the roots of container seedlings of Norway spruce, Scots pine and Sibe rian larch grow throughout the summer in which they germinate (Kin nunen and Lähde 1972). Time of germination also affects timing of root growth. Roots of Scots pine and Norway spruce seedlings sown in May stopped growing at the end of August, but roots of seedlings that germinated later continued to grow until mid-September (Valta nen et al. 1986). In the study of Mattsson (1986) the roots of Scots pine container seedlings that had germinated earlier grew more ear ly in the summer than did the roots of seedlings that germinated la ter. In container-grown birch seedlings, the roots grow most in July, declining linearly towards autumn (Rikala 1996). In some deciduous species, the roots can grow during bud dor mancy (Farmer 1975, Webb 1976, Westergaard and Eriksen 1997). However, in birch species, carbohydrate reserves, those stored mainly in the roots, are used to produce leaves in spring; and thus new roots do not grow until the leaves are able to produce carbohydrates (Abod et al. 1991). In the spring of the second year, the root growth of silver birch seedlings begins slowly and speeds up just after the new leaves are fully expanded (Rikala 1996). The temperature in the root zone strongly affects the growth and activity of the roots. Low soil temperatures decrease uptake of water (Lopushinsky and Kaufmann 1984, Örlander and Due 1986) and nutrients by roots (Bowen 1970, Karlsson and Nordell 1996). This is due to the slowed metabolism of roots and increased viscosity of water, when the temperature in the root zone is low (Ryyppö et al. 1998 a). Water and nutrient uptake of old roots is also poorer than Jaana Luoranen 25 Figure 1.3 a) Changes in night length at the latitude of Suonenjoki, b) daily mean air (solid line) and soil (at a depth of 15 cm) (broken line) temperatures as mean values for 25 or 15 years, respectively, at Suonenjoki Research Sta tion during the growing season, c) Accumulation of temperature sum (d.d.) dur ing the coolest (1987) and warmest (1972) growing season and the average accumulation of tempera ture sum in 1972-1999 observed at Suonenjoki Research Station. that of new, growing roots (Häussling et ai. 1988, Ryyppö et ai. 1994). In spring, this means that initiation and elongation of root tips are slow or are prevented completely (Lopushinsky and Kaufmann 1984, Vapaavuori et ai. 1992, Ryyppö et ai. 1994, 1998 a, Landhäusser et ai. 1996), which decreases the production of dry mass in the whole seedling (Lyr 1996). In summer, the soil temperature (Figure 1.3b) is more conducive to rapid rooting of planted birch seedlings, becau se the optimal temperature for root growth is about+ls°C (Lyr 1996). At present, most birch seedlings are planted in spring. The argu ment in favor of spring planting is that dormant seedlings are more resistant to stress (Heikinheimo 1941, Huuri 1974). This is true for bareroot seedlings, which run a great risk of exposing the roots to drying or mechanical damage (Heikinheimo 1941, Huuri 1974). Spring planted seedlings do not take up water efficiently. In decid uous species the roots cannot grow before bud burst (Abod et al. 1991, Rikala 1996) and the water and nutrient uptake capacity of the suberized, old roots is weak (Bowen 1970, Örlander and Due 1986, 26 Control of Growth and Frost Hardening of Silver Birch Container Seedlings Häussling et ai. 1988, Karlsson and Nordell 1996, Ryyppö et ai. 1994). In addition, after planting container seedlings have poor root soil contact when all roots are in the peat plug. This also limits water uptake by seedlings (Burdett 1990). In May and early June, when it is thought that there is enough water to ensure rapid rooting, soil temperatures are low but air tem peratures are relatively high (Figure 1 ,3 b). In these conditions, where water uptake by the roots is low but the water and nutrient needs of expanding leaves may be quite high (low air humidity and high evap oration), newly planted seedlings run a great risk of desiccation. The situation can be problematic for seedling establishment: seedlings have to assimilate carbon so that roots can grow, but photosynthesis depends on uptake of water and nutrients by roots (Burdett 1990). An alternative for solving the above-mentioned problems of slow rooting and increased risk for drought stress in spring is summer planting. In fact, one reason for raising container seedlings was to expand the planting season without increased risk of root damage due to drought (Leikola and Raulo 1973, Kinnunen et ai. 1974). So far, container seedlings are mainly planted in spring. In summer, the potential for root growth is high and soil temperatures (Figure 1 .3 b) should favor both growth and water and nutrient uptake by the roots. Planting throughout the growing season has been studied in all im portant silvicultural tree species in Finland and Sweden (see Luora nen 1998 and references therein), but only a little in birch seedlings (Lähde and Raulo 1977, Raulo et al. 1994, Rikala 1996). In the late 1980's, however, silver birch container seedlings were summer planted in some small-scale trials in central Finland with good re sults (A. Ruha and M. Suihkonen, personal communication). After summer planting, growth in later years has increased when seedlings were planted at the time of most rapid root growth, the seedlings were relatively short (Kinnunen et al. 1974, Valtanen et al. 1986, Kinnunen 1989, Rikala 1996) and the soil temperature has been near the seasonal maximum (Rikala 1996). However, studies of summer planting are still scarce and not all factors affecting seed ling production for summer planting and seedling establishment af ter planting have been evaluated. 27 Jaana Luoranen Figure 1.4 Schematic presentation of methods which can be used to avoid or solve problems concerning production and planting of silver birch container seedlings. The numbers after the titles in the picture tell the numbers of chapters in which each subject is described. 1.3 Methods for controlling the development of silver birch container seedlings The problems in the forestation chain of silver birch container seed lings are linked together. The cultural practices carried out in a nur sery affect the field-performance potential of seedlings. On the other hand, if one wants to change the planting window, the cultural prac tices used to produce acceptable seedlings in a nursery must be altered accordingly. Currently, however, our understanding of the annual growth cycle of silver birch seedlings is inadequate. When the growth cycle is better understood, it will be possible to try to find methods either to regulate the growth cycle or to avoid the problems. Figure 1.4 presents alternative ways to solve problems concerning the pro duction and planting of silver birch container seedlings. 28 Control of Growth and Frost Hardening of Silver Birch Container Seedlings One possibile way to solve problems of excessive height growth, delayed height growth cessation and delayed frost hardening is to avoid them. Monitoring of seedling development helps to direct nur sery practices to control growth. In this way it may also be possible to predict the risk of frost damage and keep seedling quality high. Another way to avoid these problems in a nursery is to shorten the production time in the nursery, i.e., to plant seedlings earlier or sow later. In this way it may also be possible to change the practices for production and planting of seedlings so that nurseries can respond more quickly and flexibly to fluctuations in demand for seedlings, enlarge the planting window and possibly decrease the costs of refor estation. However, there may exist other problems when these prac tices are changed: 1) how well unhgnified or only partially lignified seedlings tolerate mechanical stress during lifting, transporting and planting; 2) how summer-planted seedlings frost harden; and 3) whether there is a risk of delayed frost hardening in seedlings sown later in summer. It is also possible to control height growth of birch seedlings by retarding their growth during rapid growth phase or by inducing ear lier cessation of height growth. Growth can be retarded both by tim ing and changing nursery practices, such as irrigation and fertiliza tion. On the other hand, the growth of seedlings can be inhibited by reducing cell division and enlargement by applying chemical growth retardants. Another possible method is to induce cessation of height growth. An advantage of this method could be that with the same method it is also possible to hasten the frost hardening of seedlings. These methods might also be needed to increase stress tolerance of seedlings if duration of nursery growing is shortened. 29 Jaana Luoranen 1.4 Aims of the study The objective of this study was to evaluate possibilities to change the forestation chain of silver birch container seedlings so that it would be possible to respond more quickly and flexibly to fluctuations in the demand for silver birch seedlings and to guarantee the field per formance of seedlings. The objective was to find methods to control seedling development during the first growing season so that it would be possible: I. to decrease the risk of frost damage to silver birch seedlings in nurseries during the first autumn frosts, 11. to control the height growth of seedlings without negative influ ences on root and diameter growth or field performance, and 111. to maintain or increase the field performance of seedlings com pared to container seedlings grown and planted according to pre sent practices. To achieve these objectives 1. timing of height (Sections 3.1 and 4) and diameter growth, dry mass partitioning (Sections 3.1 and 4.2) and frost hardening (Sec tions 3.2 and 4.2) of silver birch container seedlings were char actrized during the first growing season; 2. methods of monitoring frost hardening in nurseries were compared to find suitable methods for operational use in nurseries. To do this the frost hardiness (Sections 3.2, 4.2 and 5) of birch shoots was assessed, water content was determined and leaf yellowing and abscission were scored (Chapters 3-5); 3. applications of growth retardants (Section 4.1) and short-day treat ment (Section 4.2) were studied as methods of regulating growth and frost hardening of birch seedlings; 4. the ability seedlings to grow roots into the surrounding soil (root egress) during the first year was monitored to determine the best planting windows and hence improve the ability of seedlings to survive and grow after planting (Sections 3.1 and 4.2); 5. the possibility of expanding the planting window of silver birch container seedlings by planting growing, leaf-bearing seedlings in summer was investigated (Sections 4.2 and 5). More specific objectives will be presented in each part of the study. 30 Control of Growth and Frost Hardening of Silver Birch Container Seedlings 2 Seedling material, study con ditions and common methods 2.1 Introduction This chapter presents methods used in several chapters of the study. If a method was used in only one part of the work, the method is described in that chapter. 2.2 Growing conditions These studies were carried out during 1995-1999. The studies on seedling development (Chapter 3) and growing methods (Chapter 4) were done at Suonenjoki Research Station (62°39'N, 27°03'E, al titude 142 m asl); and in all those studies, seeds of silver birch from seed orchard 379 (M 29-92-0001) were used. The seedling material used in the summer planting experiment is presented in Section 5.2. Daily mean and minimum temperatures as well as daily precipita tion at Suonenjoki Research Station for each year are presented in Figure 2.1. In each crop, accumulative temperature sums (threshold value +5 °C) were calculated from the time of sowing using mean daily temperature data collected by thermographs (Lambrecht) placed at seedling level. 2.3 Morphology and nutrient analysis Height growth of the stem was monitored weekly by measuring the same randomly selected seedlings in each treatment. Height (H) was measured (accuracy 1 mm) from the surface of peat (retained as a root collar) to the tip of the shoot apex. Diameter (D) (accuracy 0.01 mm) was measured 2 cm above the peat surface. Sturdiness (5) of seedlings (mm nr 1 ) was calculated as For monitoring dry mass partitioning, shoots (stems and branches), leaves and roots were separated, the roots washed, and all three sec tions dried for 24 h at 105° C before they were weighed (accuracy 1 mg). Samples to be used for nutrient analysis were dried for 48 h at (2 . 1} H Jaana Luoranen 31 Figure 2.1 Daily mean (thin solid line) and minimum (dotted line) air temperatures (°C) and precipitation (bars) (mm) at Suonenjoki Research Station from l May to 31 October in 1995, 1996, 1997, 1998 and 1999. Air temperatures were compared to the 27-year average air temperatures (thick solid line) recorded at Suonenjoki Research Station. 60° C. Nitrogen concentration of samples was determined with a LECO CHN-600 analyzer (Leco Co, USA); and concentrations of P, Ca, K, Mg, Cu and B of samples were determined from dry-digested 2 M HCL samples (Halonen et ai. 1983) using plasma emission spectrophotometric analysis (ICP, ARL 3800). 2.4 Water content The procedures used for determination of water content were the same as those described by Rosvall-Ahnebrink (1977) and Colombo (1990) for conifer seedlings and by Calme et al. (1995) for hard wood seedlings, but with slight modifications. Seedlings were watered the day before sampling. The uppermost 10 cm (Section 3.1) or scm (Sections 4.1, 4.2) of stem from randomly selected seedlings were cut in the morning at 0730 h-0900 h. The cut apices were put into plastic bags and weighed (accuracy 1 mg) without leaves within two 32 Control of Growth and Frost Hardening of Silver Birch Container Seedlings hours from sampling (fresh weight, FW), then put into paper bags, dried in an oven for 24 h at 105° C and weighed again (dry weight, DW). Here water relations are expressed as water content (WC) as de scribed by Timmis and Fuchigami (1982) and Colombo (1990) 2.5 Root growth potential and root-egress test Root growth potential (RGP) is defined as the ability of a tree seed ling to initiate and elongate new roots within a prescribed period of time in a standard environment that is optimized to promote root growth (Simpson and Ritchie 1997). Since RGP consists of the ini tiation, elongation and thickening of roots and can be measured quick ly, in this study, RGP is expressed as the dry mass of roots grown out from the peat plug during the test period. When the effects of treat ments were compared (Sections 4.1 and 4.2: the second crop), RGP was used as a conventional method of evaluating the vigor of seed lings in spring. RGP was tested in frozen stored and thawed seedlings in spring. RGP was determined by planting seedlings in 0.751 (the first crop in Section 4.2) or in 2.2 1 (diameter at soil surface 17 cm, experiment in growth retardant in Section 4.1) plastic pots filled with sand. Seed lings were grown in a heated and lighted (day/night temperature: +2O - +22 °C/+l5°C, photon flux density (photosynthetically active radiation, PAR) 150 (Jmol-nr2-s _I from metal halide lamps (HQI 400W Power Star, Osram), photoperiod 18 h) greenhouse for four weeks. All seedlings were watered three times a week with tap wa ter. At planting, the height and the diameter of the root collar were measured. After three weeks, all seedlings were harvested, and the height and diameter were measured again. Roots growing out of the peat plug into the sand ('new' roots) were cut and washed; roots in the peat plug ('old' roots) were washed, and the shoot was divided into stem and leaves. All plant parts (new and old roots, stems and leaves) were then dried in an oven for 24 h at 105° C and weighed (accuracy 1 mg) separately. In an attempt to find suitable planting windows for silver birch seedlings grown in containers (Sections 3.1 and 4.2: the first crop), the ability of seedlings to root egress at different times of the year was evaluated by using a modificated RGP test (root-egress test). Seedlings were planted in 0.75 1 (the first crop in Section 4.2) or in 2.2 1 (diameter at soil surface 17 cm, growth rhythm experiment in .c^)"00 <2 ' 2) 33 Jaana Luoranen Section 3.1) plastic pots filled with sand and grown as in the RGP test in spring, except that the greenhouse was unheated and under natural photoperiod and light conditions. The seedlings were water ed as necessary (at least twice a week) with tap water. Height, root collar diameter and dry mass of shoots, leaves and roots were mea sured as described above, except that roots in the peat plug ('old' roots) were excluded. 2.6 Statistical analysis 2.6.1 Statistical tests The Bartlett test was used to determine the homogeneity of the vari ances and the Lillefors test to determine whether the data were nor mally distributed. When needed, log- or other transformation was used to obtain normal distributions before analysis of variance. When the differences between treatments in weekly measured height or water content were analyzed, an analysis of variance procedure for repeated measurements was used (Chapter 4). Differences between treatments or weeks were analyzed with Tukey's HSD test or with a t-test. If no transformation was found to transform the distributions to normal and to homogenize the variances, the effects of treatments were analyzed with non-parametric methods: Kruskal-Wallis one way analysis if several treatments were compared, and Mann-Whit ney U-test if only two treatments were compared. Results shown in the figures and tables are back-transformed. Possible linear correla tion between variables was analyzed with Pearson's linear correla tion analysis (Sections 3.1 and 3.2). Data were analyzed with Systat 5.05 statistical software (Systat Inc, USA) (Wilkinson et al. 1992). 2.6.2 Estimation of growth curves Development of first-year seedlings was monitored in Sections 3.1 and 4.2. Height, diameter and dry mass of the shoot and roots of seedlings (later referred to as morphological attributes) increases with increasing age to a certain point sigmoidially (see Figures 3.1a,b and 4.7). For estimating growth responses in different crops, a simple four-parameter logistic model (Ratkowsky 1990) was fitted to the data. The developmental function is yt = f (x) +e m (2.3) ui r 4*/ + f 1 + e 34 Control of Growth and Frost Hardening of Silver Birch Container Seedlings In the function, y. is the value of a morphological attribute at the age and m, b, c and / are equation parameters. The upper and lower asymptotes of the function are determined by parameters m and I, respectively. Parameter I is the lower asymptote of function, the lev el of attribute at the beginning of development. Parameter m expresses the amount of growth between the lower and upper asymptotes. Thus, m + 1 indicates the upper asymptote, which tells the final value of an attribute in the end of growing season. Parameter c, the point of in flection, expresses the age of the maximum growth rate. Parameter b expresses the slope of the curve at the point of inflection, i.e., a maximum growth rate. This basic equation was used to estimate the development of diameter. For the height and dry mass of shoot and roots the estimated equa tion was This equation was used due to the fact that these attributes were mea sured not from the time of sowing but after transplantation. It was assumed that at the age of transplantation (x ( ) a value of an attribute is zero. The variation in the measured attributes increased with time (see Figure 4.2). To homogenize the variances both sides of equation (2.3) were divided by the statistical weight (w) In the above equation, / is the estimate of fix.) given to the first iteration and v is some fixed value (in the most of cases 10). Cessation of growth was defined by estimating the age x p . The height, diameter or dry mass at age x p is the proportion p of the total amount of growth m (expressed as fixp) = prri). An inverse function of equation (2.3) is x p is the new parameter, and one of the original parameters has to be removed and expressed by x p and other parameters to avoid overpa rametrization. The easiest way is to replace c or b. The slope b of equation (2.6) is mm A\ y ' ~ i +e Mc ~x:> ~ i +e b(c ~x ' > +£ ' W = f(v-f) (2.5) lnf-f-r —l| Vw J, JC = + c (2.6) y b b = (2.7) x p -c 35 Jaana Luoranen By substituting equation (2.7) into equation (2.3) and by using fix J = pm, we get Or replacing the inflection point c and the equation for estimation is To estimate the growth cessation or the end of dry mass accumula tion, equation (2.8) or (2.10) with a value of 0.95 for p was fitted. Development of the dry mass of leaves consists of two parts: in crease in dry mass and abscission of leaves. Both parts change with time to a certain point as a logistic function (see Figure 4.7 c). Equa tion (2.4) can be used to express the increase in dry mass of leaves. Abscission of leaves is a decreasing sigmoidial curve with parame ter h to the point of inflection and parameter k to the slope of the curve. The time when the growth curve changes to an abscission curve can be calculated using change point (Lappi 1993 and references therein). When k < 0, the equation for abscission of leaves is In change point r, the growth curve changes to an abscission curve. At this point r, or or Ä = 77 + £ (2.8) -In --1 1+ e e c = ~ + -W|LJ ,2.9, * =— T (2-10) b ,A~r- 1) x xp + - 1+ e ' yi=h(x)+ £ m-, (2.11) " l + +£ f(x ) = fi{r ) (2-12) m m _ m 2 l + eb{c ~r) l + ~l + ekr 37 Jaana Luoraneri 3 Development of first-year silver birch container seedlings 3.1 Shoot and root growth 3.1.1 Background and study objectives In nurseries, knowledge of the timing of growth and partitioning of dry mass into different parts of seedlings during the growing season helps nursery managers to time and plan growing practices (sowing time, growing density - target height, outmoving time, fertilization, etc.). However, only Rikala (1996) has published information on growth of first-year silver birch seedlings grown in containers in a nursery. Timing of shoot but especially root growth is also important for choosing planting windows for silver birch seedlings. For seed ling survival and rapid growth after planting, it is important that roots grow quickly from the peat plug into the surrounding soil (Kozlows ki and Davis 1975). The ability of a seedling to grow roots in an environment that is highly favorable for root growth has been measured by a test of root growth potential (RGP) (Mexal and Landis 1990, Ritchie and Tana ka 1990). RGP is a seedling performance attribute (Ritchie 1984). In general, the higher the RGP of a seedling the greater its chance of survival (Burdett 1987). RGP has been studied in deciduous bare root seedlings before frozen storage or in spring after storage (e.g. Farmer 1979, Webb and von Althen 1980, Struve 1990). During the first growing season, the ability of the roots of silver birch seedlings to egress from the peat plug is highest in July and early August (Ri kala 1996). In the first summer, root growth is also greater than with traditional spring planting (Rikala 1996). In the present study, the root egress (the dry mass of outgrown roots) was used as a parame ter for measuring and comparing the root growth of seedlings at dif ferent times of the first growing year. The objective of this part of the study was to investigate: 1) the time course of and relationship between height- and diameter growth and partitioning of dry mass to shoot, roots and leaves from trans planting to abscission of leaves in silver birch seedlings growing in a nursery and 2) ability of seedlings to grow roots out from the peat plug (root egress) at different times of the growing season. 38 Control of Growth and Frost Hardening of Silver Birch Container Seedlings 3.1.2 Material and methods In three crops, seeds of silver birch were sown on peat-filled germi nation trays in a greenhouse on 15 April (later this crop is referred to as APRDL96) and on 2 May (MAY 96) 1996 and on 5 May 1997 (MAY 97). Germinants were transplanted into Plantek 25 trays (25 cavities per tray, 380 cm 3 per cavity, 156 cavities per m 2, Lännen Plant Systems, Finland) filled with fertilized sphagnum peat (Vapo, XL, Finland) on 3 May (APRIL 96) and on 23 May (MAY 96) 1996 and on 29-30 May 1997 (MAY 97). In each crop, the seedlings were irrigated according to normal nursery practice by maintaining the water content of the peat at 40-70 percent by volume during the growing season. In 1996, in both crops the seedlings were irrigated and fertilized after they had been moved to the same outdoor grow ing area according to the same regime; in late summer the water content in peat plugs was lower in APRIL 96 than in MAY 96. In the APRIL 96, MAY 96 and MAY 97 crops, seedlings were fertilized with liquid fertilizer 6, 5 and 4 times, respectively. The total amounts of nutrients (including both base and liquid fertilizer) given were 144 mg N, 51 mg P and 162 mg K per seedling plus micronutrients for APRIL 96, 132 mg N, 47 mg P, 150 mg K per seedling plus micro nutrients for MAY 96, and 87 mg N or 36 mg P and 106 mg K per seedling plus micronutrients for MAY 97. Seedlings were grown in a greenhouse until they were moved to an outdoor growing area on 10 June or 20 June in 1996 and on 25 June in 1997. All seedlings were sprayed twice with Bayleton 25 (triadimefon 0.05 %) against birch rust (Melampsoridium betulinum (Pers.) Kleb.): on 12 July and 1 August in 1996 and on 29 July and 29 August in 1997. Seedlings were sprayed against fungi that cause stem lesions with Tilt 250 ES (propiconazole 0.5 1 ha -1 ) on 16 July 1996 and with Shirlan (fluazi nam 0.41 ha -1 ) on 14 August 1996. The seedlings were sprayed against aphids with Ripcord (cypermethrin 0.05 %) on 28 July 1997. Air temperatures at seedling level (about 10 cm from the ground) were monitored both years. The accumulation of the temperature sum (threshold value +5 °C) was calculated from the time of sowing. The climate in July 1996 was cool (14° C average temperature, 15-year mean daily temperature in Suonenjoki = 16° C) and rainy (monthly cumulative sum was 137 mm, 15-year mean monthly cumulative sum = 84 mm), but August was warmer (17° C average temperature, 15- year mean daily temperature 14° C) and precipitation was low (cumu lative sum = 17 mm, 15-year mean = 60 mm) (Figure 2.1). The 1997 season was warmer than the 15-year average (average temperatures were 16, 19, 17° C, 15-year mean daily temperatures were 14.5, 16, 14° C in June, July and August, respectively) (Figure 2.1). In each crop, there were 30 trays (25 seedlings in each) divided into 10 blocks, i.e., a total of 750 seedlings. In the crops of APRIL 96 39 Jaana Luoranen and MAY 96, the height of the stems of the same two randomly se lected seedlings per tray in all 30 trays were measured weekly from 2 weeks after transplanting to the middle of September. From 4 Sep tember to 9 October 1996, yellowing and abscission of leaves was recorded weekly. In each crop, two seedlings per block, a total of 20 seedlings each week, were harvested from 15 May to 15 October, from 3 June to 14 October, and from 9 June to 13 October for APREL96, MAY 96 and MAY 97 crops, respectively. Height and diameter were measured, and dry mass of the shoots (stem and branches), leaves and roots was determined. In 1996 and 1997 the roots were washed by different people, but during each year the person washing the roots was the same. Thus the results of different years are not totally comparable. Roots that grew into the ground often broke when the seedlings were lifted. Thus the results do not indicate the total dry mass of the roots. The aim of this study was to examine the amount of roots in seed lings for planting and it was not considered necessary to include bro ken roots in the samples. At the last harvesting date in October 1996, the shoot and root samples in each block were pooled for nutrient analysis. Nutrient concentrations in the stem and roots in both crops in 1996 are presented in Table 3.1. In 1997, water content in the uppermost 10 cm of all harvested MAY 97 seedlings was determined from 11 August to 13 October. For the root-egress test, 20 randomly selected seedlings per crop were planted weekly into 10 blocks from 26 June to 2 October and Table 3.1 Nutrient concentrations of stems and roots in seedlings from crops sown in April or in May 1 996. Nutrients were analyzed separately from stems and roots at the end of the first year (on 2 October or on 30 September 1 996). Values are the means of five samples (drawn from a pooled sample of two seedlings) and the standard error of the means. Nutrient Stem Roots April May April May Nitrogen, g kg -1 9.8 + 0.36 9.0 ± 0.30 7.8 ± 2.24 8.0 ± 0.28 C/N 58 ± 1.93 71 + 6.23 63 ± 2.00 68 ± 2.34 Phosphorous, g kg" 1 1.32 ±0.05 0.99 ±0.13 1.53 ± 0.09 1.53 ±0.02 Potassium, g kg -1 4.18 + 0.09 3.18± 0.42 4.05± 0.20 4.11 ±0.08 Calcium, g kg -1 2.22 + 0.21 1.46 ±0.21 2.87± 0.19 3.11 ±0.07 Magnesium, g kg -1 0.89 + 0.06 0.63 ± 0.08 1.13 ±0.05 1.33 ±0.03 Manganese, mg kg" 1 68.9 + 5.3 56.5 ± 9.4 58.5 ± 5.0 64.9 ±6.50 Copper, mg kg -1 3.0 + 0.2 3.5 ±2.1 11.5 ± 0.4 11.0± 1.1 Zinc, mg kg -1 77.8 + 7.0 52.6 + 8.4 57.2 ±3.3 56.0 ±2.88 Iron, mg kg -1 28.2 + 2.2 22.9 ± 3.3 263 ±44 210 ±36.3 Boron, mg kg -1 8.5 + 0.5 6.6 ± 1.0 7.1 ±0.74 7.8 ±0.19 40 Control of Growth and Frost Hardening of Silver Birch Container Seedlings Table 3.2 Estimated parameters of the growth curves for height (Equation 2.4), diameter (Equation 2.3) and dry mass of shoot and roots (Equation 2.4) and the parameters of the development curves for the dry mass of leaves (Equation 2.15) in seedlings sown on April 1996 (APRIL 96), on May 1996 (MAY 96) and on May 1997 ( MAY 97). The dates and accumulated temperature sums at the age of the maximum rate of growth (parameter c) and at the age when growth ceased U0.95) for each variable and dates and accumu lated temperature sums at the age when leaves began to shed (r) and when all leaves were shed (jco.os) for dry mass of leaves. Parameters of estimated growth curves Variable Crop m b c I Height APRIL96 72 0.06 80 124 MAY96 67 0.07 78 120 MAY97 84 0.09 69 101 Diameter APRIL96 6.3 0.05 73 129 0.1 MAY96 6.4 0.06 75 124 0.1 MAY97 6.9 0.07 63 104 0.1 Shoot APRIL96 5.0 0.08 107 145 MAY96 4.8 0.09 101 134 MAY97 5.4 0.09 92 124 Roots APRIL96 3.7 0.07 103 146 MAY96 3.5 0.08 98 134 MAY97 2.6 0.09 90 123 Parameters of estimated development curves for dry mass of leaves Variable Crop m b c *0.95 k h r *0.05 Leaves 2.4 0.09 75 107 -0.13 168 154 191 MAY96 2.8 0.10 78 106 -0.23 154 140 167 MAY97 2.2 0.17 63 81 -0.15 147 119 167 Dates and accumulated temperature sums at the estimated ages Max. growth rate Cessation of growth Beginning of shedding All leaves shed c *0.95 r *0.95 Variable Crop date d.d. date d.d. date d.d. date d.d. Height APRIL96 4 Jul 851 17 Aug 1288 MAY96 19 Jul 767 30 Aug 1212 MAY97 13 Jul 866 14 Aug 1306 Diameter APRIL96 27 Jun 790 22 Aug 1351 MAY96 15 Jul 741 2 Sep 1239 MAY97 7 Jul 878 17 Aug 1324 Shoot APRIL96 31 Jul 1133 7 Sep 1485 MAY96 10 Aug 987 12 Sep 1265 MAY97 5 Aug 1194 6 Sep 1536 Roots APRIL96 27 Jul 1094 8 Sep 1489 MAY96 7 Aug 949 12 Sep 1265 MAY97 3 Aug 1165 5 Sep 1527 Leaves APRIL96 29 Jun 810 31 Jul 1133 16 Sep 1493 23 Oct 1532 MAY96 18 Jul 761 15 Aug 1046 18 Sep 1269 15 Oct 1306 MAY97 7 Jul 801 25 Jul 1041 1 Sep 1500 19 Oct 1634 Jaana Luoranen 41 Figure 3.1 Development of dimensions and dry mass in birch seedlings sown on 15 April or 2 May 1 996, or 5 May 1997. a) Height, b) diameter and c) changes in dry mass of leaves, d) shoot (including stem and branches) and e) roots of seedlings were measured weekly from transplanting to the middle of October 1996 or 1997. Each symbol indicates the mean of 10 blocks (mean of 2 seedlings per block) on each harvesting date. from 1 July to 30 September 1996 for APRIL 96 and MAY 96, respec tively. Seedlings planted in sand-filled plastic pots were kept in an unheated greenhouse under natural photoperiod and light conditions for three weeks. In the greenhouse, the daily mean temperature dur ing the summer was about 2 1° C (the largest difference between mini mum and maximum temperature was 12° C in the middle of August, but otherwise the difference was 1-5° C) and the RH 58% (48-65%). On 14 October, seedlings for the root-egress test the following spring were packed in plastic bags and stored at -4° C. From 17 May until they were planted in the test on 27 May, the seedlings were thawed and kept in dark (4 d at 5°C). After 3 weeks, the seedlings were 42 Control of Growth and Frost Hardening of Silver Birch Container Seedlings Figure 3.2 Dry mass parti tioning to leaves, shoot (in cluding stem and branch es) and roots as a) dry mass (g) and b) percentage of total mass per birch seed ling sown on 2 May during the 1996 growing season. Dry mass of 20 seedlings (2 seedlings in 10 blocks) was measured weekly from 2 June to 13 October 1996. harvested and height, diameter, dry mass of shoot, leaves and new roots were assessed. Statistical analysis. The growth curves were fitted as described in Section 2.5. In the root-egress tests, due to heterogenic variances, differences between crops in each week were analyzed with the Mann-Whitney U-test. In the root egress test of the APRIL 96 seed lings, the weekly differences were tested by Kruskall-Wallis test and in the MAY 96 seedlings by ANOVA and Tukey's HSD test after lo garithmic transformation. 3.1.3 Results During the growing season, height, diameter and dry mass of stem and roots increased logistically with time for all crops (Table 3.2, Figure 3.1a,b). During the first part of the season, seedlings allo cated the most dry mass to the leaves and least to the stem (Figure 3.2). Dry mass of the leaves increased until the beginning of August. Diameter increased fairly linearly with increasing height of the stem (Figure 3.3 a), and the dry mass of the roots increased with increas ing dry mass of the shoot (Figure 3.3h). The relationship between dry mass (shoot and roots) and seedling dimensions (height and diam eter) was divided into two parts (Figure 3.3b-e): 1) until cessation of height growth in the middle or end of August, dry mass increased fairly linearly as height and diameter increased (Figure 3.3b-e); 2) then the dry mass of the stem and roots increased for about a month until the decrease in dry mass of the leaves began in the be- Jaana Luoranen 43 Figure 3.3 Correlations between seedling dimensions and dry mass of different plant parts assessed at different times during the growing season from crops sown on 15 April and 2 May 1996, and on 5 May 1997. The first harvesting dates are on the left and time moves on toward the right, except in f) and g) where time moves back to the left in autumn due to leaf senescence. To help indicate the timing of changes in relationships, mean values obtained in the middle of August (week 33, the average time for cessation of height growth) were drawn larger and with black symbols. Each symbol (for explanations of sowing times, see Figure 3.1) indicates the weekly mean of 20 seedlings (2 seedlings in 10 blocks) for each variable of each crop. ginning or middle of September (Figures 3. 1 c and 3.3f,g). After that, leaf dry mass decreased due to leaf senescence (Figure 3.4) without a change in the dry mass of shoot and roots (Figure 3.3f,g). All leaves had dropped by the end of October (Figures 3.1 c and 3.4). In 1996, the height growth of seedlings sown in April and May stopped after the middle of August (Table 3.2). According to visual scoring, the leaves started to become yellow in the second week of September (Figure 3.4) and all leaves had been shed a month later. In 1997, the height growth of the MAY 97 seedlings stopped on 14 August (Table 3.2). After cessation of height and diameter growth, water content (measured from the seedling apex) in the MAY 97 seed lings decreased until leaf abscission began in the end of September (data shown only in Figure 6.4b). At the end of the growing season, the APRIL 96 and MAY 96 seed- 44 Control of Growth and Frost Hardening of Silver Birch Container Seedlings Figure 3.4 Yellowing and abscission of leaves (% of seedlings) in seedlings sown a) in April or b) in May 1996 and assessed weekly during autumn 1996. Leaf yellowing and abscission was recorded from 60 seedlings per crop according to the following classification: 1) all leaves green, 2) some leaves green, the rest yellow, 3) all leaves yellow, half or more fallen, 4) all leaves fallen. Figure 3.5 Growth of seedlings planted in root-egress test in summer 1996 and spring 1997. a) Dry mass of new (outgrown from peat plug) roots (g), b) growth in diameter (mm) and c) height (cm) of the stem of birch seedlings sown on 15 April or 2 May 1996 during the three-week test period. Bars indicate standard errors of the means for each day. (n =lO block means, 2 seedlings per block). 45 Jaana Luoranen lings were shorter than the MAY 97 seedlings (Figure 3.1 a), but there were no clear differences in diameter (Figure 3.1b), sturdiness or dry mass of the stem (Figure 3.3b, d) between crops. Root dry mass (Figure 3.3h) and root-to-shoot ratio in crops of 1996 were higher than in the MAY 97 crop. During the growing season, the sturdiness and root-to-shoot ratio of the seedlings decreased until the begin ning of July (data not shown). In the root-egress test, the dry mass of new roots was large until the middle of August (Figure 3.5 a). At the end of August, root growth had decreased but did not cease until the leaves dropped at the end of September. The decline in amount of new roots in late summer and early autumn was correlated to the cessation of height and diameter growth. The increments in height and diameter stopped at the same time at the end of August (Figure 3.5b, c). There were no differences between crops, except at the end of July when the APRIL 96 seed lings had more root growth (Kurskal-Wallis P<0.001). The following spring, root growth was the same (Kruskal-Wallis P=0.1 12) in both APRLL96 and MAY 96 seedlings (Figure 3.5). The height at planting affected height growth, since seedlings in the AP REL96 crop that were taller at planting grew more in height (P=0.005). There were, however, no differences in diameter growth (P=0.368). Root and diameter growth in the spring test were comparable to growth in the test period that started in the middle of August, where as height growth was comparable to the test period that started in the beginning of July (APRIL 96) or August (MAY 96) the previous year. 3.1.4 Discussion 3.1 4.1 Timing of growth and dry-mass partitioning In the summer of 1997, more dry mass was allocated to shoots and less to roots compared to 1996 (Figure 3.3h). In 1996, sowing date did not affect the final height, diameter or dry mass of the stem and roots (Figure 3.1). The period of rapid growth started earlier in seed lings sown in April compared to seedlings sown two weeks later. However, at the end of the growing season there were no differences in seedlings in different crops. The weather conditions of the 1996 and 1997 growing seasons differed. July 1996 was cooler and rainy, August 1996 was warmer and dry; and both July and August 1997 were warmer and drier than the long-term average (Figure 2.1). It is probable that these differences in weather conditions also affected the height growth and dry-mass partitioning. Height and diameter growth of the APRIL 96 seedlings slowed down in late July and early August (Figure 3.1a,b), possibly due to slight drought stress. Seed lings from both crops were irrigated and fertilized according to the 46 Control of Growth and Frost Hardening of Silver Birch Container Seedlings same regime. Taller APRIL 96 seedlings were assumed to use more water due to the larger transpiring surface (not measured) compared to the MAY 96 seedlings. Thus the APRIL 96 seedlings may have suf fered slight drought stress, which could have decreased their growth late in the season. In 1997, in particular, considerably less dry mass was accumulated by roots than by shoots (Figure 3.3h). The dry mass of the roots de termined in different years was not totally comparable, due to the fact that different persons washed the roots. It cannot, however, be proved how much of the effect was due to that; but because of the increased height growth, dry-mass allocation to the roots probably was reduced in the summer of 1997 compared to the summer of 1996. The optimal temperature for shoot growth is 21° C, but temperatures between 15 and 21° C do not affect root growth (Skre 1991b). Thus the shoot-to root ratio usually increases with increasing air temperature. During the early part of the season, most of the dry mass was allo cated to leaves (Figure 3.2). Young growing leaves do not export assimilates to other tissues (Kozlowski 1991). Once leaves mature, they produce carbohydrates for growth of new leaves, stem and root tissues (Kozlowski and Pallardy 1997). In plants where height is in creasing, roots are relatively weak sinks and tend to assimilate carbo hydrates only from lower leaves or after the demand of the devel oping leaves is met (Dickson and Isebrands 1991). In this study, the dry mass of the leaves increased until height growth ceased in the middle of August (Figure 3.1). After that, the dry mass of the leaves was unchanged until the beginning or middle of September, depending on the year (Table 3.2). Both short photoperi od and decreasing temperature in autumn are needed for yellowing and abscission of leaves (Thomas and Vince-Prue 1997). After growth ceases, leaves are important for proper frost hardening, since the early stages of frost hardening are dependent on growth regulators produced by the leaves and on other active metabolic processes (Fuchigami et al. 1971). Early and heavy frosts usually retard abscission of leaves by injuring the abscission zone, thereby impeding the normal processes (Kozlowski 1991). During this study, the first autumn frosts, which did not, however, damage the seedlings, were in the middle and end of September, in 1996 and 1997, respectively (see Figure 2.1). In 1997 the seedlings were tall, mature leaves located lower on the stem were shed first, and the shedding started earlier than it had the year before (Figure 3.1 c, Table 3.2). Shading, drought stress and other environmental stresses may also lead to senescence and to shedding of mature leaves, because lower leaves are not able to maintain a favorable carbon balance (Dickson and Isebrands 1991). Due to low levels of light, mature leaves are not capable of positive net photo synthesis and usually cannot use imported sugars (Dickson and Ise brands 1991). 47 Jaana Luoranen The dry mass of the shoot and roots increased for more than a month after cessation of height and diameter growth (Figure 3.3 - e). During this period, the dry mass almost doubled. At the same time, however, growth of new roots decreased linearly toward au tumn (Figure 3.5 a). Actively growing shoots contain mostly water. However, as the shoots lignify and accumulate stored carbohydrates, the water content of the tissues decreases in association with frost hardening (Abod and Webster 1991b). The dry mass of the overwin tering parts is increased by the active processes of leaf abscission (Addicott 1970) and transfer of nutrients and mobile compounds from leaves to stem and roots (Tamm 1951, Dickson and Nelson 1982, Chapin andKedrowsky 1983, Abod and Webster 1991b, Nordell and Karlsson 1995). Abod and Webster (1991b) showed that for silver birch seedlings the roots are important sites for storage of assimi lates. Accordingly, in the present study during leaf yellowing, the increment in dry mass in the roots was probably, at least partly, due to translocation of carbohydrates to winter storage. Dry mass may also increase due to absorption of nutrients. Deciduous, broadleaf species are able to absorb nutrients during autumn and winter, pro vided that the soil temperature is above O°C (van den Driessche 1984). When all leaves had been shed, the dry mass of the stem and roots did not change (Figure 3.3f,g), thus, indicating a decrease in the meta bolic activity of the seedlings (Weiser 1970). 3.1 4.2 Ability of birch-seedling roots to egress during summer In general, the RGP test has been used to determine the ability of seedlings to grow roots in optimal radiation, photoperiod, tempera ture, moisture and nutrient conditions (Ritchie and Tanaka 1990). In an environment that is highly favorable for root growth, in a natural environment RGP usually overestimates root growth (Rikala and Put tonen 1988, Ritchie and Tanaka 1990, Folk and Grossnickle 1997). In this study, the objective of the modificated RGP test, the root egress test, was, in addition to examining the timing of root growth, also to evaluate the seedlings' performance potential at different times of the summer and to compare that with the spring values. Since the experiment was done in a greenhouse, light intensity (not measured) was probably lower and temperatures higher than outside. Differ ences between day and night temperatures were minor compared to conditions in a nursery or in the field. In spite of a few differences in growing conditions, most of the results could be generalized to natu ral growing conditions. Root growth of silver birch seedlings was strong until the middle of August (Figure 3.5 a) and continued as long as the leaves were green. Similar results were shown earlier by Rikala (1996). When 48 Control of Growth and Frost Hardening of Silver Birch Container Seedlings yellowing and abscission of leaves was induced (Figure 3.4), growth of new roots slowed down and finally stopped. Root growth depends largely on current assimilates, which are produced mainly by mature leaves (Eliasson 1968, Burdett 1990). Here root growth was mea sured as dry mass of new roots grown out of the peat plug; thus it was not possible to say whether this was due to root initiation or root elongation. According to Lyr and Hoffmann (1967), both in conifer and deciduous trees uninterrupted root growth during winter might be possible at mild winter temperatures and in frost-free soils, but the growth rate and number of growing roots would decline mark edly compared to the growing season. In spring, the amount of root egress in the seedlings was lower than in summer (Figure 3.5 a). Root growth needs carbohydrates prod uced by leaves. Storage carbohydrates are used mainly during bud flushing for growth of the leaves (Abod et al. 1991). In the early part of the growing season, expanding leaves use assimilates for their own growth and for growth of the upper leaves (Kozlowski and Pal lardy 1997). This means that few carbohydrates are available for the roots, and more rapid root growth is possible only after some leaves mature (Lee and Hackett 1976, Abod and Webster 1991b, Abod et al. 1991). 3.1.5 Conclusion After transplantation, silver birch container seedlings allocated most of their dry mass growth to leaves. Later, more and more of the dry mass was allocated to other parts of the plants. After cessation of height and diameter growth, the dry mass of the shoot and roots in creased until leaf abscission. The results of root-egress tests indi cated that the root growth of first-year birch container seedlings is strongest in July and early August. After height growth ceases, growth of the new roots declines, ending completely when the leaves are shed. In spring, root growth is not possible until new leaves are fully expanded; this means that root growth is weak compared to that of the previous summer. 49 Jaana Luoranen 3.2 Frost hardening and assessment of frost hardiness 3.2.1 Background and study objectives There are several definitions of frost hardiness. Glerum (1985) men tioned, e.g, the lowest temperature below the freezing point to which a seedling can be exposed without being damaged, and the mini mum temperature at which 50 percent of the seedlings are killed (ex pressed as lethal temperature 50 (LTS0 ). On the other hand, frost har diness can be defined as concerning only a plant part, for example, stem, buds or roots. Frost hardiness is widely expressed as LT S0 (e.g. Bittenbender and Howell 1974, Sutinen et ai. 1992, Linden et ai. 1996). However, the lowest temperature or the same kind of estimates are also used to express the frost hardniness of seedlings (Sakai et al. 1986, Lind ström 1992, Anisko and Lindström 1995, DeHayes and Willams 1989). In addition, estimates of LT 10 (Hansen 1992) or LT 20 (Perkins et al. 1993), the temperature at which only 10 or 20 percent of seed lings or tissues are dead or damaged, are used. Ketchie et al. (1972) used the temperature range at which injuries appear (T 10 and T % ). Odium and Blake (1996) recommended the use of the temperature needed to induce an index of injury of 5 percent in the electrolyte leakage test. Using the electrolyte-leakage method, Murray et al. (1989) expressed frost hardiness as the rate of electrolyte leakage. When the sigmoid function is used for calculation, frost hardiness is also described as the rate of freezing injury, or the point of inflection is used as a measure of maximum rate of freeze-thaw stress (Zhu and Liu 1987, Fry et al. 1991, von Fircks and Verwijst 1993). According to Ritchie (1991), the LT 50 value is reliable for testing intact seedlings and determines when the injuries develop complete ly during the incubation time. It might be useful to compare the grow ing methods or species using LTS0 values; but considering the needs of nurserymen, a more suitable value might be, for instance, the lowest temperature at which the evaluated tissue survives. Frost hardiness of woody plants has been assessed by several meth ods. In most tests the viability of a tissue or intact plant is assessed directly or indirectly after freezing. If frost hardiness is defined as the lowest temperature at which seedlings survive, the possibilities to use different methods to estimate this temperature have to be evalu ated. Stergios and Howell (1973) compared several viability tests in four species and concluded that, even though a method gives good results for one species in certain conditions, it is not certain that it will also work on other species or on the same species in other conditions. 50 Control of Growth and Frost Hardening of Silver Birch Container Seedlings However, according to them, in all species tested the most reliable tests were regrowth and tissue browning. The growth test evaluates growth after exposure of the plants to different temperatures. For species with no growth competence during the early part of dorman cy (rest period), this test requires considerable time before regrowth occurs. Although there may be no visible symptoms of injury and viability may not be affected, invisible minor freezing injuries may decrease growth (Ritchie 1991). In the browning test, stems are vi sually scored as dead if the cambium and phloem are brown or yel low, and alive if they are 'apple green' (Ritchie 1991). In this test the labor needed is minimal and the incubation time is only 1-2 weeks (Stergios and Howell 1973, Glerum 1985). The relative electrolyte-leakage (also called relative conductivity or relative ion leakage) method is based on the principle that plasma membranes in freeze-injured tissue allow increased leakage of electro lytes into the apoplastic water (Ritchie 1991). Since the work of Dexter et al. (1930, 1932), the electrolyte-leakage method has been widely used to assess the frost hardiness of different plant tissues. In woody plants this method has been used for both deciduous (e.g. Wilner 1960, 1961, Stregios and Howell 1973, Deans et al. 1995) and coniferous species (e.g. Aronsson and Eliasson 1970, Colombo etal. 1984). Dexter etal. (1930,1932) measured only absolute electro lyte conductivity, but later Wilner (1960, 1961) showed that more reliable results could be obtained when the relative conductivity (ratio of conductivity after freezing to total conductivity after the sample is killed) is measured. Wilner measured the conductivity of frozen samples after 24 hours incubation time and then boiled the samples in water for two minutes to release total electrolytes. Recently, Deans et al. (1995) have questioned the previously used short incubation and killing times for assessing the frost hardiness of woody species, in particular, for assessing the stem of deciduous species. According to them, a reliable method for assessing freezing injury in small pieces of leafless stem tissues required 5-7 days after temperature expo sure for electrolyte equilibration, followed by autoclaving for at least 90 min at 121° C and allowing an additional 24 h before conductivity measurement. Exposure temperature, degree of damage and frost hardiness of the samples all affect the time response of electrolyte leakage (Zhang and Willison 1987): the rate of electrolyte leakage from more dam aged samples caused by lower freezing temperatures is slower than that of less damaged samples exposed to less severe freezing temper atures. On the other hand, the hardier the tissue the slower the rate of electrolytes leakage out of the cells to deionized water. Thus, during the same, but too short, incubation time the measured electrolyte leakage of samples with different degrees of frost hardiness may dif fer, even though the damage is the same. This may cause problems 51 Jaana Luoranen to estimate a proper frost-hardiness parameter. Estimates of frost hardiness have been determined graphically from freezing response curves (Sutinen et ai. 1992). By using statistical methods, it is possible to calculate the standard errors of the estimates and thus compare frost hardiness in different populations. The meth ods used are, for example, the Spearman-Kärber method (Bitten bender and Howell 1974), calculation of critical temperature with analysis of variances and interpolation (DeHayes and Williams 1989), logistic regression models (Zhu and Liu 1987, Anderson et al. 1988, Repo and Lappi 1989, Fry et al. 1991, Sulc et al. 1991, Hansen 1992, Odium and Blake 1996, Linden et al. 1996), the Richards function (von Fircks and Verwijst 1993, Lim et al. 1998) and the Gompertz function (Lim et al. 1998). Logistic models, the Richards function and the Gompertz function are sigmoid functions with a maximum of four parameters. Logistic models are symmetrical, and the others are asymmetrical to the point of inflection of the curve (Ratkowsky 1990). According to Ratkowsky (1990, pages 140-141) the Richards function, although widely used, is statistically poor at estimating the parameters, due to its strong non-linearity with respect to parame ters. Logistic models have been widely used with good results (e.g. by Andersson et al. 1988, Repo and Lappi 1989, Odium and Blake 1996, Repo et al. 1997, Ryyppö et al. 1998b) and were also used here. The objective of this part of the study was to evaluate tests of electro lyte leakage, stem browning and viability of whole seedlings for asses sing the frost hardiness of shoots of silver birch container seedlings in autumn. More specifically, the following objectives were included: 1) to compare the frost hardiness estimates given by the three meth ods; electrolyte-leakage and stem-browning tests were compared to the whole-seedling viability test, which was used as the method giving the most reliable results, 2) to evaluate the estimates for 10 and 50 percent lethal (damaging) temperature in practical nursery use, and 3) to monitor the frost hardening of birch container seedlings raised according to normal nursery practices in Finland. 52 Control of Growth and Frost Hardening of Silver Birch Container Seedlings 3.2.2 Material and methods 3.2.2.1 Freezing tests Frost hardiness (see Section 3.1.2, crop sown in May 97) was tested five times in autumn: on 18-19 August, on 8-9 and 29-30 Septem ber, and on 13-14 and 27-28 October 1997. For each testing date, seven trays were randomly selected, so that 2 or 4 seedlings from each tray were exposed to each test temperature (a total of 20 (on the last date 10) seedlings per test temperature). Half of the seedlings were put on polystyrene trays, in which roots were protected by cov ering them with sawdust. The rest of the seedlings were cut into two parts: the water content was determined from the uppermost 10 cm (results shown only in Figures 6.4 and 6.5), the rest of the stems were exposed to freezing temperatures. In late October the air tem perature decreased below O°C, peat plugs for the survival test were frozen and it was impossible to lift seedlings from trays without dam aging stems and roots. Thus, only samples for the electrolyte-leakage and stem-browning tests were exposed. The samples were exposed to six test temperatures (presented in Table 3.3) in air-cooled chambers. Air temperature in the chambers was controlled by an external alcohol-circulating system [Lauda RUK9O Ultra-Kryomat together with a Lauda digital programmer R4lO and PM3SI (MGM Lauda, Germany)]. On the first three test dates, the temperatures in the freezers were lowered 3°C per hour, kept at the minimum for 3to 11 hours depending on test temperatu re and test date, then raised 3°C per hour to +5° C. On the fourth and fifth dates, the temperatures in the freezers were lowered 3°C per hour until -15° C was reached, lowered 5°C per hour and kept at the minimum for 3 to 21 hours, raised 6°C per hour to -15° C, and final ly 3°C per hour to +5° C. Air temperatures in the chambers and tem peratures of the stem samples or stems of whole-exposed seedlings were recorded with thermocouples. On all testing dates the control temperature was +4° C. Table 3.3 Exposure temperatures at each test date in autumn 1997. Test date +4 -2 -4 -6 -8 -10 -12 Exposure temperature, °C -14 -16 -18 -20 -22 ... -30 -32 -34 -36 ... -40 18 Aug X X X X X X X 8 Sep X X X X X X X 29 Sep X X X X X X X 13 Oct X X X X X X X 27 Oct X X X X X X X 53 Jaana Luoranen Frost hardiness was assessed by three methods: 1) visual scoring of damage on stem samples, 2) electrolyte-leakage, and 3) viability of whole seedlings. Visual scoring of stem damage After freezing, when the pieces were cut from the lower and upper parts of the stem for testing of electrolyte leakage, the rest of the stem samples were stuck onto wet floral foam for the stem-brow ning test. The cuttings for the browning test were then placed in a growth chamber [Weiss, type 10 Sp/5 DU-Pi, Lidenstruth, Germa ny; temperature: +2O/15°C, 16 h photoperiod, photon flux density (PAR) 350 nmol-irr2^ 1 from metal halide lamps (Powerstar HQIT -400/D, Osram)], where they were kept wet by adding water to plas tic boxes with foam as needed and spraying the stems daily with tap water. According to the results of a preliminary study, 14 days in these conditions were necessary for injury to become visible. Thus, after 14 days each cutting was dissected and the length of the dam aged stem measured. The stems were scored as damaged if the cam bium and phloem were brown, and alive if they were still 'apple' green (Ritchie 1991). Stem damage was calculated as relative brown ing ( RB ) where Lb is the total length of the brown tissue and L( is the total length of the sample. Electrolyte leakage After freezing, two 0.5 cm long pieces were cut from the lower and upper parts of the stem for testing of electrolyte leakage. The pieces were washed in deionized water and put into a test tube containing 18 ml of deionized water. On the first and second dates the tubes were shaken (110 revs min-1) for 22 hat room temperature (+2O - (Colombo et al. 1984) and on the other dates shaken for 120 h at +4° C (minimizing the microbial activity) in darkness (Deans et al. 1995). To minimize the risk for biased estimation of frost hardiness in samples with different degrees of hardiness, the incubation time for the last three testing dates used here was five days. It was as sumed that in the first and second test the seedlings had such low hardiness and the temperatures used were so high that there was no risk of slow electrolyte leakage (see Zhang and Willison 1987) and the incubation time was 22 h, as suggested by Colombo et al. (1984). RB = (3.1) L, 54 Control of Growth and Frost Hardening of Silver Birch Container Seedlings After the incubation, the conductivity of the solution was mea sured (Radiometer CDM 83, water bath at 25° C ±l°C temperature, cell constant 1.008, |iS cm" 1 , without temperature compensation). According to the results of the preliminary experiment, to minimize overboiling, 15 ml of water was removed from the test tubes before the sample pieces were killed by autoclaving the tubes at 120° C for 60 minutes. The removed water was returned to the tubes and the total conductivity was measured after the 22-hour shaking period at room temperature in darkness. Relative electrolyte leakage (REL) was calculated as where e, is the result of the first conductivity measurement, which expresses the electrolyte leakage caused by freezing injuries and background leakage, and e 2 is the second conductivity measured af ter autoclave killing, which represents the total electrolyte leakage of the stem tissues. Viability of whole seedlings After exposure to freezing, seedlings on polystyrene trays were kept in a dark cold room at +5° C until the peat plugs had thawed (24—48 hours, depending on test temperature). Then the seedlings were placed in an unheated greenhouse, where they were irrigated as needed. On 14 October, the seedlings were removed from the trays, packed in plastic bags and stored at -2.5±0.5°C. From 22 January to 2 Februa ry, the seedlings were thawed in darkness (11 d at +2° C) and from 2 to 3 February at +l7°C. On 3 February, the seedlings were put in Plantek 25 trays and transferred to a heated and lighted greenhouse (day/night temperature: +2O - +22 °C/ +l5°C, photon flux density, PAR 150 |jmol m 2 -s"' from metal halide lamps (HQI-400W Power Star, Osram), 18 h photoperiod) for 17 days. All seedlings were wa tered daily with tap water. On 20 February, seedlings were classified into four classes: 1) alive (undamaged), 2) alive, but smaller leaves than in class 1, 3) alive but stem apices dead and 4) dead. For statis tical analysis, seedlings were reclassified as a) damaged (class 2-4) and undamaged, for calculations of proportions of damaged seed lings (D) (used to estimate damaging temperatures (DT), or b) dead (class 4) and alive (class 1-3), for calculations of proportions of dead seedlings (L), which was used to estimate lethal temperatures (LT). REL = (3.2) «2 55 Jaana Luoranen 3.2.2.2 Estimation of frost hardiness Both in electrolyte-leakage and in stem-browning tests, the propor tions of injuries [relative electrolyte leakage (REL) or browning (RB)] and in the test of viability of whole exposed seedlings, the damage (D) or mortality (L) of seedlings increased with decreasing tempera ture to a certain temperature as a logistic function (Figure 3.6). The same logistic function as has been used previously, e.g., by Anders son et al. (1988), Repo and Lappi (1989), Odium and Blake (1996), and Repo et al. (1997), was used where where y. is the measured injury (REL, RB, the proportion of injured or dead seedlings), x. is exposure temperature (°C) and a, b, c and d are equation parameters. The inflection point c is the temperature (°C) at which the change of injuries (dyldx) is maximal as tempera ture decreases (Zhu and Liu 1987). The upper and lower asymptotes of the function are determined by parameters a and d, respectively. Parameter d is the lower asymptote of function, the background level, e.g. indicating the REL from uninjured cells at non-damaging tem peratures. Parameter a expresses the injury range between the lower and upper asymptotes. Thus, a + d indicates the upper asymptote, which tells the REL value for the temperature where injuries are at maximum. The slope of the curve at the point of inflection is ab/A. Thus, parameter b [(C°) -1 ] measures the slope at c relative to injury range a. Parameters a and d were not estimated in tests of stem-browning and whole seedling viability. It was assumed that in the stem-brown ing test, living samples have no injuries (d =0) and dead ones have brown phloem and cambium (a + d= 1). In the viability test all seed lings are assumed to be alive when the lower limit d was 0 and dead (damaged) when the upper a + d was 1. The variance of the distribution in the tests of stem browning and whole seedling viability (Figure 3.6f-r) is proportional to the vari ance of the binomial distribution (p( 1 - p)). Thus the variance is great est when p = '/ and is 0 when p=oor p = 1. In nonlinear estimation, however, homogenous error variances are assumed. To homogenize these variances, both sides of the equation (3.3) were divided by the weight (w) where / was the estimate of/(x) given to the first iteration. To avoid dividing by zero, 0.01 was added to each value. ?; = /(■*,) + £ (3.3) = l+e «'-,) + '' +e w-V/Ö-7)+001 <3-41 56 Control of Growth and Frost Hardening of Silver Birch Container Seedlings Figure 3.6 Results of freezing tests in autumn 1997. Methods for estimation of frost hardiness were a-e) electrolyte leakage, f-j) stem-browning test (relative proportion of damaged stem) and whole-seedling viability tests assessed either as proportions of k-n) damaged or o-r) dead silver birch seedlings. When whole seedlings were tested, the seedlings were classified as damaged when they were dead or their leaves were smaller than in living seedlings. For estimation of lethal temperature, seedlings with small leaves were scored as alive. The vertical bars indicate the standard error of the mean values for each method at each test temperature (n =lO samples per temperature). Lines are the fitted sigmoid functions (estimated param eters are presented in Appendix I). Injuries at temperature x p are the proportion p of the range of electrolyte leakage (or damaged stem) (expressed as f - pa + d. The temperature x p can be estimated as described earlier in the case of estimation of growth cessation (see section 2.6.2). The equations for estimations are or y,= 77— + d + e (3.5) -In 1 -(e-*,) l+ e x "~ c Jaana Luoranen 57 Equations (3.3) and (3.5) or (3.6) with a value of 0.1 for p were fitted for data from each testing date and each method by using non linear regression with the Quasi-Newton method. The logistic equa tion used is symmetrical (Ratkowsky 1990), and the estimated point of inflection c was used to express the temperature at which the REL is 50 percent of the observed range of REL (ET 50 ), half of the stem is brown (BT50), and half of the seedlings are damaged (DT50) or dead (LT 50 ) (Zhu and Liu 1987, Andersson et ai. 1988, Repo and Lappi 1989, Fry et ai. 1991, Von Fircks and Vervijst 1993, Odium and Bla ke 1996). The estimated parameter x p expresses 10 percent of the range of electrolyte leakage (ET |0 ), stem browning (BT ]0 ), damaged seedlings (DT10) or mortality (LT |0). Frost hardiness estimates given by electrolyte-leakage and stem-browning tests were compared to estimates assessed by whole-seedling viability; relationships between estimates were determined by fitting linear regressions. Figure 3.7 Changes in electrolyte leakage from stems of silver birch seed lings after freezing tests during autumn. Mean, maximum and minimum electrolyte leakage (mea sured as conductivity of the incubated solution) calcu lated from all measured conductivities after freezing (broken lines) and killing by autoclaving (solid line) on each testing date in autumn 1997. 3.2.3 Results In early autumn (test on 18 August) the tissues were damaged at -4 to -5 °C, and the response curves were steep (Figure 3.6 a, f, k, o). Later the slope of the curve (b ) decreased with increasing frost har diness (Figure 3.6). Frost hardening decreased the total amount of leaked electrolytes (Figure 3.7); and the differences between electro lyte leakage after freezing and killing decreased, making curve fit ting more difficult or impossible (on the last date only BT could be estimated). a , yt = 7 +d + £ in i-1 (3.6) b X„ + X; l+ e ' 58 Control of Growth and Frost Hardening of Silver Birch Container Seedlings Figure 3.8 Position of damaged cambium and phloem in samples used in the stem-browning test on each testing date. Proportion of the length of the brown base, middle or top part of the total length of the sample at each test temperature is shown. The black part of the bar indicates that the cambium and phloem were damaged in the whole sample. In the stem-browning test, the position of the injuries on the stem was also observed. In most cases the entire stem was either damaged or alive. However, when cuttings were only partially damaged, the lower part of the stem was damaged more often than the apices or the middle part of the stem (Figure 3.8). Frost hardiness of silver birch seedlings increased slowly from the middle of August to the middle of September. After that, frost harden ing became more rapid. The differences between the estimated BT50 and ET J0 or BT |0 and ET 10 were minor (Appendix I). There were strong linear relationships between estimates, but BT S0 and ET 50 un derestimated the viability of the stem compared to estimates of LT 50 or DT 50 (Figure 3.9 b, d). BT 50 and ET 50 estimated, however, quite well DT 10 (Figure 3.9 c). The differences between estimates given by different tests increased with frost hardening. Jaana Luoranen 59 Figure 3.9 Correlations of frost-hardiness estimates by different assessment meth ods. The X-axis presents estimates for a) 10 and b) 50 percent lethal tempera ture (LT) or c) 10 and d) 50 percent damaging temper ature (DT) assessed by who le seedling viability tests. In the Y-axis are the frost har diness estimates from elec trolyte-leakage (ET) or stem browning (BT) tests. 3.2.4 Discussion In this study all samples in a test were subjected to the same treat ment conditions: the same chamber for samples at a given exposure temperature, the same shaker, all test tubes in the same water bath during the conductivity measurements etc. Thus, samples were pseu doreplicates. This might cause bias to the estimates of frost hardiness and its error variances. According to Repo and Lappi (1989), this does not cause bias in the estimated parameters of the response func tion, but their variances increase and are underestimated. There were only three chambers and seedlings for each test and the number of test temperatures was limited. Thus, in order to decide on exposure temperatures, before freezing some assumptions were made about the frost hardiness of seedlings. In the second and third test the frost hardiness of the seedlings was assumed to be higher and the last time lower than it actually was. In all tests, there should have been enough temperatures so that both the upper (all samples dead, maximal REL) and lower limits (all samples alive, background level in electrolyte leakage) would have been reached (Anisko and Lindström 1995). Because this was not the case in the last test, fit ting of reliable curves was difficult (browning test) (Figure 3.6j) or 60 Control of Growth and Frost Hardening of Silver Birch Container Seedlings impossible (electrolyte leakage) (Figure 3.6e). On the other hand, test temperatures should be so close each other that there are also values between asymptotes (partially damaged, only some seedlings dead) (Raymond et al. 1986). On the third date, there were too few temperatures between -5 and -10° C for reliable estimation of frost hardiness, especially by the electrolyte-leakage method (Figure 3.6 c, h, m, q, Appendix I). In early autumn, when the frost hardiness of seedlings was low, seedlings were damaged in a narrow temperature range and the re sponse curve was steep. With frost hardening, the response curves became less steep due to decreased electrolyte leakage (Figure 3.7). The reduced leakage has been explained by cellular differentiation, lignification and suberization of cell walls (Sutinen et al. 1992, Repo et al. 1997) and decreased amount of free water in tissues (Levitt 1980). At the end of October, electrolyte leakage was so slight that it was not possible to fit the sigmoid curve (Figure 3.6e). Decreased leakage could not be caused by undamaged samples, since in the browning test most stem cuttings were scored as dead at the same exposure temperatures (Figure 3.6j). However, also in the stem-brow ning test, parameter estimation would have been more reliable if test temperatures had been lower. Difficulties in estimating frost hardiness by the electrolyte-leakage test during late autumn and winter have also been observed by Sutinen et al. (1992) using pine needles. Ac cording to them, low electrolyte leakage from hardy samples might be due to changes in the properties of the cell wall, which provide resistance to diffusion of electrolytes from cells into the extracellu lar water. Thus, the only suitable method used in this study for hardy seedlings was the stem-browning test. According to the definition of Steponkus and Lanphear (1967), an ideal method of testing for frost hardiness "should eliminate bias associated with visual observations, be based on a quantitative sys tem that could be analyzed statistically, utilize small quantities of tissue, be relatively quick, and be capable of predicting the future performance of the plant." The advantages and disadvantages of each method used for frost hardiness estimation are presented in Table 3.4. The electrolyte-leakage method fulfills all requirements for ideal methods, and the stem-browning test fulfills all except the require ment for elimination of visual observation. Observing the viability of whole seedlings does not fulfill any of Steponkus' and Lanphear's (1967) requirements except that it predicts the field performance of seedlings well. In this study, visual observation of viability of whole seedlings was used as the method that gives an estimate of the real frost hardiness of seedlings, as suggested by Stergios and Howell (1973) and Ritchie (1991); the frost hardiness estimates of the electro lyte-leakage and stem-browning tests were therefore compared to the results of whole-seedling viability test. 61 Jaana Luoranen Both ET 50 and ET 10 were higher than BT S0 and BT |0, but the differ ences were not statistically significant (Figure 3.9). The estimates of ET 50 and ET10 might have been more appropriate if the samples ex posed to freezing had been incubated in light in heated conditions some time before the pieces were cut into tubes for determination of electrolyte leakage. Paita et al. (1977) suggested that to give enough time for cell membranes to recover, measurements of conductivity should be made 1-2 weeks after exposure to freezing. In the present study, during five days of incubation the injuries to cell membranes probably recovered in water solution, but how much they recovered is uncertain. Thus it is possible that estimates of frost hardiness by both electrolyte leakage and stem-browing test could have been the same after two weeks incubation time before the electrolyte-leakage test. The electrolyte-leakage test identifies injury to tissues indirectly by measuring damage to cell membranes: how many electrolytes (mainly K + ) leak through the membranes or transport proteins em bedded into plasma membranes to the apoplast and then into deion ized water (Palta et al. 1977, Pukacki and Pukacka 1987). Therefore Table 3.4 Advantages and disadvantages of electrolyte-leakage, stem-browning and whole-seedling sur vival tests for assessment of damage after freezing in silver birch seedlings, partially using the definition of an ideal method by Steponkus and Lanphear (1967). Advantages Disadvantages Electrolyte leakage - relatively rapid (a week) compared - laborious (2 measurements, several steps) to other tests - only 0.5 cm stem samples needed - skills needed - could be analyzed statistically - less suitable for hardy samples - large number of samples per test - relatively expensive equipment needed - assessing early stages of injuries - hardiness differences between different parts of seedlings Stem browning - relatively rapid (two weeks) - visual observation - not laborious (only one relatively - difficulties in scoring damaged and rapid measurement) living tissues - quantitative when injuries are measured - hardiness differences between different parts as a length of brown tissues of seedlings, when using only a part of stem* - suitable for hardy samples - small samples are sufficient (see *) - could be analyzed statistically Whole-seedlings - accurately predicts the mortality - not practical for large number of seedlings survival (chronic injuries) of whole seedlings (space need in chamber, in storage or in greenhouse during incubation) - could be analyzed statistically - weak separation of living/dead/damaged seedlings - not laborious (visual scoring once. - space needed in chamber (height of storage needed) the seedlings) - time required before observation (rest and chilling period, months) - elimination of other stresses after exposure during storage, before evaluation 62 Control of Growth and Frost Hardening of Silver Birch Container Seedlings it is the most sensitive method for observing the minor injuries caused by freezing (Odium and Blake 1996). Some of the injuries to cell membranes might be repaired and some be so minor that they do not kill the seedlings (Paita et al. 1977). A disadvantage of the electrolyte leakage test is that it measures injuries in the whole sample. How ever, cambium is the most important tissue for survival of the plant; and even if it is dead, the other parts of the sample might be alive and decreasing the average leakage (Stergios and Howell 1973). The stem-browning test measures damage in the cambium and ph loem. It is also important to observe the position of injury in the stem tissues: cambial death, when it occurs in the terminal leader, is not lethal to the whole plant, whereas the same damage would be lethal if it occurred at ground level (Ritchie 1991). When damage was observed in dissected stems at the beginning of frost hardening, the most tender part of the stem was at ground level. However, the whole stem was usually damaged by freezing (Figure 3.8). These results in silver birch seedlings agree with observations made in coni fer seedlings that seedlings harden from the top down (Ritchie 1991, Colombo et al. 1995). A disadvantage of the stem-browning test is that it is subjective. Separation of colors in living (green) and dead (brown or yellow) tissues is not always easy. The green color is due to chlorophyll in the stem tissues, and a marked decline in chlorophyll levels is ob served from the growing season towards winter (Kauppi 1991). This means that in late autumn the green color in the stem is not as bright as it is during summer and early autumn. Observations made with a microscope aid the assessment of injuries (Ritchie 1991), allowing the observer to view the breakdown of cells in addition to discolora tion (Lindström 1992), but at same time the method would become more laborious. The electrolyte-leakage and stem-browning tests underestimated frost hardiness compared to the survival test (Figure 3.9). The under estimation of the electrolyte-leakage test was also observed by Zhang and Willison (1987), Sutinen et al. (1992) and Odium and Blake (1996). Although the stem-browning test describes the survival of stem tissues, some of the stem damage was moderate and/or not lethal and did not affect the survival of seedlings. If the seedlings exposed to freezing had been planted in the field and the viability observed after one season, those seedlings classified as damaged would pro bably have been dead and the estimates of lethal temperature and estimates of 50 % electrolyte leakage or stem browning more simi lar. According to Ryyppö et al. (1998b), survival of Scots pine seed lings was much lower at the end of the first season after exposure to freezing than prior to bud burst. It was assumed that the ET 50 and BT 50 estimates of the electrolyte leakage and the stem-browning test would estimate the LT 50 values 63 Jaana Luoranen and thus would not give useful prediction of frost hardiness for nur sery managers, so estimates of ET [0 and BT10 were calculated. ETJ0 or BT50, however, predicted the temperatures causing damage of 10 % of seedlings quite well (Figure 3.9 c). Thus by estimating the tem perature causing 50 % electrolyte leakage in the observed range of relative electrolyte leakage or the temperature at which half of the stem tissues are damaged, it would be possible to predict the tempera ture at which only 10 % of seedlings are damaged. In the whole seedling viability test, using the damaging temperature instead of the lethal temperature as an estimate of the viability of seedlings will give more accurate estimates of the damage due to freezing be cause frost affects not only mortality of seedlings but also suppresses growth (Raymond et al. 1986). The method used for estimation of frost hardiness depends on the objective of the study. Odium and Blake (1996) concluded that elec trolyte leakage is a suitable method for assessing acute damage, whereas the whole-seedling method may be more appropriate for detecting chronic damage. If, for example, hardening methods are compared, the electrolyte-leakage or stem-browning tests, which give faster results, are suitable. When the aim is to evaluate the real sur vival of seedlings after hardening, assessment of the viability of en tire seedlings gives more reliable and appropriate results. None of the methods used in this study give estimates of frost har diness just after exposure to freezing. Reliable estimation of frost hardiness in deciduous seedlings requires at least a week (electrolyte leakage). Results of tests for making decisions about nursery opera tions are needed much more rapidly, and therefore other methods should be evaluated. Other possible methods are, for example, imped ance (Evert and Weiser 1971, Blazich et al 1974, Timmis and Fuchi gami 1982, Glerum 1985, Jozefek 1989, Repo 1994, Repo et al. 1997) and chlorophyll fluorescence (Binder et al. 1996, Ögren 1999 a). In the fertilizer experiment of Jozefek (1989), high pre-freezing im pedance readings indicated frost hardy tissue and low readings indi cated the opposite. In deciduous species, chlorophyll fluorescence is possible as long as there is chlorophyll in stem tissues (Binder et al. 1996). For example, Larcher and Nagale (1992) determined the photo synthetic activity of the stem tissues of beech (Fagus sylvatica L.) in winter using chlorophyll fluorescence. Ögren (1999 a) also used this method in willow, with good results. Both methods might be useful if frost hardiness could be assessed without exposure to freezing. Promising results have also been obtained with impedance spectro scopy in willow seedlings (Repo et al. 1997). 64 Control of Growth and Frost Hardening of Silver Birch Container Seedlings 3.2.5 Conclusion Estimates of temperature-induced 50 % electrolyte leakage or injury to 50 % of dissected stem tissues would be useful for comparison of the effects of frost-hardening regimes on frost hardiness of silver birch seedlings. With these estimates, it is possible to predict the temperatures that induce 10 % of the seedlings to damage. The electrolyte-leakage test is more laborious, but it is quantitative and requires only small quantities of tissue. It is suitable for assessing frost hardiness in early autumn. When seedlings are hardy, the only suitable method (used in this study) for reliable assessment of frost hardiness is the stem-browning test. 65 Jaana Luoranen 4 Artificial control of growth and frost hardening of silver birch seedlings 4.1 Growth retardants 4.1.1 Background and study objectives Growth retardants have been used to control stem growth (Gianfag na 1987) and studied for possible use in enhancing frost hardening, but with contradictory results (Carter and Brenner 1985). When Apha lo et al. (1997) studied the use of CCC (chlormequat chloride) and paclobutrazol in the production of silver birch container seedlings, height growth and accumulation of dry mass were partially inhibited after CCC application, but there were no negative effects on growth the following spring. Height growth in the nursery decreased after paclobutrazol application, but the growth of seedlings was also inhib ited the following season. Aphalo et al. (1997) concluded that CCC could be a useful tool in nursery management, especially for control ling end-of-season growth during warm autumns. However, they did not study the influences of CCC on frost hardening or the long-term effects after planting. Other compounds have also been used to control both height growth and frost hardening of birch seedlings. Cathey (1975) compared five growth retardants (ACPC, phosfon D, CCC, daminozide and ancy midol) in several plant species. According to his results, CCC retarded growth in most plants but not in paper birch ( Betula papyrifera Marsh.). Application of daminozide (B-Nine), however, also retarded the height growth of paper birch. An advantage of daminozide, com pared to CCC, was that it was nontoxic at all concentrations, while CCC induced temporary paling of the foliage and marginal burn (Cat hey 1975). There are also newer compounds, such as tetcyclasis and prohexa dione calcium, which inhibit GA biosynthesis at later stages than CCC and ancymidol (Arteca 1996). Tetcyclasis acts like the triazo les (e.g. paclobutrazol, uniconazole), but is more effective than pac lobutrazol in retarding height growth in one-year-old silver birches (Abod and Webster 1991 a). Thus tetcyclasis might also inhibit the growth of seedlings the following year as paclobutrazol did in the experiment of Aphalo et al. (1997). The inhibition of GA biosynthesis 66 Control of Growth and Frost Hardening of Silver Birch Container Seedlings by prohexadione calcium is quite similar to the action of SDs (Junttila 1993). Thus, the most potential retarding compounds are CCC and dami nozide. There have been only a few reports on the effects of CCC and daminozide in controlling height growth of silver birch seed lings in nursery conditions (Aphalo et al. 1997) and, to my knowledge, none on control of frost hardening. The objective of this part of the study was to compare the effective ness of CCC and daminozide applications for retardation of height growth of silver birch seedlings. A further objective was to identify the concentration of these chemicals that would be effective in con trolling height growth and hastening frost hardening of silver birch container seedlings, but without negative effects on vitality or after effects on field growth. 4.1.2 Material and methods Silver birch seeds were sown on peat-filled germination trays in a greenhouse on 2 May 1995. Germinants were transplanted on 22- 26 May to Plantek 25 trays (25 cavities per tray, 380 cm 3 per cavity, 156 cavities per m 2, Lännen Plant Systems, Finland) filled with fer tilized sphagnum peat (Vapo, XL, Finland). Seedlings were irrigated according to normal nursery practice by keeping the water content of peat at 30-60 percent by volume during the growing season. Seed lings were fertilized seven times with liquid fertilizer. The total amounts of nutrients (including basic and liquid fertilizer) given were 126 mg N, 56 mg P and 156 mg K per seedling plus micronutrients. Seedlings were grown in an unheated greenhouse until the plastic cover was removed on 20 June. Daminozide (N-dimethylaminosuccinamic acid, trade name B- Nine, Alar 85, Uniroyal) was supplied in powder form with an active ingredient of 85.0 % ww _1 . CCC (chlormequat chloride, 2-chloro ethyl-N,N,N-trimethyl-ammonium chloride, trade name Cycocel, Kemira Oy, Helsinki, Finland) was applied in an aqueous suspension at a concentration of 750 g H. Seedlings were sprayed once (on 28 June) with daminozide dissolved in water at concentrations of 1.0,2.0, 3.0,4.0,5.0 or 6.0 g pure daminozide per liter (the concentrations used were decided according to the results of Cathey (1975) and instruc tions for flowers) and twice (on 29 June and 27 July) with CCC dis solved in water at concentrations of 0.5,1.0,1.5,2.0,2.5,3.0 g of pure chlormequat chloride per liter (for CCC the concentrations were the same as there used by Aphalo et al. 1997). Control seedlings were sprayed with water. Each tray was sprayed inside a non-ventilated plastic-film hood for 20 s, which gave an applied amount of 200 ml per tray per date and about 8 ml per seedling per date. 67 Jaana Luoranen Thirteen treatments, 30 trays with 25 seedlings in each treatment, were divided into six complete blocks, five trays of each treatment in each block, giving a total of 390 trays and 9750 seedlings. Height of the stem of the same randomly selected two seedlings per tray in all 30 trays in each treatment were measured weekly from 26 June to 18 September. Cessation of height growth was defined as the date when 90 percent of the seedlings had stopped growing. Water content was determined weekly from the uppermost five centimeters of 10 ran domly selected seedlings per treatment from 27 June to 4 October. In the autumn, after leaf abscission, 30 randomly selected seed lings per treatment were harvested, the height and diameter were mea sured, dry mass of shoots (stems and branches) was determined and amounts of nutrients were analyzed. Determination of root dry mass failed because of difficulties to separate the roots from the peat me dium. For the nutrient analysis, stems were pooled by blocks into six samples per treatment. Nutrient concentrations in each treatment are presented in Appendix 11. On 11 October, 24 seedlings per treatment were randomly selected for tests of root growth potential (RGP); and 60 seedlings per treat ment were selected for a planting experiment. Seedlings were packed in plastic bags and stored at -4° C until 19 February when the RGP seedlings were thawed in darkness for 4 d at +5° C followed by 4 d at + 10° C. After thawing, the potted seedlings were divided into four blocks, six seedlings per treatment in each block in a heated green house (day/night temperature: +2o—(-22°C/+l5°C, with PAR 150 pmol nr2 s_1 from metal halide lamps (HQI-400W Power Star, Os ram), photoperiod: 18 h) for four weeks (until 27 March). In order to determine the timing of bud burst and leaf elongation, the length of the same three buds and after bud burst the leaves of each seedling were measured weekly. On 22 May, 60 seedlings from each treatment were planted in the field in a randomized block design (six blocks, 10 seedlings per treat ment per block). At planting and at the end of the first, second and third growing seasons the height was measured. Root collar diame ter was measured at planting and at the end of the first and second growing seasons. Statistical analysis. Height growth and water content were ana lyzed using a multivariate general linear hypothesis (MGLH) pro cedure for repeated measurements. Before analysis of variance, the diameter-to-height ratio of daminozide-sprayed seedlings in the autumn test and new roots in the RGP test were log-transformed to obtain normal distributions. Results shown in the figures and tables are back-transformed. After the analysis of variance, differences be tween applications were analyzed using Tukey's HSD multiple com parisons. ANOVA tables are shown in Appendix V. 68 Control of Growth and Frost Hardening of Silver Birch Container Seedlings 4.1.3 Results CCC induced chlorosis to the leaf edges at all doses sprayed except 0.5 gI 1 level, but the damage was later reversed. Later in the sum mer the leaves on both daminozide- and CCC-sprayed seedlings were darker green than the leaves on the control seedlings. Retardation of height growth by both compounds appeared two weeks after the appli cations (on 10 July) and continued for about four weeks (Figure 4.1). Neither daminozide nor CCC affected cessation of height growth (all seedlings stopped growing by 20 August). In the middle of Au gust, however, only 30-35 percent of the CCC-sprayed seedlings had stopped height growth. At the same time, only percent of the control seedlings and seedlings sprayed with daminozide con centrations of 1.0-3.0 g I -1 and 57-62 percent of the higher dami nozide concentrations had stopped their height growth. Daminozide decreased the final height (P