Academic Press is an imprint of Elsevier
30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 525 B Street, Suite 1900, San Diego, California 92101-4495, USA 84 Theobald’s Road, London WC1X 8RR, UK
Copyright # 2010 Markus Keller, Published by Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.
Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone: (þ44) 1865 843830, fax: (þ44) 1865 853333, E-mail: [email protected]. You may also complete your request online via the Elsevier homepage (http://elsevier.com), by selecting “Support & Contact” then “Copyright and Permission” and then “Obtaining Permissions.”
Library of Congress Cataloging-in-Publication Data
Keller, Markus. The science of grapevines : anatomy and physiology/Markus Keller. p. cm. Includes bibliographical references and index. ISBN 978-0-12-374881-2 (hard cover : alk. paper) 1. Grapes–Anatomy. 2. Grapes–Physiology. I. Title. QK495.V55K44 2010 634.8–dc22 2009033159
British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.
ISBN: 978-0-12-374881-2
For information on all Academic Press publications visit our Web site at www.elsevierdirect.com
Printed in China 091011987654321 Preface
Grapes were among the first fruit species to stomatal pores may be an exception) are rooted be domesticated and today are the world’s in biochemistry. They are driven or at least most economically important fruit crop. facilitated by enzymes, which in turn are built According to 2009 statistics by the Food and based on blueprints provided by genes. I have Agriculture Organization of the United therefore taken it for granted that it is Nations, grapevines were planted on almost understood that a developmental process 7.3 million hectares producing more than 67 or change in chemical composition implies a million metric tons of fruit in 2007. This makes change in enzyme activity, which implies grapes the number 25 food crop in terms of a change in the activity of one or more genes. planted area and number 16 in terms of ton- This does not imply, as used to be thought, that nage. More than 70% of this crop was used to “one gene makes one enzyme,” nor that “one make wine, 27% consumed as fresh fruit (table enzyme makes one chemical,” but it merely grapes), 2% as dried fruit (raisins), and less means that all enzymatic processes have a than 1% was processed to grape juice or dis- genetic basis. tilled to brandy. Many biochemical and biophysical processes This book is an introduction to the physical apply to many or even all plants. Perhaps no structure of the grapevine, its various organs process is truly unique to grapevines. Chances and tissues, their functions, their interactions are that if grapes do it, some or many other with one another, and their responses to the species employ the same solution to a survival environment. It focuses essentially on the phys- issue because they share a common ancestor ical and biological functions of whole plants that invented the trick a long time ago. For rather than the metabolism and molecular example, microbes hit upon photosynthesis biology of individual cells. It is nonetheless and respiration long before these discoveries necessary to review some fundamental enabled some of them to combine forces and processes at the cell, tissue, and organ levels evolve into plants. Consequently, although this in order to build up an appreciation of whole- book is about grapevines, and primarily about plant function. The book covers those elements the wine grape species Vitis vinifera, I have bor- of physiology that will enhance our under- rowed heavily from research done with other standing of grapevine function and how they plant species, both wild and cultivated, peren- relate to grape production. Although of neces- nial and annual, woody and herbaceous, sity the text contains a plethora of technical including the “queen of weeds”—at least in terms and details, I have tried to resist the the fast-paced world of modern molecular biol- temptation to dwell in biochemical and molec- ogy—Arabidopsis thaliana, the otherwise incon- ular biological jargon. Most physiological spicuous thale cress. I have even taken the processes (water movement through the vine’s liberty to borrow insights gained using micro- hydraulic system and evaporation from the organisms, such as the yeast Saccharomyces
ix x PREFACE cerevisiae that gives us wine and beer and The magnitude of the task of reviewing as bread, that enable us to think about these much of the pertinent literature as possible issues. often forced me to rely on review papers, This book aims to be global in scale. It covers where they were available. I apologize to those physiological aspects of tropical viticulture all friends and colleagues whose work I did not the way to those that pertain to the production cite or cited incompletely or incorrectly. of ice wine at the temperate northern margins Science—and scientists—can only ever hope to of grape growing. It moves from vineyards at approximate the truth. This and the simple fact sea level to vineyards at high altitude. It con- of “errare humanum est” will guarantee a number siders the humid conditions of cool, marine of errors throughout the text. These are entirely climates, the moist winters and dry summers my responsibility, and I would be grateful for of Mediterranean climates, as well as the arid any feedback that might help improve this book environment typical of continental climates in and further our understanding of the world’s the rain shadow of massive mountain ranges. most important and arguably most malleable Yet a book of this nature is necessarily incom- fruit crop. After all, the full quote from Seneca plete, and so is the selection of published infor- the Younger, who was a contemporary of mation included in the text. No one can read Columella, the Roman author of agriculture everything that has been and is being pub- and viticulture textbooks, reads “errare humanum lished, even in the admittedly relatively narrow est, sed in perseverare diabolicum” (“to err is field of grapevine anatomy and physiology. human, to persevere is devilish”). About the Author
Markus Keller is the Chateau Ste. Michelle University in Wagga Wagga, Australia, before Distinguished Professor of Viticulture at coming to eastern Washington. Dr. Keller is Washington State University’s Irrigated Agri- the author of numerous scientific and technical culture Research and Extension Center in papers and industry articles in addition to Prosser, Washington. He received his master’s being a frequent speaker at scientific confer- degree in agronomy (plant science) in 1989 ences and industry meetings and workshops. and a doctorate in natural sciences in 1995 from He also has extensive practical experience in the Swiss Federal Institute of Technology in both the vineyard and the winery as a result Zu¨ rich. He has taught and conducted research of work in the family enterprise, and he was in viticulture and grapevine physiology on awarded the Swiss AgroPrize for innovative three continents, beginning at the Swiss Federal contributions to Switzerland’s agriculture Research Station for Fruit-Growing, Viticulture industry. His current research focuses on and Horticulture in Wa¨denswil (now Agro- environmental factors and management prac- scope Changins-Wa¨denswil), Switzerland, and tices as they influence crop physiology and then moving to Cornell University in Geneva, production of wine and juice grapes. New York, and from there to Charles Sturt
vii Acknowledgments
Completing this book would have been and encouragement from Nancy Maragioglio impossible without the help and support of and Carrie Bolger at Elsevier are greatly many individuals to whom I express my deep acknowledged. They provided numerous sug- gratitude. Many of the illustrations were gestions that improved this book and were skillfully drawn by Adrienne Mills, and Lynn always quick to answer my countless ques- Mills helped with data collection and assisted tions. Special thanks to my wife, Sandra Wran, with some of the most recalcitrant illustra- for scanning my entire slide collection and tions. I thank Gregory Gasic for reviewing supporting this project in many ways from the entire manuscript, identifying errors, and beginning to end. offering many insightful suggestions. The help
xi Elsevier: The Science of Grapevines: Keller Página 1 de 2
The Science of Grapevines
Keller
Table of Contents
Chapter 1 Botany and Anatomy
1.1 Botanical classification and geographical distribution
1.2 Cultivars, clones and rootstocks
1.3 Morphology and anatomy
Chapter 2 Phenology and Growth Cycle
2.1 Seasons and daylength
2.2 Vegetative cycle
2.3 Reproductive cycle
Chapter 3 Water Relations and Nutrient Uptake
3.1 Osmosis, water potential and cell expansion
3.2 Transpiration and stomatal action
3.3 Water and nutrient uptake and transport
Chapter 4 Photosynthesis and Respiration
4.1 Light absorption and energy capture
4.2 Carbon uptake and assimilation
4.3 Photorespiration
4.4 Respiration
Chapter 5 Partitioning of Assimilates
5.1 Photosynthate translocation and distribution
5.2 Canopy-environment interactions
5.3 Nitrogen assimilation and interaction with carbon metabolism
http://www.elsevierdirect.com/toc.jsp?isbn=9780123748812 5/4/2011 Elsevier: The Science of Grapevines: Keller Página 2 de 2
Chapter 6 Developmental Physiology
6.1 Yield formation
6.2 Grape composition and fruit quality
Chapter 7 Environmental Constraints and Stress Physiology
7.1 Responses to stress
7.2 Water: too much or too little
7.3 Nutrients: deficiency and excess
7.4 Temperature: too cold or too warm
7.5 Living with other organisms: defense and damage
Glossary
References
Internet Resources
http://www.elsevierdirect.com/toc.jsp?isbn=9780123748812 5/4/2011 CHAPTER 1
Botany and Anatomy
OUTLINE
1.1. Botanical Classification and Geographical 1.3. Morphology and Anatomy 20 Distribution 1 1.3.1. Root 21 1.3.2. Trunk and Shoots 27 1.2. Cultivars, Clones, and Rootstocks 9 1.3.3. Nodes and Buds 33 1.2.1. Variety versus Cultivar 9 1.3.4. Leaves 36 1.2.2. Cultivar Classification 13 1.3.5. Tendrils and Clusters 41 1.2.3. Clones 15 1.3.6. Flowers and Grape Berries 43 1.2.4. Rootstocks 16
1.1. BOTANICAL CLASSIFICATION species is a closed gene pool, an assemblage AND GEOGRAPHICAL of organisms that does not normally exchange DISTRIBUTION genes with other species. Their genes compel the individuals belonging to a species to per- The basic unit of biological classification is petuate themselves over many generations. the species. According to the “biological species Yet all life forms on Earth are interrelated; they concept,” a species is defined as a community all ultimately descended from a common of individuals—that is, a population or group ancestor and “dance” to the same genetic code, of populations, whose members can interbreed whereby different combinations of three con- freely with one another under natural condi- secutive nucleotides of each organism’s deoxy- tions but not with members of other popula- ribonucleic acid (DNA) specify different tions (Mayr, 2001; Soltis and Soltis, 2009). In amino acids that can be assembled into pro- other words, such communities are reproduc- teins. Because they are thus interrelated, organ- tively isolated. Although each individual of a isms can be grouped according to the degree of sexual population is genetically unique, each their genetic similarity, external appearance,
The Science of Grapevines 1 Copyright # 2010 Markus Keller. Published by Elsevier Inc. All rights reserved. 2 1. BOTANY AND ANATOMY and behavior. In the classification hierarchy, expressed it clearly: "Wherever many closely closely related species are grouped into a allied yet distinct species occur, many doubtful genus, related genera into a family, allied forms and varieties of the same species likewise families into an order, associated orders into a occur" and, furthermore, "there is no funda- class, similar classes into a division (plants) or mental distinction between species and vari- a phylum (animals), related divisions or phyla eties," and, finally, "varieties are species in the into a kingdom, and, finally, allied kingdoms process of formation" (Darwin, 2004). Indeed, not into an empire but a domain. The “evolu- modern genetic evidence indicates that the tionary species concept” recognizes this ances- various Vitis species evolved relatively recently tor–offspring connection among populations from a common ancestor so that they have not that may follow distinct evolutionary paths to yet had time to develop the complete repro- occupy separate ecological niches but may con- ductive isolation that normally characterizes tinue to interbreed for some time (Soltis and biological species. Thus, Vitis species are Soltis, 2009). defined as populations of vines that can be As is the case with many plants, the species easily distinguished by morphological traits, of the genus Vitis are not very well-defined such as the anatomy of their leaves, flowers, because of the extreme morphological variation and berries, and that are isolated from one among and within populations of wild vines another by geographical, ecological, or pheno- (Currle et al., 1983; Hardie, 2000; Mullins et al., logical barriers; such species are termed ecospe- 1992). This implies the following: (1) All Vitis cies (Hardie, 2000; Levadoux, 1956; Mullins species are close relatives that share a relatively et al., 1992). The following is a brief overview recent common ancestor, and (2) evolution is of the botanical classification of grapevines, still at work, throwing up new variants all the starting with the domain at the top of the time (see Chapter 2.3). Many vine species are hierarchy and finishing with a selection of actually semispecies—that is, populations that species at the base. partially interbreed and form hybrids under natural conditions, which is in fact common Domain Eukarya All living beings, making among plants and may be an important avenue up the earth’s biological diversity or biodiver- for the evolution of new species (Soltis and sity, are currently divided into the three great Soltis, 2009). Despite some hybridization where domains of life: the Bacteria, the Archaea, and their natural habitats overlap, however, the the Eukarya. The Eukarya (eukaryotes; Greek eu various Vitis gene pools usually stay apart so ¼ true, karyon ¼ nucleus) include all terrestrial, that the populations remain recognizably dif- sexually reproducing “higher” organisms with ferent. Grapevines are a good example of the relatively large cells (10–100 mm) containing a limits of taxonomic systems, demonstrating true cell nucleus, in which the DNA-carrying that there is a continuum of differentiation chromosomes are enclosed in a nuclear mem- rather than a set of discrete, sexually incompat- brane, and cell organelles such as mitochondria ible units. As early as 1822, the Rev. William and plastids (Mayr, 2001). They evolved fol- Herbert asserted that "botanical species are lowing injections of oxygen into the atmo- only a higher and more permanent class of sphere caused by abiotic (i.e., nonbiological) varieties," and in 1825 the geologist Leopold factors such as plate tectonics and glaciation von Buch postulated that "varieties slowly (Lane, 2002). The vast majority of life and become changed into permanent species, which the bulk of the world’s biomass—the small are no longer capable of intercrossing" (both (1–10 mm), single-celled prokaryotes (Greek cited in Darwin, 2004). Charles Darwin later pro ¼ before) with cell walls composed of 1.1. BOTANICAL CLASSIFICATION AND GEOGRAPHICAL DISTRIBUTION 3 peptidoglycans (protein–polysaccharides)—is fermentation, other fungi cause diseases of grouped into the other two domains. However, grapevines (see Chapter 7.5); for example, gray both the photosynthetic organelles (chloro- rot is caused by Botrytis cinerea Pers.:Fr. and pow- plasts, from cyanobacteria) and the “power dery mildew by Erysiphe necator [aka Uncinula plants” (mitochondria, from proteobacteria) of necator (Schwein.) Burr.]. Animals can also be eukaryotic cells have descended from (symbi- important pests of grapevines, especially certain otic) bacteria that “infected” other single-celled insects (e.g., phylloxera, Daktulosphaira vitifoliae organisms (or were “swallowed” by them) over Fitch), mites, and nematodes. 1 billion years ago. These organelles still retain some of their own DNA (i.e., genes), although Division (Phylum) Angiospermae (Synonym more than 95% of their original genes have Magnoliophyta) The angiosperms or, in new since been and are still being lost or transferred terminology, the magnoliophytes are the flow- (donated) to their host’s nucleus (Timmis et al., ering plants, which include approximately 2004). Some bacteria cause diseases of grape- 270,000 species. They are believed to have vines; for example, crown gall is caused evolved from a common ancestor that lived by Agrobacterium vitis and Pierce’s disease by approximately 160 million years ago during Xylella fastidiosa (see Chapter 7.5). the late Jurassic period, and they make up the most evolutionarily successful group of plants. Kingdom Plantae The Eukarya comprise at Angiosperms are the plants with the most least four kingdoms; the number changes as the complex reproductive system: They grow their relationships among organisms become better seeds inside an ovary (Greek angeion ¼ pot, known. The Animalia comprise the multicellular vessel) that is itself embedded in a flower. After animals with two sets of chromosomes, cells the flower is fertilized, the other flower parts without cell walls, except in arthropods (insects, fall away and the ovary swells to become a spiders, and the like), which have chitin cell fruit, such as a grape berry. Indeed, the produc- walls, and are the domain of zoology. The Plantae tion of fruits is what defines the angiosperms or plants have a haplo-diploid life cycle and cell and sets them apart from the gymnosperms, walls composed of cellulose, and they are studied with whom they are classed in the Superdivi- in the field of botany. The Fungi include the hap- sion Spermatophyta or seed plants. loid mushrooms, molds, and other fungi, with cell walls composed of glucans and chitin; they Class Dicotyledoneae (Synonym Magno- are studied in mycology. The Protista or Proto- liopsida) This class is large and very diverse, ctista are a catch-all group of all other “higher and its members are often called dicot plants. order” organisms from single-celled microbes, The vast majority of plants ( 200,000 species), including unicellular fungi, plants (green algae), including most trees, shrubs, vines, and flowers, and animals (protozoans), to large, multicellular and most fruits, vegetables, and legumes belong seaweeds (algae and kelp). There are approxi- to this group. Like all members of the Dicotyledo- mately 500,000 plant species, which are classified neae, grapevines start their life cycle with two into 12 phyla or divisions based largely on repro- cotyledons (seed leaves) preformed in the seed. ductive characteristics. The vascular or higher plants, to which grapevines belong due to their Order Rhamnales (Vitales According to the water conduits, form the Subkingdom Tracheo- Angiosperm Phylogeny Web) Grapevines bionta. Whereas one group of fungi (singular fun- belong to the order Rhamnales, which gets its gus), the yeasts (especially Saccharomyces name from the genus Rhamnus, the buckthorns. cerevisiae), turns grapes into wine through The order has three families: Rhamnaceae 4 1. BOTANY AND ANATOMY
(e.g., Ziziphus jujuba Mill., jujube tree), Leeaceae • Fused flower petals that separate at the base, (the oleasters), and Vitaceae. Plants of the family forming a “calyptra” or cap Leeaceae are more recognizable as being related • Soft and pulpy berry fruits to grapevines than those belonging to the Rham- naceae, being shrubs or trees with flowers aggre- Genus Muscadinia Members of the genus gated in inflorescences, black berries, and seeds Muscadinia usually have glabrous (hairless) that resemble grape seeds, also named pips. leaves, simple tendrils, nonshredding bark, Some taxonomists have now separated the Vita- nodes without diaphragms, and hard wood ceae (Jansen et al., 2006) and Leeaceae from the (Currle et al., 1983; Mullins et al., 1992). Because Rhamnales and placed them in the order Vitales. they do not root from dormant cuttings, they are usually propagated by layering, although Family Vitaceae The members of this family they do root from green cuttings. The “home- are collectively termed grapevines. The family land” of this genus extends from the southeast- contains approximately 1000 species assigned ern United States to Mexico. The genus has to 17 genera that are typically shrubs or woody only three species, which are all very similar lianas that climb by means of their leaf-opposed and may not even deserve to be classed as tendrils—hence the name Vitaceae (Latin viere ¼ separate species (Currle et al., 1983; Mullins to attach). Although most species of this family et al., 1992; Olien, 1990). reside in the tropics or subtropics, a single spe- • Muscadinia rotundifolia Small (formerly Vitis cies from the temperate zones has become the rotundifolia Michaux): A dioecious plant, world’s leading fruit crop grown in almost 90 although breeding has yielded perfect- countries for wine and juice production or as flowered and female cultivars, such as fresh table grapes or dried grapes (raisins). Vita- Noble, Carlos, or Magnolia, known as ceae roots are generally fibrous and well “muscadines” that are grown as table, jelly, branched, and they can grow to several meters or wine grapes. The species is native of the in length. The leaves are alternate, except dur- southeastern United States. The musky ing the juvenile stage in plants grown from flavor and thick skins of the fruit can be seeds, and can be simple or composite. The unattractive. The species has co-evolved fruits are usually fleshy berries with one to four with and therefore resists or tolerates the seeds. All cultivated grapes belong to either grapevine diseases and pests native to North the genus Muscadinia (2n ¼ 40 chromosomes) America, including the fungi powdery or the genus Vitis (2n ¼ 38 chromosomes). The mildew and black rot, the slime mold downy former classification of Muscadinia and Euvitis mildew, the bacterium causing Pierce’s as either subgenera or sections of the genus disease, the aphid phylloxera (Daktulosphaira Vitis has fallen out of favor among taxonomists vitifoliae), and the nematode Xiphinema index (Mullins et al., 1992). Because of the different (which transmits the grapevine fanleaf numbers of chromosomes, crosses between virus), but is sensitive to winter frost and these two genera rarely produce fertile hybrids. lime-induced chlorosis (Alleweldt and Key morphological characteristics of the two Possingham, 1988). Although usually genera include the following: incompatible in both flowering and grafting • Simple leaves with Vitis species, it does produce fertile • Simple or forked tendrils hybrids with V. rupestris, which allows it to • Generally unisexual flowers—that is, either be used in modern (rootstock) breeding male (staminate) or female (pistillate) programs. 1.1. BOTANICAL CLASSIFICATION AND GEOGRAPHICAL DISTRIBUTION 5
• Muscadinia munsoniana Small (Simpson): dioecious (Greek dis ¼ double, oikos ¼ house), Native to Florida and the Bahamas, with containing imperfect male (i.e., female sterile better flavor and skin characteristics than or staminate) or female (i.e., male sterile or pis- M. rotundifolia, but not cultivated. tillate) flowers on different individual plants, • Muscadinia popenoei Fennell: Native to whereas the cultivated varieties of V. vinifera southern Mexico (“Totoloche grape”), have perfect or, in a few cases, physiologically relatively unknown. female flowers (Negrul, 1936; Pratt, 1971; see also Figure 1.1). Members of this genus are very Genus Vitis The genus Vitis occurs predomi- diverse in both habitat and form. Nevertheless, nantly in the temperate and subtropical climate all species within the genus can readily inter- zones of the Northern Hemisphere (Mullins breed to form fertile interspecific crosses called et al., 1992; Wan et al., 2008a). All members of hybrids, which implies that they had a rela- this genus are perennial vines or shrubs with tively recent common ancestor. Moreover, all tendril-bearing shoots. This genus probably species can be grafted onto each other. The comprises 60–70 species (plus up to 30 fossil genus is often divided into two major groups: species and 15 doubtful species) spread mostly the American group and the Eurasian group. throughout Asia ( 40 species) and North The dominant species of the two groups differ America ( 20 species) (Alleweldt and Possing- greatly in their useful agronomic traits ham, 1988; Wan et al., 2008b,c). The Eurasian (Table 1.1), which makes them attractive breed- species Vitis vinifera L. gave rise to the over- ing partners (Alleweldt and Possingham, 1988; whelming majority of grape varieties cultivated This et al., 2006). Unfortunately, none of the today. Plants that belong to this genus have many attempts and thousands of crosses that hairy leaves with five main veins, forked ten- have been tested to date has truly fulfilled the drils, bark that shreds when mature, nodes breeders’ hopes to combine the positive attri- with diaphragms, and soft secondary wood. butes while eliminating the negative ones They all can form adventitious roots, which contained in the natural genetic variation of permits propagation by cuttings, yet only V. the two groups. Perhaps the genes conferring vinifera, V. riparia, and V. rupestris root easily disease resistance are coupled to those respon- from dormant cuttings. Although the ancestor sible for undesirable fruit composition. Indeed, of all Vitis species may have had perfect (i.e., hybrids have often been banned in European bisexual or hermaphroditic) flowers (McGo- wine-producing countries because of their per- vern, 2003), the extant wild species are ceived poor fruit (and resulting wine) quality.
FIGURE 1.1 Flower types in the genus Vitis: perfect flower (left), female flower (center), and male flower (right). Illus- trations by A. Mills. 6 1. BOTANY AND ANATOMY
TABLE 1.1 Broad Viticultural Traits of American grow near a permanent source of water, such as and Eurasian Grapevine Species a river, stream, or spring (Morano and Walker, 1995; Figure 1.2). Following is an incomplete list Trait Eurasian Species American Species of some of the more important species: Fruitfulness Good Poor or highly • Vitis labrusca L.: Vigorous climber (“northern variable fox grape”) native to the eastern United Fruit quality Good Poor States from Georgia to southeastern Canada, Usefulness Highly diverse Niche products with Indiana as its western limit. This products species differs from all others in that it Propagation Good Variable usually has continuous tendrils (a tendril at capacity every node). Some of its cultivars (e.g., Lime Good Highly variable Concord and Niagara) are commercially tolerance grown in the United States for juice, jam, Phylloxera Poor Good or variable jelly, and wine production. The classification tolerance of these cultivars, however, is still debated; Disease Poor Good or variable Concord (which has perfect flowers) resistance probably is a natural hybrid of V. labrusca and V. vinifera, and thus it has been classed as Vitis labruscana L. Bailey (Mitani et al., The only unequivocal success story thus far has 2009; Mullins et al., 1992). The distinct foxy been the grafting of phylloxera-susceptible flavor (caused by methyl anthranilate) European wine grape cultivars to rootstocks unique to this species is popular in the that are usually hybrids of tolerant American United States but strange to Europeans. The Vitis species (see Chapter 1.2). species is cold tolerant and resistant to powdery mildew and crown gall, but it is American Group Depending on the taxono- susceptible to phylloxera, downy mildew, mist, this group contains between 8 and 34 spe- black rot, and Pierce’s disease and has poor cies, of which several have become economically lime tolerance (i.e., prefers acid soils). important as wine or juice grapes. Because of Hybrids of V. labrusca were exported to their varying resistance to the North American Europe at the beginning of the 19th century. grapevine diseases and pests, members of this Some of these plants carried powdery group are also being used as rootstocks (see mildew, downy mildew, black rot, and Chapter 1.2) or crossing partners in breeding pro- phylloxera, which drove most populations of grams (Alleweldt and Possingham, 1988; This wild vines extinct and brought the European et al., 2006). As an aside, crosses are always listed wine industry to the verge of destruction. as maternal parent paternal parent (i.e., the • Vitis aestivalis Michaux: Vigorous climber mother’s name comes first). The species of this native to eastern North America, growing in group generally have thinner shoots with longer dry upland forests and bluffs. It is very cold internodes and less prominent nodes than the hardy (to approximately 30 C), drought Eurasian species. They also have small buds, tolerant and also tolerates wet and humid and the leaves have very shallow sinuses and summers (“summer grape”), and is resistant often a glossy surface. All grape species native to powdery and downy mildew and Pierce’s to North America are strictly dioecious (i.e., none disease. The species is very difficult to of them has perfect flowers), and most of them propagate from cuttings. Its fruits are used 1.1. BOTANICAL CLASSIFICATION AND GEOGRAPHICAL DISTRIBUTION 7
FIGURE 1.2 The “bank grape,” V. riparia, growing in a forest in upstate New York (left) and the “canyon grape,” Vitis arizonica Engelmann, growing up a riverside tree in Utah’s Zion National Park (right). Notice the large size of the wild vines and the long “trailing trunks” at the bottom right. Photos by M. Keller.
to make grape jelly, and the cultivars Norton phylloxera, and is resistant to fungal and Cynthiana are commercially grown as diseases but susceptible to Pierce’s disease. wine grapes in the southern and midwestern • Vitis rupestris Scheele: Native to the United States (Tarara and Hellman, 1991). It southwestern United States from Texas to is possible that the two names are synonyms Tennessee, the species is now almost extinct. for the same cultivar and/or that they are It is found in rocky creek beds (“rock hybrids between V. aestivalis and V. labrusca grape”) with permanent water, and it is or even V. vinifera. vigorous, shrubby, and rarely climbs. It has • Vitis riparia Michaux: Widespread in North deep roots for anchorage but is not very America from Canada to Texas and from the drought tolerant on shallow soils, and its Atlantic Ocean to the Rocky Mountains. This lime tolerance is variable. The species species climbs in trees and shrubs along tolerates phylloxera and is resistant to riverbanks (“bank grape”) and prefers deep powdery mildew and downy mildew, but it alluvial soils, but does not do well in is susceptible to anthracnose. calcareous soils (i.e., prefers acid soils), and • Vitis berlandieri Planchon: Native to central its shallow roots make it susceptible to Texas and eastern Mexico, this species drought (a trait it also confers to the climbs on trees on deeper limestone soils rootstocks derived from its crosses with between ridges. It is one of very few other species). It is the earliest to break buds American Vitis species that have good lime and ripen of all the American species, tolerance. Its deep root system makes it matures its canes early, is very cold hardy relatively drought tolerant, but it is very (to approximately 36 C), is tolerant of susceptible to waterlogging. The species 8 1. BOTANY AND ANATOMY
breaks buds and flowers much later than and Hopf, 2001). This is the most well- other species and is the latest ripening of the known species of the Eurasian group American group with very late cane because it gave rise to most of the cultivated maturation. It is somewhat tolerant of grapes grown today. Because the cultivated phylloxera and resistant to fungal grapevines have hermaphroditic or, rarely, diseases and Pierce’s disease, but it is very physiologically female flowers, they are difficult to propagate and to graft often grouped into the subspecies V. vinifera (Mullins et al., 1992). ssp. sativa (also termed V. vinifera ssp. • Vitis candicans Engelmann: Very vigorous vinifera). However, the concept of subspecies climber native to the southern United States (or geographical races) as taxonomically and northern Mexico. The species is drought recognizable populations below the species tolerant, relatively tolerant of phylloxera, and level is biologically virtually worthless and, resistant to powdery and downy mildew and according to many taxonomists, V. vinifera Pierce’s disease, but it is difficult to sativa is merely the domesticated form of propagate. Other southern species, such as V. vinifera (ssp.) sylvestris. According to this V. champinii Planchon and V. longii Prince, are view, the differences between the two forms probably natural hybrids of V. candicans, are the result of the domestication process V. rupestris, and other native species. They are (This et al., 2006)—that is, they arose through highly resistant to nematodes. human rather than natural selection. The species is highly tolerant of lime, even more Eurasian Group There are approximately 40 so than V. berlandieri, and drought. known species in this group, most of them • Vitis sylvestris (or silvestris) (Gmelin) Hegi: confined to eastern Asia. Chinese species are Native to an area spanning central Asia to particularly diverse, growing in the dry south- the Mediterranean region, this group west, the northern and southern foothills of contains the dioecious wild vines (now also the Himalayas, the very cold northeast, and termed Lambrusca vines) of Asia and the hot and humid southeast. Although some Europe, growing mainly in damp of them are resistant to fungal diseases and woodlands (“forest grape”) on alluvial soils may tolerate high humidity (Li et al., 2008; of riverbanks and hillsides. Taxonomists Wan et al., 2007, 2008a), most of these species debate whether this group of vines deserves are little known, and there may be several addi- species status or whether it is a subspecies of tional species that have not yet been described. V. vinifera (ssp. sylvestris) because, apart Most Eurasian species are not resistant to the from their flowers, the two look very similar North American grapevine diseases, and yet and interbreed readily (Mullins et al., 1992). It one of them has come to dominate the grape has almost disappeared in Europe and is and wine industries throughout the world. considered an endangered species in some countries, mainly because of the destruction • Vitis vinifera L.: Native to western Asia and of its natural habitats and its susceptibility to Europe between 30 and 50 N, but phylloxera and the mildews introduced from temporarily confined to the humid and North America (Arnold et al., 1998). The high forested to arid, volcanic mountain ranges of humidity and periodic flooding in wooded the southern Caucasus between the Black river valleys may have protected the Sea and the Caspian Sea, and to the remaining populations from destruction by Mediterranean region during the ice ages phylloxera. These wild grapes are thought to (Hardie, 2000; Mullins et al., 1992; Zohary be more cold tolerant than their cultivated 1.2. CULTIVARS, CLONES, AND ROOTSTOCKS 9
siblings and to be resistant to leafroll and Cultivated grapevines of sufficiently similar fanleaf viruses (Arnold et al., 1998). vegetative and reproductive appearance are • Vitis amurensis Ruprecht: The genetically usually called “grape varieties” by growers diverse native of northeastern China and and “cultivars” by botanists. Botanically Russian Siberia (“Amur grape”) is said to be speaking, a variety includes individuals of a the cold-hardiest of all Vitis species (wild) population that can interbreed freely, (Alleweldt and Possingham, 1988; Wan et al., whereas, according to the International Code 2008a). However, budbreak occurs of Nomenclature for Cultivated Plants (Interna- approximately 1 month earlier than in tional Society for Horticultural Science, 2004), a V. vinifera, and bloom and fruit maturity are cultivar is "an assemblage of plants that has also advanced, which seems to be true for been selected for a particular attribute or com- many other Chinese species (Wan et al., bination of attributes, and that is clearly dis- 2008c). It is resistant to downy mildew and tinct, uniform, and stable in its characteristics Botrytis but susceptible to phylloxera (Du and that, when propagated by appropriate et al., 2009). Some cultivars with female or, means, retains those characteristics." A cultivar rarely, hermaphroditic flowers and small can be produced both sexually (i.e., from a clusters are grown in northeastern China. seedling) and asexually (i.e., as a clone), but • Vitis coignetiae Pulliat: Native to Japan, and only the latter method will give grapevine des- used locally for jam production, the species cendents that are genetically identical (i.e., strongly resembles the American V. labrusca “true to type”). Seedlings of so-called grape (Mullins et al., 1992). varieties are not identical copies of their mother plant. Because grapevines are heterozygous across a large number of chromosome positions 1.2. CULTIVARS, CLONES, AND or loci (singular locus) in their genomes, each ROOTSTOCKS seed may give rise to a cultivar with distinct characteristics (Mullins et al., 1992, Thomas 1.2.1. Variety versus Cultivar and Scott, 1993; also see Chapter 2.3). Grapes were among the first fruit crops to be The study of the botanical description, iden- domesticated, along with olives, figs, and dates tification, and classification of plants belonging (Zohary and Hopf, 2001; Zohary and Spiegel- to the genus Vitis and of their usefulness for Roy, 1975). Although all four wild fruits had viticulture is termed ampelography (Greek ampe- ranges stretching far beyond the eastern Medi- los ¼ vine, graphos ¼ description). The descrip- terranean region, the deliberate cultivation of tors have traditionally included such visual grapes for winemaking as well as fresh fruit traits as shoot tips (shape, hairs, and pigmen- and raisin production probably started about tation), leaves (shape of blade with lobes, 7000 to 8000 years ago in the eastern part of sinuses, and serrations), fruit clusters (size the Fertile Crescent that spans the modern- and shape), and berries (size, shape, and day countries of the Near East: Egypt, Israel, pigmentation) (Galet, 1985, 1998; Viala and Lebanon, Jordan, Syria, eastern Turkey, Iraq, Vermorel, 1901). The advent of DNA finger- and western Iran. This is approximately the printing has led to its adoption for the identifi- same time that farmers invented irrigation cation of selections in grape collections and and precedes by approximately 2000 years the is increasingly being used to uncover the his- earliest civilization of Sumer in southern Meso- torical origins and genetic relationships of potamia (McGovern et al., 1996). As an interest- grapevines. ing aside, the Sumerian symbol for “life” is a 10 1. BOTANY AND ANATOMY grape leaf, and the “tree of life” of many in a region comprising the Iberian Peninsula, ancient civilizations is the grapevine. These Central Europe, and Northern Africa (Arroyo- early societies regarded wine as “nectar of the Garcı´a et al., 2006). Whereas almost all wine gods,” and wine drinking was considered to and juice grapes contain seeds, many raisin be a hallmark of civilization: “Savages” or “bar- and most table grape cultivars are seedless, a barians” drank no wine. The biblical legend, selected trait that seems to originate from a modeled on the earlier Mesopotamian Epic of very narrow genetic base, mainly Sultanina Gilgamesh (McGovern, 2003), of Noah’s ark (synonyms Sultana, Thompson Seedless, and stranding near volcanic Mount Ararat on the Kishmish), an ancient Middle Eastern grape Turkish/Iranian/Armenian border and his variety (Adam-Blondon et al., 2001; Iba´n˜ez subsequent planting of the first vineyard also et al., 2009). Most current cultivars are not pro- reflects precisely the probable origin of viticul- ducts of deliberate breeding efforts but are ture. Nevertheless, new archaeological discov- the results of continuous selection over many eries show that wine might have been made centuries of groups of grapevines that were from wild grapes in China as far back as 9000 spontaneously generated by mutation and years ago (McGovern et al., 2004). In Europe, intraspecific crosses via sexual reproduction grapes were grown at least 4000 years ago (Riv- (see Chapter 2.3) and via somatic (Greek soma era Nu´ n˜ez and Walker, 1989). Clearly, humans ¼ body) mutations—that is, mutations occur- quickly learned how to use fermentation as ring in the dividing cells of the shoot apical one of the most important food preservation meristem (see Chapter 1.3), typically during technologies; they also learned to preserve bud formation (so-called bud sports). grapes as raisins by drying them. Domestica- Mutations (Latin mutatio ¼ change) in the tion of V. vinifera was also accompanied by a grape genome arise through rare, chance mis- change from dioecious to hermaphroditic takes in DNA replication during the process reproduction and an increase in seed, berry, of cell division. Whenever a cell divides, it first and cluster size (This et al., 2006); perhaps early has to double (i.e., copy) its DNA so that each viticulturists selected not only more fruitful daughter cell gets a complete set of chromo- vines but also plants that were self-pollinated, somes. Millions of nucleotides are regularly thus eliminating the need for fruitless pollen copied with amazing fidelity, and sophisticated donors. repair mechanisms attempt to fix most mis- Today, an estimated 10,000 or so grape culti- takes that do occur. However, every so often, vars are being grown commercially, although a nucleotide is switched with another one, or DNA fingerprinting suggests that a more accu- a whole group of nucleotides is inserted, rate figure may be approximately 5000 (This repeated, omitted, or moved to a different et al., 2006). Many cultivated grapes are closely location on the chromosome. Many DNA related to one another, and many are known by sequences, called “jumping genes” or transpo- several or even many synonyms (different sons, even move on their own, some of them names for the same cultivar) or homonyms employing a “cut-and-paste” strategy, and (identical name for different cultivars). The may cause mutations if they cannot be silenced vast majority of cultivars belong to the species by the grape genome (Benjak et al., 2008; Lisch, V. vinifera, and comparisons of the DNA 2009). New genes quite often seem to arise from contained in chloroplasts suggest that they duplication of an existing gene and subsequent might have originated from (at least) two geo- mutation of one of the two copies (Dı´az- graphically distinct populations of V. sylvestris: Riquelme et al., 2009; Firn and Jones, 2009). one in the Near and Middle East and the other Many mutations are also caused by damage to 1.2. CULTIVARS, CLONES, AND ROOTSTOCKS 11 the DNA by so-called reactive oxygen species and because they accumulate over time, the that are produced during oxidative stress variation also increases. Although ancient culti- (Halliwell, 2006; Mller et al., 2007; see also vars (e.g., Pinot noir) can be quite heterogeneous Chapter 7.1). Mutations, if they are not lethal, and are planted as many different clones, the can give rise to slightly different gene variants genetic similarity of these clones typically is termed alleles (Greek allelos ¼ each other). The still on the order of 95–99% (Bessis, 2007; genome of the Vitis species is approximately 500 Wegscheider et al., 2009). Similarly, almost all of million DNA base pairs long and comprises an the approximately 20 cultivars with “muscat” estimated 30,000 genes (Jaillon et al., 2007; Velasco aroma, most of them used as raisin or table grapes et al., 2007) that form both a vine’s “building and many of them with a number of synonyms, plan,” including its flexible architectural and are direct descendants of Muscat a` petits grains engineering design, and its “operating manual” (synonym Muscat blanc) or Muscat of Alexandria by encoding (i.e., prescribing) the amino acid (Crespan and Milani, 2001), which is also one makeup of proteins. In comparison, humans are parent of the Argentinian Torronte´s cultivars thought to have fewer than 35,000 genes. Because (Agu¨ero et al., 2003). there are genes that influence growth habit, leaf Hybridization, on the other hand, occurs by shape, disease resistance, cluster architecture, cross-pollination and fertilization (see Chapter berry color, and other quality attributes, some of 2.3) of flowers from different plants, which are the mutations also affect these traits (Bessis, genetically distinct, and thus give rise to a 2007). For instance, dark-skinned (i.e., anthocya- genetically novel individual. Both deliberate nin-accumulating) fruit is the “default” version interbreeding and natural hybridization have in the Vitaceae, and it appears that virtually all occurred many times in the history of viticul- V. vinifera cultivars with green-yellow fruit (so- ture. Many cultivars were originally selected called white cultivars) have a single common from domesticated local wild grapes (V. sylves- ancestor that arose from mutations of two neigh- tris or V. vinifera ssp. sylvestris) and further boring genes of an original dark-fruited grape- developed by interbreeding with other wild vine (Cadle-Davidson and Owens, 2008; This grapes or by the introduction of exotic varieties et al., 2007; Walker et al., 2007). These mutations (Arroyo-Garcı´a et al., 2006). Intriguingly, were probably caused by the insertion of a jump- although the majority of the French wine grape ing gene that renders grapes unable to “switch cultivars are genetically closely related to each on” the anthocyanin assembly line (Kobayashi other and to the wild grapes of western Europe, et al., 2004). The chances of both mutations most Italian and Iberian cultivars are different occurring together are extremely small; this from these and also more distinct among them- might have been a one-time event in the history selves, and most table grapes are altogether in a of grapevines, perhaps as recently as approxi- completely different “genetic league.” Never- mately 200,000 years ago (Mitani et al., 2009). theless, many “noble” grape lineages are as Although this implies that white-fruited cultivars interwoven as those of their human selectors. are likely to be genetically more closely related to For instance, it is thought that Cabernet franc each other than are cultivars with dark fruit, (which is closely related to Petit Verdot) and roughly half of all current grape cultivars have Traminer (synonym Savagnin) may have been fruit with “white” skin. Instability of at least selected from wild grapes, whereas Cabernet one of these mutations may account for the occa- Sauvignon resulted from a natural cross sional appearance of dark-skinned variants of between Cabernet franc and Sauvignon blanc white cultivars (Lijavetzky et al., 2006). Mutations (Bowers and Meredith, 1997; Levadoux, 1956). also add to the genetic variation within a cultivar, Cabernet franc also fathered Merlot and 12 1. BOTANY AND ANATOMY
Carmene`re (Boursiquot et al., 2009). Syrah (syn- genetically similar to many southeastern Euro- onym Shiraz), is derived from a cross between pean wine grapes and, surprisingly, also to the Dureza and Mondeuse blanche, whereas Syrah red-fleshed “teinturier” cultivar Alicante crossed with Peloursin gave rise to Durif (syno- Bouschet, may have been the result of a Gouais nym Petite Sirah) (Meredith et al., 1999; Vouilla- blanc cross with an unknown partner. In fact, moz and Grando, 2006). Durif and Peloursin, based on genetic evidence the ancient eastern on the other hand, are closely related to Malbec European (Croatian) cultivar Gouais blanc has and Marsanne, which are distinct from the thus far been proposed to be involved in more (other) Bordeaux and Burgundy cultivars. than 75 European cultivars. This prominent posi- Chardonnay, Gamay noir, Aligote´,Auxerrois, tion as a parent of many premium wine grape Melon, and other French cultivars all originated varieties is somewhat ironic because its fruit from the same two parents, Pinot and Gouais quality has long been considered inferior. Yet its blanc (synonym White Heunisch), which means vigorous growth and high fertility made it a they are full siblings (Bowers et al., 1999). Pinot favorite with growers, who grew it alongside blanc and Pinot gris are somatic berry-pigment selections from wild vines (e.g., Traminer with mutants of Pinot noir (Figure 1.3); indeed, highly variable yields) for centuries so that it variation in berry color seems to be a fairly com- became widespread throughout Europe in the mon outcome of somatic mutation (Bessis, 2007; Middle Ages. Moreover, because of their sour Furiya et al., 2009; Mu¨ller-Stoll, 1950; Regner berries, Gouais vines were often planted as buf- et al., 2000). The Pinot family in turn is thought fers around vineyards to deter potential grape to be derived from a spontaneous cross between thieves. Another old cultivar from the Adriatic Traminer and Pinot Meunier (synonym Schwarz- coast has also attained some popularity in riesling). Whereas Pinot also appears to be one of both the Old and the New World, albeit under the more distant ancestors of Syrah, Traminer fea- different names: Primitivo in southern Italy, tures as a parent of Sauvignon blanc, Sylvaner, Crljenak kasˇtelanski in Croatia, and Zinfandel in and other cultivars (Sefc et al., 1998). Gouais California (Fanizza et al., 2005; Maletic´ et al., 2004). blanc, crossed with a Traminer variety, probably As novel DNA fingerprinting tests show, gave rise to Riesling, and with Chenin blanc it what breeders get is not always what they set yielded Colombard and other cultivars. Even out to develop, despite their best efforts to Lemberger (synonym Blaufra¨nkisch), which is exclude nondesired pollen from their crosses (Bautista et al., 2008; Iba´n˜ez et al., 2009). The most well-known example is the German wine grape cultivar Mu¨ller-Thurgau, which has long been disseminated as a deliberate cross of Riesling flowers with Sylvaner pollen (hence the synonym Riesling Sylvaner) (Becker, 1976). More than 100 years after its introduction, however, genetic analysis exposed Madeleine Royale, a likely progeny of Pinot noir, as the illegitimate father (Dettweiler et al., 2000). Similarly, Cardinal, a table grape cultivar with worldwide distribution, may be derived from Alphonse Lavalle´e (a possi- ble progeny of Gros Colman and Muscat FIGURE 1.3 Pinot noir (left) growing next to Pinot gris/ Hamburg) and Ko¨nigin der Weinga¨rten (proba- Pinot blanc (right; photo by M. Keller). bly a descendant of Pearl of Csaba) and not, as 1.2. CULTIVARS, CLONES, AND ROOTSTOCKS 13 has been assumed, from Flame Tokay and cultivar name Traminer was first mentioned Alphonse Lavalle´e(Iba´n˜ez et al., 2009). Natural in 1349, Pinot gris (synonym Rula¨nder) in hybridization can also cross species boundaries, 1375, Pinot in 1394, Riesling in 1435, Chasselas as exemplified by the North American juice (synonym Gutedel) in 1523, and Sangiovese in grape cultivar Concord, which was selected from 1590, although it is not certain that the same awildV. labrusca seedling that may have been name was consistently applied to the same cul- pollinated by an unknown V. vinifera cultivar or tivar (von Bassermann-Jordan, 1923). Tradi- a seedling with partial V. vinifera parentage tional European vineyards were—and in some (Mitani et al., 2009; Mullins et al., 1992). Misidenti- areas still are—composed of a population of fication of planting material imported to new heterogeneous vines, and sometimes several grape growing regions is another common cultivars were planted together in the same problem. For instance, the so-called Bonarda in block so that many cultivated varieties have Argentina and Charbono in California are not no defined origin. The identification and culti- related to the various Italian cultivars of the same vation of “pure” cultivars is quite recent. names but are in fact both identical to the old French cultivar Corbeau (Martı´nez et al., 2008). 1.2.2. Cultivar Classification Although each cultivar originally began as a Following several millennia of cultivation and single vine that grew from a seedling, most repeated selection of spontaneous mutants and major cultivars grown today have been propa- natural as well as man-made intra- and interspe- gated vegetatively for a long time, some for cific crosses in many different regions, there is a many centuries and perhaps even millennia vast range of cultivated forms and types of grape- (Bessis, 2007; Hardie, 2000). Propagation by vines. Because thousands of grape cultivars are cuttings was undoubtedly in use by Roman being grown commercially, various attempts times 2000 years ago (Columella, 4–ca. 70 AD). have been made to group them into families or For instance, Gouais blanc is thought to have tribes. Unlike botanical classifications, these been brought from Croatia (then Dalmatia) to groupings are not always based on phenotypic France (then Gallia) along with Viognier (the or genotypic differences, whereby the genotype Croatian cultivar Vugava bijela seems to be is the sum of the genetic material of an organism, identical with Viognier and a close relative of and the phenotype arises from the interaction of the Italian Barbera and the Swiss Arvine) the genotype with the environment during around 280 AD by the Roman Emperor Probus, development (Mayr, 2001). The most common who encouraged vineyard development to methods involve classification on the basis of ensure economic stability. In fact, if a map of place or climate of origin, viticultural charac- the distribution of viticulture in Europe, the teristics, final use, or winemaking characteristics. Near East, and North Africa were to be over- Arguably the first cultivar classification was laid with a map of the Roman Empire at its attempted by Plinius (ca. AD 50) in Rome, who greatest extent, there would be an almost exact created three groups according to berry color geographic overlap. It is believed that the rare and yield: extant Swiss variety Re`ze is, if not identical with, at least a direct descendant of Raetica, • Anemic cultivars: Varieties with small, white which was considered by Roman writers to be grapes one of the two best wine grape varieties of the • Nomentanic cultivars: Varieties with red Empire (Vouillamoz et al., 2007). A long gap grapes, low yielding in cultivar description and identification fol- • Apianic cultivars: High-yielding varieties, lowed the demise of the Roman Empire. The poor quality 14 1. BOTANY AND ANATOMY
During the Middle Ages, European wine from the fact that grape cuttings have forever grapes were simply divided into two groups been carried to distant locations and used according to perceived wine quality (von for (deliberate or natural) interbreeding with Bassermann-Jordan, 1923): locally domesticated and selected variants. Therefore, cultivars have also been grouped • Vinum francicum:“Frentsch”grapes,low- according to their viticultural characteristics, yielding varieties of high quality (e.g., Traminer) although these can be modified by local soil • Vinum hunicum: “Huntsch” grapes, high- and climatic conditions and cultural practices: yielding varieties of poor quality (e.g., Heunisch) • Time of maturity (Table 1.2): The early The division proposed by the Russian bota- maturing Chasselas is used as a basis or nist Negrul (cited in Levadoux, 1956, and Mul- standard. lins et al., 1992) distinguishes three main • Vigor: Rate of shoot growth. ecological or ecogeographical groups of vari- • Productivity: Yielding ability. proles ¼ eties, called “proles” (Latin offspring), Classification is also possible in terms of based on their region of origin: what the grapes will be used for (Galet, 2000), • Proles pontica: Fruitful varieties with although some cultivars are used for various medium-sized, round berries that originated purposes: from the banks of the Aegean and Black Seas • Table grapes: Large, fleshy or juicy grapes, and spread throughout eastern and southern often seedless, some of them with muscat or Europe (e.g., Furmint, Clairette, Black foxy aromas. Examples include Cardinal, Corinth, and Rkatziteli) Cinsaut, Chasselas, and Muscat of Proles occidentalis • : Wine grape varieties of Alexandria. western Europe with small clusters and • Raisin grapes: Predominantly seedless small berries (e.g., Riesling, Chardonnay, grapes. Examples include Thompson Se´millon, Sauvignon blanc, Gewu¨ rztraminer, Seedless, Flame Seedless, Black Corinth the Pinots, and the Cabernets) (synonym Zante Currant), and Delight. Proles orientalis • : Mostly table grape varieties • Juice grapes: Some highly aromatic grapes, with large clusters and large, elongate berries especially in the United States. Examples originating from the Near East, Iran, include Concord and Niagara. Afghanistan, and central Asia (e.g., Thompson • Wine grapes: Very sweet, juicy grapes, often Seedless, the Muscats, and Cinsaut) low yielding. Examples include Riesling, Chardonnay, Se´millon, Sauvignon blanc, Modern genetic research has mostly borne Gewu¨ rztraminer, the Pinots and out Negrul’s grouping (Aradhya et al., 2003). Cabernets, Merlot, Tempranillo (synonyms Moreover, it also confirmed that many cultivars Aragonez, Tinta Roriz), Nebbiolo, and many of the proles occidentalis are closely related to the others. wild V. sylvestris (or V. vinifera ssp. sylvestris), • Brandy (distillation) grapes: Generally white that almost all of the proles pontica varieties are grapes producing bland, acidic wines. closely related to each other, and that the proles Examples include Ugni blanc (synonym orientalis are genetically distinct from the other Trebbiano), Colombard, and Folle blanche. groups. Yet Negrul’s classification is not clear- cut and contains some ampelographic errors. Of course, divisions can also be made on the In addition, all of these classifications suffer basis of winemaking characteristics, but this 1.2. CULTIVARS, CLONES, AND ROOTSTOCKS 15
TABLE 1.2 Classification of Some Grape Cultivars Based on Their Relative Heat Requirement to Reach Acceptable Fruit Maturity, Subjectively Defined as 18–20 Brix
Group Red Wine Cultivars White Wine Cultivars
1 (early) Madeleine Angevine 2 Blue Portuguese Chasselas, Mu¨ ller-Thurgau 3 Gamay, Dolcetto Pinot gris, Pinot blanc, Aligote´, Gewu¨ rztraminer 4 Pinot noir Chardonnay, Sauvignon blanc, Sylvaner 5 Cabernet franc, Lemberger Riesling 6 Merlot, Malbec, Zinfandel, Tempranillo, Cinsaut, Se´millon, Muscadelle, Chenin blanc, Marsanne, Barbera, Sangiovese Roussanne, Viognier 7 Cabernet Sauvignon, Syrah, Nebbiolo Colombard, Palomino 8 Petit Verdot, Aramon, Carignan, Grenache, Muscat of Alexandria, Ugni blanc Mourve`dre 9 (late) Tarrango, Terret noir Clairette, Grenache blanc
Modified from Viala and Vermorel (1909), Gladstones (1992), and Huglin and Schneider (1998).
method is strongly influenced by the market intense fruit pigmentation or other perceived environment and changing consumer demand: quality attributes, and resistance to a particular disease. Due to vegetative propagation, such • Grape composition: Basic characteristics clonal traits arise from somatic mutations (see (sugar, acid, pH, tannins, flavors, and Chapter 1.3) rather than during sexual repro- aroma) important for winemaking. duction (i.e., in germ cells, see Chapter 2.3). • Varietal aroma: White grapes can be Nonetheless, unlike in animals that keep aromatic (e.g., Riesling, Gewu¨ rztraminer, somatic and germ cells separate, somatic muta- and Muscats) or nonaromatic (e.g., tions may ultimately be propagated both vege- Chardonnay, Se´millon, and Sylvaner). tatively and by seeds and result in individual • Production costs: Value for winemaking plants of the same cultivar having slightly dif- reflected in wine and grape price structure ferent genotypes and sometimes phenotypes and appellation systems. (Franks et al., 2002; This et al., 2006). This geno- typic diversity, which accumulates over time, is 1.2.3. Clones termed clonal variation (Mullins et al., 1992; Riaz et al., 2002). If the change is sufficiently distinct In viticultural parlance, a clone (Greek klon (e.g., major change over a short time or many ¼ twig) is a group of grapevines of a uniform small changes over a long period), such clones type that have been vegetatively propagated may come to be called cultivars, as exemplified (usually by cuttings) from an original mother by the example of the fruit skin color mutants vine that would normally have been selected of Pinot: Pinot noir, Pinot gris, and Pinot blanc for a particular desired trait. Such traits include (Bessis, 2007; Regner et al., 2000). In fact, many low vigor, high yield, loose clusters, large or clones of the Pinot family are chimeras, which small berries, seedlessness, different or more are defined as plants with more than one 16 1. BOTANY AND ANATOMY genetically different cell population that arose cultivar are typically very small. For example, from a mutation in only one of the two function- all 39 Sangiovese clones analyzed using a ally distinct cell layers or lineages of the shoot DNA assay were found to have been derived apical meristem (Riaz et al., 2002; Thompson and from the same mother vine (Filippetti et al., Olmo, 1963). This includes another old Pinot 2005). (gris) clone, Pinot Meunier (synonym Schwarz- It is currently fashionable to plant clonal riesling), whose leaves are densely coated with selections in vineyards, although choices are white hair (Franks et al., 2002; Hocquigny et al., often based on the “fallacy of the perfect 2004). Chimeras also exist among clones of Char- clone”—that is, on the false assumption that donnay and Cabernet Sauvignon (Moncada et al., the “best” clone found on one site will also per- 2006), which suggests that layer-specific muta- form best in a new location. Yet the perfor- tions in the apical meristem may be an important mance or suitability of a clone on a particular source of clonal and varietal variation in grape- site is strongly modified by the site’s environ- vines. Therefore, as in the case of species dis- ment (soil, climatic conditions, cultural prac- cussed in Chapter 1.1, there is no clear-cut tices, etc.) and also depends on the desired distinction between clones and cultivars, a fact end use of the grapes produced on that site recognized by Charles Darwin while he devel- and even regulatory circumstances (e.g., yield oped his revolutionary theory of evolution by restrictions). This makes it nearly impossible (random) mutation and (directional) natural to predict clonal performance in new environ- selection (Darwin, 2004). ments. Planting several clones of a cultivar in Although clonal vines were selected and the same vineyard block not only enables vegetatively propagated for fruitfulness, fruit growers to conduct their own evaluation at a flavor, and wine quality in Roman times (Colu- particular site but also provides some insur- mella, 4–ca. 70 AD), organized and methodical ance against fluctuations in yield, fruit quality, clonal selection only began in Germany in the and disease susceptibility that come with the second half of the 19th century and did not absence of genetic variation of a single clone. begin in France until the 1950s. Selection is usu- ally based on the absence of symptoms of virus 1.2.4. Rootstocks diseases (e.g., leafroll and fanleaf), healthy growth, and good performance such as consis- Rootstocks are specialized stock material to tent yields and high wine quality, but criteria which grape cultivars with desirable fruit can also include specific viticultural traits such properties are grafted; the shoot portion of the as fruit set, disease resistance, or drought toler- two grafting partners is termed the scion, ance. Nevertheless, many so-called clones are whereas the rootstock provides the root system actually phenotypes caused by various combi- to the fused combination of genotypes. For the nations and degrees of virus infections rather grafting operation to be successful, the vascular than genuine genetic differences among the cambium responsible for cell division (see Chap- plants themselves. Today, clones exist for most ter 1.3) of the two grafting partners must make major grape cultivars, and their success is contact with each other so that they can “build” based on the ability of some clones to perform a connection between their separate “plumbing” differently in diverse environments. Differ- systems for water and nutrient supply. Grafting ences among clones, and thus within a cultivar, of grapevines was common in ancient times; in offer viticulturally relevant diversity and his De Re Rustica, the Roman writer Columella potentially better planting material. Nonethe- described techniques known to his ancestors. less, the genetic differences among clones of a The use of rootstocks derived from American 1.2. CULTIVARS, CLONES, AND ROOTSTOCKS 17
Vitis species saved grapegrowing in Europe diseases. Infested propagation material can from extinction due to the introduction, on thus be a potential source of inoculum (infec- imported planting material, of the aphid-like tious material) for new vineyards and new viti- insect phylloxera (Daktulosphaira vitifoliae)in cultural regions. the late 19th century (see Chapter 7.5). On the The technique of grafting combines the downside, rootstocks or their progeny that tolerance of soil-borne pests of American Vitis escaped, for example, from abandoned vine- species with the winemaking quality (i.e., yards have also become feral across Europe, consumer acceptance) of V. vinifera. Today, and their interbreeding populations are in some grape rootstock varieties are used not only for areas behaving as invasive species (Arrigo and their tolerance or resistance to root parasites, Arnold, 2007; Laguna, 2003). such as phylloxera and nematodes, but also for Commonly used rootstocks are either indi- their ability to influence crop maturity or their vidual Vitis species or crosses of two or more tolerance of adverse soil conditions such as species (Table 1.3), and due to the dioecious drought, waterlogging, lime, and acid or saline nature of their parents, they are either male or soils (May, 1994; Pongra´cz, 1983; see also female plants. Examples of male rootstocks Table 1.3). Rootstocks can also contribute to the include Teleki 5C and SO4 and Riparia gloire management of vine vigor and of grape maturity de Montpellier, whereas female rootstocks and composition (Currle et al., 1983; Dry and include Kober 5BB, 101-14 Millardet et de Coombe, 2004; Galet, 1998). Although the influ- Grasset, and Fercal (Meneghetti et al., 2006). ence on scion vigor is an important consider- The majority of rootstocks in use today are ation in rootstock choice, there are no truly hybrids of three species: V. riparia, V. rupestris, dwarfing rootstocks, unlike in the case of, for and V. berlandieri (Galet, 1998). Nevertheless, instance, apple. The vertical and horizontal the genetic basis of the world’s rootstocks is distribution of the root system of various root- extremely narrow because as many as 90% of stocks, it seems, is mostly a function of soil all V. vinifera vines are grafted to fewer than texture, composition, and water availability 10 different rootstock varieties. As discussed rather than an inherent trait of the rootstock geno- previously for scion cultivars, this strategy puts type (Smart et al., 2006). Grafting does not directly vineyards at risk from mutant strains of soil affect the color or flavor profile of the grapes pro- pests, including phylloxera, with potentially duced on the scion variety because the chemicals devastating effects if a resistance breaks responsible for these quality-relevant traits are through. An example is the failure in the produced inside the berry (see Chapter 6.2) and 1980s after many years of widespread use of are therefore determined by the genotype of the V. vinifera V. rupestris rootstock AxR1 in the scion. However, an indirect effect on fruit California. Moreover, with the exception of composition, especially on acidity, is possible those derived from M. rotundifolia, the so-called due to the potential influence of the rootstock on “resistant” rootstocks are not immune to phyl- scion vigor, canopy configuration, yield forma- loxera or nematodes (microscopic round- tion, and, possibly, nutrient uptake (Keller et al., worms); they just tolerate them better than do 2001a; Ruhl et al., 1988; Schumann, 1974). V. vinifera cultivars and therefore suffer less Some of the lesser known eastern Asian species, from infestation (Mullins et al., 1992). Tolerant such as V. amurensis, are used increasingly in species or cultivars may grow quite well in modern rootstock breeding programs for their the presence of these pests, but they do not pro- cold tolerance (Alleweldt and Possingham, hibit pest numbers from building up over time. 1988), although this diminishes the phylloxera They also can be symptomless carriers of virus resistance of the resulting crosses. 18
TABLE 1.3 Agronomic Characteristics of Important Grapevine Rootstocks ucpiiiyt gDeficiency Mg to Susceptibility ucpiiiyt Deficiency K to Susceptibility htptoaResistance Phytophthora aeo ec Grafting Bench of Ease rw alResistance Gall Crown co ri Maturation Fruit Scion hloeaResistance Phylloxera eaoeResistance Nematode ad olTolerance Soil Sandy rfe co Vigor Scion Grafted cdSi Tolerance Soil Acid lySi Tolerance Soil Clay loigTolerance Flooding ruh Tolerance Drought aiiyTolerance Salinity aeo Rooting of Ease ieTolerance Lime .BTN N ANATOMY AND BOTANY 1.
Rootstock Parent Species
Riparia Gloire V. riparia NA Rupestris V. rupestris ND St. George Rupestris du Lot V. rupestris D 420A Millardet V. berlandieri V. riparia YD et de Grasset 5BB Kober V. berlandieri V. riparia YYD SO4 V. berlandieri V. riparia YN 8B V. berlandieri V. riparia 5C Teleki V. berlandieri V. riparia A 161-49 Couderc V. berlandieri V. riparia 99 Richter V. berlandieri V. rupestris YYD 110 Richter V. berlandieri V. rupestris YYD 1103 Paulsen V. berlandieri V. rupestris NY D 140 Ruggeri V. berlandieri V. rupestris NY D 44–53 Male`gue V. riparia V. rupestris YNA 3309 Couderc V. riparia V. rupestris NA 101-14 Millardet V. riparia V. rupestris A et de Grasset Schwarzmann V. riparia V. rupestris Gravesac V. berlandieri V. riparia V. rupestris 1616 Couderc V. solonis V. riparia A Salt Creek V. champinii (Ramsey) Dogridge V. champinii Harmony V. champinii V. solonis
V. riparia ROOTSTOCKS AND CLONES, CULTIVARS, 1.2. Freedom V. champinii V. solonis V. riparia
Excellent/high...... Poor/low A: Advanced; D: Delayed N: No; Y: Yes Modified from Currle et al. (1983), Pongra´cz (1983), Galet (1998), and Dry and Coombe (2004). 19 20 1. BOTANY AND ANATOMY
1.3. MORPHOLOGY AND 50–75% of the biomass (dry matter) of ANATOMY cultivated grapevines. This proportion is much lower in young plants and increases with age and plant size; it is less in heavily cropped Grapevines are very vigorous, woody clim- vines and is also lower in humid climates than bers named lianas that are perennial (i.e., they in dry climates. The entire plant body is rigid live more than 2 years), polycarpic (i.e., they and strong yet flexible and adaptive enough flower many times during their life), and decid- to explore soil resources, intercept sunlight, uous (i.e., they shed their leaves each year). and develop seeds for propagation. This combi- With the aid of their tendrils and flexible nation of strength and flexibility is possible trunks, wild vines climb on trees to a height because plants employ a “bricks and mortar” of 30 m or more and spread out their foliage strategy for growth. The vine’s basic building over the tree canopy. Using the trees for sup- blocks, the cells (bricks), have the ability to port (i.e., seizing them as a natural trellis sys- divide and expand, and they are joined tem), they can grow to tremendous size; the together by an adhesive matrix, the cell walls canopy of a single vine may cover a surface (mortar). More precisely, the two cell walls of area of dozens of square meters. Grapevines adjoining cells are “glued” together by the so- can live to several hundred years of age: The called middle lamella, which prevents sliding famous Great Vine in London’s Hampton between cells. Growing cells secrete a primary Court Palace is thought to have been planted wall, which is reinforced in mature cells that before 1770 and continues to produce 200–300 have ceased expanding with a lignified second- kg of fruit each year. Vegetative propagation, ary wall for extra mechanical strength. The cell whether naturally by layering or artificially walls partly take on the function of bones in from cuttings, bud grafts, tissue culture, and animals: They give the plant body form and other means, extends a vine’s lifespan virtually rigidity, and like bones, cell walls are rich in indefinitely. In fact, both rooted cuttings and calcium. The cells differentiate into various grafted plants show clear signs of rejuvenation types, each “designed” to carry out specific compared to their “mother” plants (Munne´- physiological tasks under the direction of dif- Bosch, 2008). For instance, growth rates, leaf ferent sets of genes; although each cell of the gas-exchange rates, and fruiting of recently same plant contains the same complement of propagated vines are those of young plants and genes, different genes can be active or inac- are independent of the age of the vines from tive—that is, switched “on” or “off.” Such sim- which the propagation material was taken. ilar groups of cells that form functional units Like other higher plants, grapevines com- are termed tissues, and different tissues, such prise vegetative organs (roots, trunk, shoots, as epidermis, parenchyma, phloem, and xylem, leaves, and tendrils) and generative or repro- in turn make up the plant’s organs, which are ductive organs or fruiting structures (clusters described next. with flowers or berries). The vegetative part The evolution of such complex multicellular can be divided into a belowground portion structures from single-celled ancestors required (roots) and an aboveground portion (trunk the simultaneous development of innovative and shoots). The aboveground portion together solutions to solve the problem of communication with the reproductive organs is termed the between the cells forming a tissue, between vine’s canopy. The roots, trunk, and its trained different tissues within an organ, between extension, the cordon, together form the per- the diverse organs of the plant body, and among manent structure of the vine, which makes up plants of a community. As will become apparent 1.3. MORPHOLOGY AND ANATOMY 21 throughout this book, plants master the art of True roots develop from the hypocotyl intercell communication by cell-to-cell channels (Greek hypo ¼ below, kotyledon ¼ depression) named plasmodesmata, whereas a vascular of the embryo (Greek embryon ¼ unborn child). system composed of the xylem and phloem In vines grown from seeds, seed germination ensures interorgan communication (Lough and begins with water absorption, called imbibi- Lucas, 2006). The roots and shoots “converse” tion, that enables the embryonic root, or radicle with each other, at least in part, via the produc- (Latin radicula ¼ little root), to grow and rup- tion and release of different hormones (Greek ture the seed coat, forming a primary root or hormaein ¼ to stimulate); cytokinin and abscisic tap root, from which multiple secondary roots, acid are the main root hormones, whereas auxin branch roots or lateral roots, grow. This type and gibberellin are the main shoot hormones. of root morphology is typical of dicotyledons Hormones are defined as extracellular signaling and is called an allorhizic root system (Osmont molecules that act on target cells distant et al., 2007). In vegetatively propagated vines, from their site of production. Perhaps even more roots originate from the cambium layer (Latin fascinating, information exchange between cambiare ¼ to exchange) of woody cuttings neighboring plants is achieved by emitting and (Pratt, 1974). Most of these so-called adventi- receiving volatile chemicals transmitted through tious roots form near the nodes, but they also the air and by releasing soluble chemicals grow on the internodes. These roots are often into the soil and recruiting symbiotic micro- termed main roots, and they branch off into organisms that infect and thereby interconnect secondary, tertiary, etc. lateral roots to form a the roots of different plants. Of course, successful root system with a complicated architecture, communication requires not only the production which resembles the secondary homorhizic and export of appropriate signaling molecules root system typical of monocotyledons from transmitting cells (output signals) but (Osmont et al., 2007). In other words, main also the presence of a sophisticated network of roots develop directly from a cutting, whereas signal perception, decoding, and transformation lateral roots develop from other roots. Unlike in the receiving cells (input signals). Such the arrangement of shoot lateral organs, the signals integrate genetic programs that perceive number and placement of lateral roots are not and respond to both developmental and envi- predetermined (and hence inflexible) but ronmental cues to direct cell, tissue, organ, depend on the availability of water and nutri- and plant growth and physiology (Lough and ents in the soil (Malamy and Benfey, 1997). Lucas, 2006). Because the grapevine is unable to move away from poor soils, it relies on the ability to detect 1.3.1. Root these environmental cues and extend lateral roots to take advantage of favorable soil The root system is the interface between the regions. Moreover, in contrast to many other grapevine and the soil. It provides physical plant species, lateral root initiation in grape- support for the vine in the soil and is responsi- vines is not restricted to the unbranched apical ble for water and nutrient uptake. The roots zone. Grapevines are also able to grow new lat- also serve as storage organs for carbohydrates eral roots on older parts of the roots that have and other nutrients, which support the initial already developed a vascular cambium. Under growth of shoots and roots in spring, and for conditions of very high humidity and high tem- water. In addition, they are a source of plant perature, trunks and other aboveground parts hormones (cytokinins and abscisic acid), which of the vine can also form aerial roots. Whereas modify shoot physiology. this phenomenon is mostly restricted in the 22 1. BOTANY AND ANATOMY genus Vitis to greenhouse conditions (and rare mucilage secreted by the cap covers the surface even there), it is quite common in the natural of the maturing root. The mucigel harbors habitat of the Muscadinia (Viala and Vermorel, microorganisms and probably aids in the estab- 1909). lishment of symbiotic mycorrhiza (Greek mykes The root system of established grapevines ¼ fungus, rhı´za ¼ root) in the root tips, which comprises a far-reaching, highly branched form a network of fungal mycelium that con- structure with a surface area far exceeding that nects plants belowground. This network helps of the leaf canopy it supports. A mature, with the uptake of nutrients and water in cultivated grapevine can have more than 100 exchange for carbohydrates supplied to the km of total root length with a surface area fungi by the plants (Bais et al., 2006; Marschner, greater than 100 m2, whereas its leaf area is 1995; Smith et al., 2001). The site of active cell usually less than 10 m2. The woody roots, division (the apical meristem in the center) whose diameter rarely exceeds 3 or 4 cm, serve and differentiation (the first phloem cells to anchor the vine and transport and store soil- appear to the outside of the meristem) lies just derived nutrients, whereas the small absorbing behind the cap and is followed (0.7–1.5 mm roots (“fine roots” 0.1–1 mm in diameter) are behind the tip) by the elongation zone, a short responsible for acquisition of resources such region of cell expansion and further differentia- as water and nutrients. The woody roots of tion (the first xylem cells appear; see Figure 1.4). mature vines are widely distributed, with hori- Cell division is stimulated by the hormone zontal roots exploring the soil for distances of auxin (indole-3-acetic acid and related com- up to approximately 10 m from the trunk pounds) that is delivered from the shoot tips (Smart et al., 2006; Figure 1.4). Although the via the phloem and via active cell-to-cell trans- majority of roots, especially the fine roots, are port down the vascular parenchyma and is normally concentrated in the top 0.5–1 m, roots then also recycled, or refluxed, within the root can grow to a depth of more than 30 m when tip (Friml, 2003; Kramer and Bennett, 2006). they encounter no impermeable barriers (Galet, Cell elongation and differentiation, on the other 2000; Lehnart et al., 2008; Morlat and Jacquet, hand, are promoted by a group of growth hor- 1993; Pourtchev, 2003; Viala and Vermorel, mones termed gibberellins that are produced, 1909). Indeed, grapevines are among the most upon stimulation by auxin, close to or at their deeply rooted plants, and their root biomass site of action—that is, in tissues with rapidly can range from 5 to 40 t ha 1, which may be a expanding cells (Yamaguchi, 2008). The activity reflection of the competition for water and of the meristem also requires high concentra- nutrients during the vines’ co-evolution with tions of the “cell-division hormone” cytokinin their “trellis” trees (Smart et al., 2006). (trans-zeatin and related compounds), pro- The growing root tip or apex is covered by a duced in the root tips, and low concentrations slimy root cap whose starch-containing central of the “cell-expansion hormone” gibberellin, portion is named the columella. The cap pro- whereas cell elongation requires the opposite tects the root meristem from abrasion, facili- (Wang and Li, 2008; Weiss and Ori, 2007). tates penetration of the soil, and contains Because the phloem and xylem cells differenti- gravity sensors, which guide the root down- ate behind the root tip, hormones as well as ward through the soil. As the root advances the water and nutrients feeding the dividing and encounters new regions of moisture and cells must move across the cells after leaving nutrients, the cap is continuously sloughed off the vascular tissues. and replenished from the inside, whereas the The cells produced in the meristem and polysaccharide-rich slime called mucigel or elongation zone form a central ring called the 1.3. MORPHOLOGY AND ANATOMY 23
FIGURE 1.4 Distribution of Merlot root system in a drip-irrigated vineyard (left; photo by M. Keller) and diagrammatic longitudinal section of the apical region of the root (right; reproduced from Taiz and Zeiger, 2006). endodermis, which divides the root into two primary phloem, and primary xylem and is regions—the cortex toward the outside and thus also called the vascular cylinder, whereas the stele toward the inside. The cortex is the cortex develops the hypodermis and epi- responsible for nutrient uptake from the soil dermis, which form the dermal (“skin”) tissues and storage of starch and other nutrients, (Evert, 2006; Galet, 2000). This epidermis, how- whereas the stele is responsible for nutrient ever, is short-lived, and the underlying hypo- transport up and down the plant (see Chapter dermis forms the boundary between root and 3.3). The stele differentiates into the pericycle, soil even a few millimeters behind the root tip 24 1. BOTANY AND ANATOMY
(Storey et al., 2003). The endodermis consists of consisting of interactions between auxin, a single layer of cells with thickened radial and ethylene, cytokinin, brassinosteroid, and ABA transverse cell walls, named Casparian strips or regulates root architecture in response to Casparian band, impregnated with suberin and environmental (mainly water and nutrient lignin (see Chapter 3.3). The pericycle, rather availability) and developmental influences than the apical meristem as in the shoot, (Osmont et al., 2007). beneath the endodermis gives rise to the lateral Several millimeters behind this region fol- roots by renewed cell division, following dedif- lows the absorption zone, which is densely ferentiation of pericycle founder cells, where covered by root hairs. The colorless root hairs the meristem transitions into the elongation are long protrusions of epidermal cells (epider- zone (Mapfumo et al., 1994a; Nibau et al., mis cells that form root hairs are termed tricho- 2008; Osmont et al., 2007; Viala and Vermorel, blasts) and are mostly responsible for water 1909). Lateral root growth seems to be initiated and nutrient uptake (Gilroy and Jones, 2000; by shoot-derived auxin and aided by locally Pratt, 1974). They are very thin (10–15 mmin produced growth hormones termed brassinos- diameter but can flatten to squeeze through soil teroids (which resemble animal steroid hor- pores <2 mm) and can make up more than 60% mones) and small amounts of gaseous of the root’s surface area, which greatly ethylene (C2H4), which is released from the dif- increases the contact surface between root and ferentiating xylem. Apparently, the incoming soil, or the plant–soil interface, and the auxin induces this ethylene release, which in exploited soil volume (Clarkson, 1985; Sonder- turn blocks further auxin movement, and the gaard et al., 2004; Watt et al., 2006). As the root resulting local auxin accumulation then grows, new root hairs constantly grow behind induces lateral root growth adjacent to the pri- the elongation zone while the older ones are mary xylem (Aloni et al., 2006b). On the other worn away along with the epidermis, the origi- hand, higher concentrations of ethylene and nal cortex, and the endodermis. Hence, the cytokinin moving up from the root tips inhibit absorption zone advances along with the root lateral root formation (Ivanchenko et al., 2008; tip, leaving behind the conducting zone that Osmont et al., 2007). The fact that cytokinin con- continues into the trunk and all aboveground centration is highest closest to the site of its organs of the vine. production ensures that lateral roots are not Procambial cells generate both the primary formed too close to the tip, which would inter- phloem and the primary xylem and then fere with its continued growth (Aloni et al., develop into the vascular cambium, which 2006b). Moreover, the “dormancy hormone” serves as a lateral meristem and forms a contin- abscisic acid (ABA) seems to control the activa- uous sleeve of one to several cell layers within tion of the new lateral root meristem after the root–shoot axis. Once elongation growth emergence of the lateral root from its parent has ceased, the cambium forms new, secondary root: A high ABA concentration (e.g., due to phloem cells toward the outside and secondary water stress or high nitrogen availability) inhi- xylem cells, which collectively form the wood, bits meristem activation and keeps the lateral toward the inside (Viala and Vermorel, 1909). root in a dormant state (De Smet et al., 2006; The cambium is therefore responsible for the Malamy, 2005). Once activated, auxin usually radial growth of a root and moves outward as maintains the cell-producing activity of the the root grows thicker. Its cells can be regarded meristem. Thus, the meristem and elongation as “stem cells,” and the transcellular flow (i.e., zones together form the zone of active root cell-to-cell movement across the cell mem- growth in length, and a signaling network branes and cell walls) of auxin from the shoot 1.3. MORPHOLOGY AND ANATOMY 25 tip toward the roots in addition to cytokinin along with the phloem flow; van Bel, 2003) are produced in the root tips and physical pressure the conduits for assimilates, whereas the stimulate cambial activity and subsequent companion cells (which are crammed with cell differentiation (Aloni, 2001; Aloni et al., mitochondria) carry out the metabolic func- 2006b; Dengler, 2001; Ye, 2002). The cambium tions abandoned by the sieve elements and act becomes dormant in late summer and is reacti- as transfer cells loading sugar and other materi- vated in spring by root-derived cytokinin in als from and to the surrounding parenchyma response to auxin released by the swelling buds cells into and out of the sieve elements. (Aloni, 2001). Simultaneously, a cork or bark To this end, pairs of companion cells and cambium named phellogen forms in the pericy- sieve elements (which arise from a common cle, which later generates cork (phellem) mother cell; see Viala and Vermorel, 1909) are toward the outside and secondary cortex (phel- interconnected through numerous “lifelines” loderm) toward the inside (Mullins et al., 1992; termed pore/plasmodesma units, and the com- Pratt, 1974). Because of the sloughing off of panion cells and parenchyma cells are connected the outermost cell layers, the mature root con- through plasmodesmata (membrane-lined chan- sists of cork on the outside, cork cambium, sec- nels connecting plant cells). The sieve elements, ondary cortex, phloem, vascular cambium, and with a diameter up to approximately 35 mm (Viala xylem at the center (Viala and Vermorel, 1909). and Vermorel, 1909), have thicker cell walls than The deposition of the waxy, hydrophobic the adjacent parenchyma cells and are stacked (water-repellent) biopolyester suberin in the end-to-end and interconnected by large (up to cork cell walls and the oxidation of phenolic 1 mm) sieve pores in the end walls (named sieve compounds from dying cells give these roots a plates) to form the sieve tubes (van Bel, 2003). In brownish color. contrast to the xylem cells, they retain The phloem, vascular cambium, and xylem their plasma membrane, which is necessary to together are called the vascular tissue, or vascu- generate osmotic gradients for assimilate lar system, which forms discrete vertical strands transport in a direction opposite to transpiration named vascular bundles, separated by rays con- (see Chapter 5.1). sisting of parenchyma cells that are rich in nutri- The xylem (Greek xylos ¼ wood) consists of ent reserves (e.g., starch and proteins) and vessel elements, tracheids, fibers, and xylem tannins (Currle et al., 1983; Stafford, 1988; Viala parenchyma. The vessel elements and tracheids and Vermorel, 1909). The phloem and xylem form strong secondary cell walls that become are placed in parallel within each vascular bun- heavily lignified by the incorporation of lignin dle (a pattern termed collateral vascular bundle) (Latin lignum ¼ wood) and together are called and are interconnected by the parenchyma rays, the tracheary elements. These are arranged which allows for the transfer of nutrients longitudinally and embedded in the fibers and between the two “circulatory systems.” The parenchyma, the latter forming radial rays phloem (Greek phloios ¼ bark) is composed of that separate bundles of tracheary elements (living) sieve elements, companion cells, and (Figure 1.5). The tracheary elements are the phloem parenchyma. In a unique division of actual xylem conduits or water pipes of the vine. labor, the hollow and almost “clinically” dead Before they can assume their function, however, sieve elements (before they become functional, these cells must die; their organelles and cyto- they resorb their nucleus and vacuole and only plasm are disassembled, digested, and removed retain their cell membrane, a thin layer of cyto- in an orderly manner termed programmed cell plasm, plastids, and a few enlarged mitochon- death, and the dead cells become the wood (Jones dria, tied together so they are not dragged and Dangl, 1996; Plomion et al., 2001; Turner et al., 26 1. BOTANY AND ANATOMY
FIGURE 1.5 Cross sections of 1-year-old (left), 2-year-old (center), and 3-year-old V. vinifera root (right). Reproduced from Viala and Vermorel, 1909).
2007). Their thickened secondary cell walls perforation plates of the large vessel elements provide mechanical strength to prevent the pipes formed early in the growing season are simple from collapsing under the negative pressure (i.e., have a single hole), whereas those of the generated by transpiration (Ye, 2002; see also vessels formed later (latewood) are sometimes Chapter 3.3). The lignification also renders the scalariform (i.e., have several elongated holes that cell walls relatively waterproof, although lignin make the remaining cell wall appear like rungs of is not completely impermeable to water. The fiber a ladder). The much rarer tracheids, which look cells also commit suicide in a tightly controlled much like vessels without perforation plates, manner after the death of the tracheary elements occur in the latewood and, because they are single (Courtois-Moreau et al., 2009) so that the only cells, are much shorter (1–10 mm) than the live cells in the mature xylem are those of the vessels; many vessels in grapevines are more than parenchyma, and even these cells die after a few 100 mm long, and some exceed 1 m. Both vessels years (Plomion et al., 2001). These cells serve as and tracheids are radially interconnected storage sites for water and nutrients such as car- (although not with each other) through pairs of bohydrates and proteins or amino acids (Zapata porous (water-permeable) cell wall depressions et al., 2004), contain raphide crystals to deter named pits. Whereas the connections between herbivores, and are involved in the defense individual vessel elements within a vessel are against pathogen penetration. The thick ( 4 mm) provided by either pits or perforation plates, cell walls of the very long ( 0.5 mm) and thin those between one vessel and another or between ( 20 mm) fibers give the root mechanical support. neighboring tracheids are provided by pits (De Multiple vessel elements are stacked end-to- Boer and Volkov, 2003; Sun et al., 2006). The pits end and interconnected by perforation plates are generally scalariform and bordered; bordered (holes created by targeted removal of the cell pits are pores with overarching secondary cell walls and membranes) in the end walls to walls with a narrower aperture or pit channel form long hollow columns termed vessels. The opening into a wider pit chamber. The two 1.3. MORPHOLOGY AND ANATOMY 27 adjoining vessels or tracheids are separated (so-called spur pruning, which generally inside the chamber by a permeable “membrane,” retains one to three buds per spur) or longer which consists of the middle lamella and primary canes (so-called cane pruning, which typically cell walls. Although not a true cell membrane retains eight or more buds per cane). Both spur because it does not contain a lipid bilayer, the fine and cane pruning are usually done manually, pore size ( 5 nm) of pit membranes tends to slow which permits maximum control over both water flow but also limits the spread of air bub- node number and position on the vine. bles (see Chapter 3.3) and pathogens. Because of Cordon-trained grapevines may also be pruned their more open construction, vessels can conduct by machine, leaving a range of shorter and lon- water more efficiently than can tracheids. Water ger spurs centered on the cordon; this is also also moves laterally through pits (more accu- called a box cut. Machine pruning is sometimes rately through pit pairs because the pits of the done as a prepruning operation only, followed two cells line up to connect the cells), both by manual touch-up to control bud numbers. between parallel vessels or tracheids, and from Alternatively, very light mechanical pruning, and to the surrounding parenchyma cells, but which trims off only the ends of the shoots or not to fibers. Due to their length ( 0.5 mm) and canes, may be done in summer and/or winter, large diameter (10–150 mm), individual vessel ele- a method termed minimal pruning (Clingeleffer, ments can be surrounded by and connected with 1984; Possingham, 1994). This practice relies on many dozens of the much smaller parenchyma the vines’ self-pruning ability, in which the cells (Sun et al., 2006; Viala and Vermorel, 1909). immature apical portion of the shoots generally dies back and falls off in winter (see Chapter 1.3.2. Trunk and Shoots 5.1). An assortment of combinations of training systems, pruning methods, and other cultural The aboveground or aerial axis of the grape- practices are employed in commercial viticul- vine, which comprises the shoots, arms, and ture. They all have as their common goals to trunk, is called the stem by botanists. It pro- constrain the natural vigor of the vines and vides support for the growing vine and is maintain them at a small, manageable size; to responsible for water, nutrient, and assimilate sustain their shape and productivity over transport. The flexible stem (in addition to the the several decades of a vineyard’s life; to opti- roots) also serves as storage organ for carbohy- mize fruit production and quality depending drates and other nutrients, which support early on the intended end use of the grapes; and growth in spring and during stress periods, to facilitate labor and permit various degrees and for water. Because of their liana natures, of mechanization. cultivated grapevines typically require a trellis Phyllotaxy (Greek phyllon ¼ leaf, taxis ¼ order) system for support, unless they are trained describes the regular arrangement of leaves in very close to the ground. The trunk is often space or the pattern of lateral organs on the extended along a horizontal wire to form one shoot, whereby the points at which the lateral or more permanent arms or cordons that sup- organs connect to the shoot are called nodes. In port the 1-year-old wood, which in turn gives vines grown from seeds, the epicotyl (embryonic rise to the fruiting shoots. Such vines are said shoot) emerges from the embryo, forming a to be cordon trained, whereas vines without primary shoot without tendrils, and leaves cordons are said to be head trained. The shoots, arranged in a spiral fashion, with leaves which are called canes after they have matured offset by 137.5 (the so-called golden angle), and the leaves have fallen off, are generally displaying 2/5 (sometimes 3/7) phyllotaxy pruned back in winter to either short spurs (Galet, 2000; Viala and Vermorel, 1909). This 28 1. BOTANY AND ANATOMY means that two revolutions around the shoot The apical meristem is the site of cell division, must be made to find a leaf that is located where all shoot organs are initiated (in contrast directly above the reference leaf, which is five to the root apical meristem that does not form leaves below. This juvenile phase generally ends lateral roots) and where the pattern of the when 6–10 leaves have developed and the vine shoot system is established. It consists of three enters adult development (Mullins et al., 1992). subpopulations of cells: the central zone at the All new leaves are now being produced tip and the rib and peripheral zones below it alternately—that is, on two opposite sides of the (Clark, 1997; Evert, 2006; Kerstetter and Hake, shoot with a single leaf at each node (distichous 1997). The central zone with very small, rela- or 1/2 phyllotaxy with 180 angles)—and the first tively slowly dividing cells with small vacuoles, tendrils are formed. In vegetatively propagated located at the apex of the apical meristem, serves vines or adult plants, primary shoots usually as a source of cells (“stem cells”) for the originate from the buds at nodes of woody canes, other two zones, which are involved in cell dif- and their leaves always display 1/2 phyllotaxy. ferentiation. The rib zone at the base of the apical Shoot growth from buds is sometimes divided meristem forms the vascular and other tissues in into “fixed” growth and “free” growth (Mullins the central part of the shoot. The peripheral et al., 1992). Fixed growth occurs from leaf pri- (or morphogenetic) zone with small, rapidly mordia and compressed internodes preformed dividing cells with small vacuoles surrounds during the previous growing season that the central zone on the flanks of the apical overwinter in the dormant bud. It is responsible meristem and produces the outer shoot tissues for the rapid growth of the first 6–12 leaves in and the lateral meristems that give rise to the spring; the number increases with increasing primordia, which in turn develop into the lateral temperature during bud development (Buttrose, organs (Figure 1.6). In the great majority of 1970a; Morrison, 1991). Free growth, on the species of the family Vitaceae (including all other hand, occurs later in the season from the species of economic importance), there are two production of new leaf primordia and internodes types of lateral meristems: One is responsible in the shoot’s apical meristem. for leaf formation and the other for inflorescence
FIGURE 1.6 Diagrammatic longitudinal section of Concord shoot tip (left; reproduced from Pratt, 1971, reprinted by permission of AJEV); lateral organs arising from various positions in the dormant bud, illustrating the repeating three-node pattern unit of the shoots of many Vitis species (center; reproduced from Carmona et al., 2002); and chimeric Cabernet Sau- vignon shoot (right; photo by M. Keller). 1.3. MORPHOLOGY AND ANATOMY 29 and tendril formation (Carmona et al., 2007; the soil surface and leads to the formation of Gerrath and Posluszny, 2007). The various zones colorless, etiolated shoots that elongate rapidly overlap with another structural feature of the under the influence of gibberellin in search of apical meristem—the horizontal layering of cells light (Hartweck, 2008). The initial set of cells divided into the tunica (which usually consists from which all subsequent cells of a lateral of two parallel cell layers, the outermost of organ are derived are termed “founder cells” which generates the epidermis) on the outside or “anlagen” (German Anlage ¼ conception, and the corpus on the inside (Evert, 2006; layout, investment) that result in a primor- Thompson and Olmo, 1963). Because the cells dium. The lateral organs include leaves, ten- of the tunica divide in an anticlinal plane drils, clusters (inflorescences), and axillary (perpendicular to the surface) and those of the buds. Axillary buds contain meristem primor- corpus in different directions, the two layers dia that are called secondary, axillary, or lateral perpetuate separate cell lineages (Clark, 1997; meristems and form several compressed phyto- Kerstetter and Hake, 1997). Mutations occurring mers before becoming dormant; they can be in only one of the two layers can therefore give reactivated later to produce lateral shoots, rise to chimeric plants (see Figure 1.6) and to whose leaves form at right angles to those of clonal variation within grape cultivars (see the main shoot (Gerrath and Posluszny, 1988a; Chapter 1.2). McSteen and Leyser, 2005, Pratt, 1974; In contrast to the mature cells of the pith, Figure 1.7). Likewise, the inflorescence primor- rays, and cortical parenchyma (Kriedemann dia contain their own flower meristems that and Buttrose, 1971), the meristem cells do not later produce the grape flowers and their contain chlorophyll, so they cannot perform different floral organs, which are themselves photosynthesis and must import all carbon modified leaves—a concept originally pro- and other nutrients. The “installation” of the posed by the German writer and philosopher photosynthetic machinery begins as soon as Johann Wolfgang von Goethe (1790). Contrary the cells are released from the meristem, unless to their influence in the root tip, cytokinins those cells remain in darkness. This occurs, for stimulate cell division in the shoot apical meri- example, in a trunk sucker arising from below stem and are necessary for the transition from
FIGURE 1.7 Repeating three-node pattern of a Syrah shoot (left); mistakes do happen in nature—three consecutive ten- drils on V. vinifera shoot (center); and dormant bud and lateral shoot in a leaf axil (juncture between petiole and shoot) of a Malbec main shoot (right). Photos by M. Keller. 30 1. BOTANY AND ANATOMY undifferentiated cells to differentiated pri- the other members of the family leave a mordia (Dewitte et al., 1999; Mok and Mok, “blank” at every third node and how they keep 2001; Werner et al., 2003). count is still mysterious. The unusual position Just how the vine “knows” when and where of clusters and tendrils opposite leaves is a to initiate the various lateral organs is still result of the lack of elongation of the internodes somewhat of a mystery. Local accumulation of separating nodes that bear only rudimentary auxin and perhaps cytokinin seems to be the leaves termed bracts from those that bear true signal for organ formation by activating the cell leaves (Morrison, 1991; Tucker and Hoefert, wall-loosening protein expansin (see Chapter 1968). The first (i.e., lowest) two or three nodes 3.1), and the identity of each lateral organ usually carry only leaves, the next two nodes may be determined by hormone gradients are generally the ones with clusters, followed (Reinhardt et al., 2000; Wang and Li, 2008). by a node with a leaf only and then a succes- Once initiated, the primorida become sinks for sion of repeating mirror-images of three-node auxin, draining auxin from surrounding cells units with the clusters replaced by tendrils and thereby preventing their growth, which (see Figure 1.6). Some cells do not become pri- ensures the regular, phyllotactic positioning of mordia and will instead develop into the inter- lateral organs (Benjamins and Scheres, 2008; nodes of the shoot axis. These cells differentiate Berleth et al., 2007; Kuhlemeier, 2007). Meris- into the epidermis, cortex, endodermis, cam- tems and leaves are in constant communication bium, phloem, and xylem (Galet, 2000; Viala and clearly influence each other; preexisting and Vermorel, 1909). The repeating units are leaves are required for the correct positioning called phytomers, and each phytomer consists of a new leaf, possibly by defining the routes of a node, an internode, a leaf, and an axillary of auxin transport to the meristem (Piazza bud containing an axillary meristem for et al., 2005; Reinhardt et al., 2000). The older branching (McSteen and Leyser, 2005). leaves even influence specific traits of newly The epidermis develops a waxy cuticle developing leaves, such as leaf size and the (Latin cutis ¼ skin) on its outer cell walls as a density and size of stomata, so that the unfold- protective layer of all aboveground organs. ing leaves are adapted to the environment into The epidermis also has stomata, various types which they are “born” (Lough and Lucas, of hair, or trichomes (Greek trichos ¼ hair), 2006). Moreover, the axillary buds are initially and, on young organs, often small and short- not connected to the vascular bundles of the lived translucent structures variously named shoot; this connection is established at the time pearls, pearl glands, pearl bodies, or sap balls of bud outgrowth. It appears that the buds (Pratt, 1974; Figure 1.8). Pearls also develop on release auxin just before they break, and this tendrils, inflorescences, petioles, and the abax- auxin induces differentiation of vascular tissues ial surface of leaves, and they are especially that connect the buds (and hence the growing frequent under warm, humid conditions and lateral organs) to the preexisting vascular net- high nutrient availability promoting vigorous work (Aloni, 2001). growth (Paiva et al., 2009; Viala and Vermorel, The production of leaf-opposed tendrils and 1909). Although they remain covered by a thin clusters appears to be unique to the Vitaceae epidermis that sometimes develops a stoma, family and is typically discontinuous; that is, they are formed of highly vacuolated, thin- two of every three nodes bear a tendril (Gerrath walled subepidermal cells accumulating and Posluszny, 2007; Pratt, 1974). One notable sugars, oils and, perhaps, proteins. Pearls have exception is V. labrusca, which has a continuous no glandular activity but instead are also pattern—that is, a tendril at every node. Why termed food bodies because they ostensibly 1.3. MORPHOLOGY AND ANATOMY 31
bundles is probably induced by acropetal (toward the tip) polar, active transcellular movement of auxin (at a rate of 1cmh 1), perhaps in the epidermis, from developing leaf primordia and young leaves (Benjamins and Scheres, 2008; Berleth et al., 2007; Woodward and Bartel, 2005). In addition, gibberellin imported via the phloem and/or produced by the growing xylem cells may regulate xylem cell differentiation and elongation (Israelsson et al., 2005). As in the root tip, the first phloem cells appear before the first xylem cells, providing a conduit for the import of carbohy- drates and other material at the earliest stages of leaf development. Each new leaf generated by the apical meristem is connected with pre- existing vascular bundles in the shoot by, normally, five divergent bundles called leaf FIGURE 1.8 Ant feeding on pearls on a Chardonnay traces (Figure 1.9). Most leaf traces arise three shoot. Photo by M. Keller. nodes below their point of departure to a leaf, whereas tendril and axillary bud traces usually begin one node below their point of departure attract and feed ants, which in turn help defend (Chatelet et al., 2006; Gerrath et al., 2001). The the vulnerable young plant organs against shoot vascular bundles that continue their herbivorous insects (Paiva et al., 2009). course through the next internode and ensure The cortex consists mostly of parenchyma the continuity of the vine’s vascular system cells that contain chloroplasts, starch, calcium are termed sympodial bundles. Their xylem oxalate crystals, and phenolic compounds vessels are connected with the vessels of the (including anthocyanins on the sun-exposed leaf traces via intervessel pits, whereas the leaf side of the shoots of many varieties). Some traces ensure continuous vessels between the parenchyma cells below the epidermis differen- shoot and the leaf blade. The vessels terminate tiate into strands of collenchyma with thick cell in the leaf at approximately 50–60% of the leaf walls of cellulose, strengthening the growing length; the xylem conduits beyond this limit shoot. The endodermis is also called the starch are tracheids rather than vessels (Chatelet sheath because of its high concentration of et al., 2006). amyloplasts. These are specialized plastids res- The procambium develops into the vascular ponsible for starch production from imported cambium, which is divided into intrafascicular sucrose and its storage in starch grains or gran- cambium and interfascicular cambium. The for- ules (Emes and Neuhaus, 1997; Martin and mer produces secondary phloem toward the Smith, 1995; Pratt, 1974). In addition, paren- outside and secondary xylem toward the inside, chyma cells also accumulate and store protein and the latter forms parenchyma cells that reserves inside specialized protein storage become the rays separating the conducting tis- vacuoles, which are disassembled into trans- sues or vascular bundles (see Figure 1.9). As in portable amino acids during remobilization the root, the cambium is responsible for radial in spring. The differentiation of the vascular growth, but in contrast to its action in the root, 32 1. BOTANY AND ANATOMY
FIGURE 1.9 Cross section of V. vinifera shoot prior to periderm formation (left) and of cane after two layers of phloem have been discarded (center), and longitudinal section through Vitis shoot (right: A, pith; B, stele; C, cortex; D, diaphragm). Reproduced from Viala and Vermorel (1909). cytokinin stimulates this cambial activity in the time the grapes begin to ripen). Annual rings shoot (Werner et al., 2003). Cytokinins are pro- are composed entirely of xylem cells because duced in the root tips and unfolding leaves and the vascular cambium moves outward with the are transported upwards in the xylem and growing trunk, leaving only one live ring of downwards in the phloem (Aloni et al., 2005; phloem at the end of each growing season, Nordstro¨m et al., 2004). Cambium activity which is reactivated in spring (Davis and Evert, results in annual rings of secondary xylem; 1970; Esau, 1948). The xylem remains functional when we count annual rings, we are counting for several years; older, dysfunctional xylem rings of xylem. The annual rings of grapevines then becomes the heartwood. are called diffuse-porous (Mullins et al., 1992; Some parenchyma cells form outgrowths Sun et al., 2006) because the size of the xylem into the lumen of tracheary elements, entering cells produced in spring is similar to that of via the interconnecting pits; such outgrowths those produced in summer, before the cambium are termed tyloses (Greek tylos ¼ knob, knot) activity ceases (usually at approximately the and serve mainly to seal injured xylem (e.g., 1.3. MORPHOLOGY AND ANATOMY 33 due to summer or winter pruning or to freez- and is accompanied by deposition of starch in ing), perhaps to prevent entry of pathogens the xylem and phloem parenchyma cells (Eifert (Evert, 2006; Sun et al., 2006; Viala and et al., 1961; Plank and Wolkinger, 1976). It ends Vermorel, 1909). Within the ring of vascular with the death and abscission (Latin abscissio ¼ bundles there is a core of dead, unsclerified separation) of the apical meristem and sealing parenchyma cells called the pith. Some second- of the sieve pores and plasmodesmata inside ary phloem cells also form a cork cambium the sieve tubes by callose for overwintering (phellogen), which later produces cork (phel- (Davis and Evert, 1970; Esau, 1948; Viala and lem) toward the outside and secondary cortex Vermorel, 1909). Like cellulose (b1!4-glucan) (phelloderm) toward the inside (Galet, and starch (a1!4- and a1!6-glucan), callose 2000; Pratt, 1974). Phellem, phellogen, and (b1!3-glucan) consists of long chains of glu- phelloderm together form the outer bark cose molecules. The callose is degraded in the or periderm, whose multilayer combination of following spring, restoring the transport func- dead but elastic and heavily suberized cork tion of the phloem for another growing season cells and waxes, by protecting the shoot (and (Pratt, 1974). The reactivated phloem from the later the cane, cordon, and trunk) from wound- previous season is important during the early ing as well as water and nutrient loss, not only growth period, before it is replaced with newly takes over the function of the epidermis/cuticle formed phloem and discarded around mid- it replaces but also provides some insulation season, concurrent with periderm formation against temperature fluctuations (Franke and (see Figure 1.9). Depending on the species, the Schreiber, 2007; Lendzian, 2006). The cork layer vascular cambium resumes its function from is interspersed with pores called lenticels approximately 2 weeks (V. vinifera) to 2 months that, it seems, assume the role of the vanished (V. riparia) after the phloem has been reacti- stomata, permitting continued gas exchange vated, and a new phellogen is formed (Davis between the interior tissues and the atmo- and Evert, 1970; Esau, 1948). This cambium sphere. Additional protection, especially reactivation (early cell divisions), which is against fungal infection, arises from the incor- induced by auxin, is extremely fast and occurs poration of phenolic compounds termed stil- almost concurrently in the canes and trunk, benes (mainly resveratrol and ¼ E-viniferin), before any bud swelling can be observed which is one of the reasons woody tissues (Aloni, 2001). The new phellogen isolates most decay only very slowly (see Chapter 7.5). of the secondary phloem produced in the pre- All tissues outside the vascular cambium are vious growing season, and the dead phloem, collectively referred to as the bark; these together with the old periderm, is eventually include epidermis, cortex, and phloem. The sloughed off as strips of dead bark (Davis and periderm, along with the outermost layers of Evert, 1970; Esau, 1948). The tissue layers that secondary phloem, dies and turns brown (from are eliminated annually (i.e., the dead parts oxidation of cellular components) starting at of the bark) are collectively termed the rhyti- the base and moving up toward the tip as the dome (Greek rhytidu`ma ¼ wrinkle) (Viala and shoot becomes a cane. Thus, browning is Vermorel, 1909). caused by the death of the green cortex rather than by lignification, which occurs in all sec- 1.3.3. Nodes and Buds ondary cell walls but especially in those of the xylem cells, independently of and long before The lateral organs of the shoot are attached browning. Browning is also called shoot “matu- at nodes, which can be distinguished from the ration” (French aouˆtement from aouˆte´ ¼ mature) rest of the shoot or cane by their characteristic 34 1. BOTANY AND ANATOMY swelling due to a thicker pith and cortex arise in the axil (upper angle) between a shoot (Figure 1.10). In all Vitis species, a diaphragm and a leaf, directly above the petiole insertion consisting of hard, thickened pith cells with point (Galet, 2000; Viala and Vermorel, 1909). sclerified (Greek skleros ¼ dry, hard) cell walls All buds are formed by shoot apical meristems. divides the pith at the node. It is thought that There are three types of buds: prompt buds, the diaphragm aids in directing water and dormant buds, and latent buds. Prompt buds nutrients into the leaves. The leaf and bud are are the only true axillary buds of the primary inserted just above the node, and each leaf is or main shoot and are also called lateral buds connected to the shoot via its own vascular (Gerrath and Posluszny, 2007; Pratt, 1974). bundles inside a petiole. When the petiole falls They develop from the earliest formed (lateral) off, it leaves a leaf scar on the node. meristem and can break in the current growing Buds are young, compressed shoots (i.e., the season and give rise to secondary, so-called lat- internodes have not yet elongated; see Fig- eral shoots. Species such as V. rupestris, V. arizo- ure 1.6) enclosed in scales or bracts that are nica,orV. monticola often grow many long green in summer and turn brown during shoot laterals, which in turn can give rise to tertiary maturation. The bracts (leaflike structures with- shoots (Viala and Vermorel, 1909). By contrast, out petiole or lamina) protect the bud from des- tertiary shoots are rare in V. vinifera, and lat- iccation and freezing. In addition, long, woolly erals form mainly if the apical dominance of epidermal hairs growing on the inside of the the primary shoot is broken (Alleweldt and bracts form a down that cushions the bud and Istar, 1969), for example, by removal of the gives it a woolly appearance when it opens shoot tip (see Chapter 2.2). Although prompt (breaks). Buds are always axillary; that is, they buds are usually not fruitful, some clusters,
FIGURE 1.10 Location of the main features of a Vitis cane (left; illustration by A. Mills) and one-node Concord spur with one count node and three basal buds (right: A, spur with buds; B, cross section of basal bud; C, longitudinal section of basal bud; D, cross section of compound bud; E, longitudinal section of compound bud; reproduced from Pool et al., 1978, reprinted by permission of AJEV). 1.3. MORPHOLOGY AND ANATOMY 35 referred to as “second crop,” can grow on lat- Initially, 3 or 4 leaf primordia are produced, eral shoots, especially on those growing from each of them flanked by a pair of scales and the preformed part of the main shoot. Remov- wrapped in abundant hair (Morrison, 1991; ing the shoot tip, a cultural practice variously Snyder, 1933; Srinivasan and Mullins, 1981). named shoot tipping, topping, or hedging in The bud’s apical meristem then starts to pro- viticultural parlance, appears to promote the duce lateral meristems opposite to leaf primor- development of such second-crop clusters. dia. The first two or three of these will These clusters are typically small and ripen normally give rise to inflorescences, and the well after the clusters on the main shoot, if at remainder will become tendrils. Intriguingly, all. Many lateral shoots fail to form a proper in the majority of cultivars only the buds pro- periderm and are abscised in late fall or winter. duced from fixed growth (i.e., at the first 6–10 Dormant buds are formed later than prompt nodes), it seems, are able to initiate inflores- buds and arise from their own shoot apical cences; all buds formed during the shoot’s meristems in the axil of the first leaf (reduced subsequent free-growth phase only initiate ten- to a small, scale-like leaf called a prophyll) of drils (Morrison, 1991; Sa´nchez and Dokoozlian, lateral shoots, whether or not these shoots 2005). The inflorescence meristem produces emerge from the prompt buds (Gerrath and several branch meristems in spiral phyllotaxy, Posluszny, 1988a, 2007; Pratt, 1974; Viala and each of them subtended by a bract (Gerrath Vermorel, 1909). Some of these dormant buds and Posluszny, 1988b; May, 2004; Srinivasan will give rise to next year’s crop, but they and Mullins, 1981). The secondary buds gener- remain dormant until the following spring, ally remain small and are less fruitful. They which is why they are also called “winter often grow out when the primary bud has been buds.” A dormant bud is in fact an overwinter- damaged, for example, by spring frost or ing compound bud (also called an “eye”) insects, but also when vines have been pruned because it generally contains three separate severely so that their shoot system is out of bal- buds (see Figure 1.10)—a larger primary bud ance with the capacity of the root system. in the center flanked by two smaller secondary Latent buds are buds that remain dormant buds. The latter are also termed accessory buds or hidden for several years and often become and are often separately named a secondary part of the permanent structure of the vine. and a tertiary bud because they develop in the Even decades-old cordons and trunks still carry axils of the first and second prophyll, respec- viable latent buds that can give rise to water tively, of the shoot primordium that forms the shoots or suckers, especially following frost primary bud (Morrison, 1991; Pratt, 1974). events or on vines that have been pruned too Higher order buds have been reported, but severely (Galet, 2000; Lavee and May, 1997; these usually remain rudimentary (Zelleke Sartorius, 1968). Such shoots are usually not and Du¨ ring, 1994). Just like the shoot apical fruitful, although exceptions do occur. meristem, the apical meristem of the primary Basal buds, located at the base of a shoot or bud generates two types of lateral meristems: cane (see Figure 1.10), are formed in the shoot’s One is responsible for leaf production and the prophyll axils and thus can be prompt buds, other for inflorescence and tendril production dormant buds, or latent buds. Because the (Carmona et al., 2007). These form as many as basal internodes elongate only insignificantly, 12 leaf primordia (8–10 on average) and their these buds appear to be situated in a whorl associated prompt bud, up to several inflores- (“crown”) around the shoot (cane) base (Pratt, cence primordia, and several tendril primordia 1974; Viala and Vermorel, 1909). They are before they become dormant (see Chapter 2.2). therefore also called “crown buds” and can be 36 1. BOTANY AND ANATOMY quite numerous, although typically inconspicu- primordia is termed the plastochron (Greek ous. Although basal buds are usually less fruit- plastos ¼ formed, chronos ¼ time). We distin- ful than the other buds, this varies by cultivar: guish four types of leaves according to their Zinfandel, Sangiovese, Gamay, and Muscat of position on the shoot: cotyledons, scales, bracts, Alexandria have particularly fruitful basal and foliage leaves. Cotyledons (embryonic buds, whereas fruitfulness is very low in Viog- leaves) are preformed in the embryo and are nier, Gewu¨ rztraminer, and Thompson Seedless. the first two leaves to emerge from the embryo Of course, this has implications for the type of during seed germination. They are short-lived winter pruning; cultivars with low basal bud and fall off soon after germination. The scales fruitfulness are better suited for cane pruning, form around buds and protect them from water whereas those with high fruitfulness can be loss and mechanical injury. Bracts are small, pruned to short spurs. scalelike leaves at branch points on the stem of inflorescences and tendrils. The first two 1.3.4. Leaves leaves of a shoot grown from a bud ordinarily also develop as bracts and are separated by A leaf differentiates along with a very short internodes. Subsequent leaves corresponding node from a leaf primordium develop into mature leaves termed foliage (Latin primus ¼ first, ordiri ¼ to begin to weave) leaves, which are separated by elongated inter- produced by the apical meristem in the shoot nodes. Foliage leaves consist of lamina (blade), tip (Figure 1.11). The production of leaf primor- petiole (stalk) and, at the base of the petiole, a dia is induced by the “cell-expansion” protein pair of stipules, which are short-lived leaflike expansin, which in turn is induced by the plant structures or sheaths nearly surrounding the hormone cytokinin produced either locally near shoot’s node (Viala and Vermorel, 1909). the apical meristem or in the root tips, from The petiole (Latin petiolus ¼ little foot) is the where it is imported via the xylem (Mok and leaf stem that connects the lamina to the shoot Mok, 2001, Pien et al., 2001). The (thermal) time and contains multiple (typically 12–14) vascu- between the formation of two successive leaf lar bundles (Viala and Vermorel, 1909). Several vascular bundles are often derived from the same shoot vascular bundle (i.e., leaf trace); this branching is called anastomosis. Depending on species and cultivar, the petiole can be between 2 and 12 cm long and grows toward the light to position the leaf for optimal sunlight inter- ception (Figure 1.12). Where the petiole joins the blade, it divides to form the five main veins of the lamina. An abscission layer forms at each end of the petiole toward the end of the grow- ing season and leads to leaf fall. The basic function of the lamina (Latin lam- ina ¼ leaf) is to capture sunlight for energy FIGURE 1.11 Longitudinal section of dormant Cabernet (ATP) production and carbon dioxide (CO2) Sauvignonbud.A,shootapicalmeristem;B,lateralbudpri- for carbohydrate production to support the mordium; C, inflorescence primordium; L, leaf primordium; U, uncommitted primordium. Reproduced from Morrison, vine’s metabolism and growth (see Chapter 4). J.C. (1991). Bud Development in Vitis vinifera L. International To maximize light absorption, leaves must be Journal of Plant Sciences, The University of Chicago Press. as wide as possible, whereas to maximize gas 1.3. MORPHOLOGY AND ANATOMY 37
FIGURE 1.12 Cross section of a V. vinifera leaf with prominent main vein (left; reproduced from Viala and Vermorel, 1909), Zinfandel petiole bending toward the light (center), and pattern of Concord leaf venation (right; photos by M. Keller).
exchange, they must be as flat as possible; both evenly to and from all over the lamina. The of these properties, however, make leaves veins are enclosed by a set of thickened paren- inherently vulnerable to overheating and dehy- chyma cells termed the bundle sheaths, which dration (Tsukaya, 2006). Different species and are visible as almost transparent, projecting cultivars of grapevine differ in their leaf shape, “ribs,” especially on the abaxial side of the leaf and leaf morphology forms the main basis of (Pratt, 1974; see Figure 1.12). The bundle ampelography (see Chapter 1.1). The predomi- sheaths function to conduct water away from nant leaf form of the genus Vitis is palmate, in the xylem to the mesophyll cells and serve as which the five main vascular bundles (major “windows” transferring light to the photosyn- veins; see Figure 1.12) serve the leaf’s five lobes thetic mesophyll cells in addition to providing that in most V. vinifera cultivars are partly sepa- mechanical support to protect the leaf blade rated from each other by more-or-less deep against collapse during severe dehydration gaps termed sinuses (Mullins et al., 1992; Viala and other stresses. and Vermorel, 1909; see Figure 1.12). The main The leaf margins are serrated (toothed), and vascular bundles divide into an ever finer, elab- both the number of teeth and their sharpness orate network whose branches (minor veins) or sinus depth vary greatly among Vitis species are interconnected or eventually (at the fifth and cultivars (Figure 1.13). It is thought that the order) end blindly between the leaf mesophyll sinus depth, but not the teeth number, is cells. Interconnected veins form enclosed areas inversely related to the rate and/or duration or islets termed areoles, approximately 1 or of cell division during leaf formation so that 2 mm across, and 10 mm of minor veins services leaves with greater cell numbers should have up to 10 mesophyll cells (Wardlaw, 1990). The shallower sinuses (Tsukaya, 2006), but this vein network acts as an “irrigation/collection association has not been investigated in grape- system,” with the major veins acting as a rapid vines. The teeth “funnel” the tracheids of the supply network and the minor veins forming a terminal veins into water pores called slow distribution and collection network hydathodes, which can discharge drops of (Canny, 1993). This system delivers water and xylem sap (Evert, 2006; Pratt, 1974). This pro- nutrients, and it collects assimilates relatively cess is called guttation (Latin gutta ¼ drop) 38 1. BOTANY AND ANATOMY
which in turn determines the extent and direction of expansion of these cells by means of sugar and water import (see Chapter 3.1). As a side note, the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) is a synthetic, nontransportable auxin that stimu- lates cell division but inhibits cell expansion and differentiation, and interferes strongly with pat- terning, leading to severely distorted leaves. Thus, the leaf’s shape and the basic architecture of its vascular system are determined by genetic and hormonal controls during the initial stages FIGURE 1.13 Chardonnay leaf (left) and Merlot leaf of leaf formation. Newly formed leaf cells require (right). Illustrations by A. Mills. approximately 2 weeks to expand to their full size (in contrast to root cells, which finish expanding and should not be confused with dew, which is within a few hours), which gives leaves sufficient the formation of droplets of atmospheric water developmental flexibility to adjust their final that condenses on cool surfaces. Guttation is size to environmental conditions prevailing especially pronounced during warm nights during leaf formation. Leaf size and shape can with high relative humidity (i.e., when transpira- be markedly altered by patterns (gradients) of tion is minimal and root pressure may force water water availability across the leaf during the up the vine, see Chapters 2.2 and 3.3). So-called period of leaf expansion. The final size of a leaf scavenging cells surrounding the xylem vessels is usually closely tied to its cell number, although near their endings actively and selectively collect greater cell expansion often partially compen- xylem solutes (e.g., nutrient ions and hormones sates for lower cell number (Tsukaya, 2006). such as cytokinin) and recycle them back to the Because both cell division and cell expansion are phloem for re-export (Sakakibara, 2006), whereas dependent on sugar supply, leaves developing undesirable solutes (e.g., excess calcium) may be when a vine is photosynthetically “challenged” excreted along with the water. Thus, the will produce few and small cells (Van Volken- hydathodes may function as safety (overflow) burgh, 1999). Therefore, leaves can grow to a valves that can rid the leaves of excess water. In large size if nutrient availability is unlimited leaf primordia and unfolding leaves, the during their expansion, but they will remain hydathodes are also thought to be primary sites small if they form while environmental condi- of auxin production, which is, at least in part, tions are unfavorable. responsible for the formation of the characteristic The lamina grows by cell division and leaf shapes (patterning) and differentiation of expansion in basal and intercalar (Latin interca- vascular bundles, whereby cells that experience lare ¼ to insert) meristems and consists mainly rapid auxin flow will differentiate into veins of primary tissue; there is no secondary growth (Aloni, 2001; Aloni et al., 2003, 2006b; Benjamins to speak of (Evert, 2006). Cell division generally and Scheres, 2008; Berleth et al., 2007; Dengler, stops when the leaf is only half or less of its 2001). Auxin stimulates cell division at high con- final size so that the remainder of leaf growth centration and cell expansion at low concentra- is caused solely by the expansion of preformed tion. It may exert its growth-promoting cells, which occurs mainly due to enlargement influence via differences in its content among of their vacuoles (see Chapter 3.1). In addition cells (i.e., auxin gradients arising from active to the influence of auxin, both cell division cell-to-cell transport away from the source), and cell expansion are stimulated by cytokinins 1.3. MORPHOLOGY AND ANATOMY 39 produced by the dividing leaf cells as well as with a thin, continuous, extracellular mem- by cytokinins imported via the xylem from the brane termed the cuticular membrane or cuticle root tips. Moreover, cell expansion is also pro- (Bargel et al., 2006; Kerstiens, 1996). The cuticle moted by gibberellins. Gibberellin production consists of cutin, polysaccharides, phenolics, by expanding cells just behind the shoot apical and soluble lipids that make up the intracuticu- meristem and by the young, rapidly expanding lar wax embedded in the outer parts of the leaves and petioles near elongating internodes polymer matrix. It is covered by a layer of over- is induced by auxin (Hartweck, 2008). lapping solid, semicrystalline lipid platelets— At the same time as the leaf cells expand and the epicuticular wax. The intra- and epicuticu- the rate of cell division decreases, they also lar waxes are typically referred to as cuticular build up their photosynthetic machinery, wax, which seals the plant surface. Although namely the chlorophyll-rich chloroplasts (see waxes are complex mixtures of different lipids, Chapter 4.1). By the time the various tissues of they are essentially (>90%) made up of chains the leaf blade begin to differentiate, the leaf is of more than 20 methylene (CH2) groups that generally only six cell layers thick (Pratt, are assembled within expanding epidermis 1974). The outer surface layers differentiate into cells and then secreted to the surface along the adaxial (upper, top, and sun-exposed) and with cutin as the leaf unfolds and expands abaxial (lower, bottom, and shaded) epidermis. (Samuels et al., 2008; Scho¨nherr, 2006). The The epidermis does not contain chloroplasts self-cleaning properties of this water-repellent but can have various types of hairs called tri- or hydrophobic (Greek hydros ¼ water, phobos chomes, particularly along the veins on the ¼ fear) surface film keep the leaves clean by abaxial side (Viala and Vermorel, 1909). The greatly reducing the adhesion of water and par- number and length of trichomes varies with ticles (Bargel et al., 2006). Together with the species and cultivar and gives the underside thick outer cell walls of the epidermis, the cuti- of some leaves (e.g., Vitis candicans; Clairette, cle provides mechanical support that maintains Meunier) a mat whitish appearance, whereas the integrity of leaves and other plant organs others (e.g., Vitis monticola; Grenache) are gla- and determines their growth rate and extent brous (Latin glaber ¼ hairless, bald). Trichome of growth. This epidermis/cuticle skin also density is especially high in young leaves, but forms the first mechanical barrier to invading it decreases after unfolding due to gradual pathogens as well as physical or chemical shedding of trichomes, especially on the adax- injury by repelling fungal spores and dust par- ial side (Karabourniotis et al., 1999). Trichomes ticles, and it reduces nutrient leaching (Bargel probably help reduce water loss (by regulating et al., 2006; Kerstiens, 1996). Its most important the leaf surface temperature and increasing the function, however, is the protection of the plant boundary layer resistance; see Chapter 3.2), from desiccation. The waxy cuticle restricts deter insects (as mechanical barriers), provide gas diffusion so that most water vapor and protection from damaging ultraviolet radiation other gases must pass through openings called (due to their phenolic compounds absorbing stomata (Greek stoma ¼ mouth; Figure 1.14). and scattering radiation; see Chapter 5.2), and These small pores (15–40 mm in length) are may even afford some insulation. The outside formed by degradation of the cell wall joining walls of epidermis cells, which are much two specialized cells, the kidney-shaped guard thicker than the cell walls of interior tissues cells, which, in contrast to the surrounding (Kutschera, 2008a,b), contain cutin—a strong, epidermis cells from which they are derived, elastic biopolyester composed of highly poly- contain a few chloroplasts (Lawson, 2008). Vitis merized hydroxy fatty acids—and are covered leaves are said to be hypostomatous because 40 1. BOTANY AND ANATOMY
FIGURE 1.14 Cross section of a leaf (left; illustration by A. Mills and M. Keller) and stoma on a Chardonnay leaf, surrounded by crystalline platelets of epicuticular wax (right; photo by M. Keller). the stomata are located almost exclusively in cannot increase in length and therefore bend the abaxial epidermis. outward. The aperture (pore width) of open Grapevine leaves typically have between 50 stomata is approximately 8 mm, whereas it is and 400 stomata/mm2, with V. vinifera and almost zero when the stomata are closed, for most rootstocks at the lower end (<250 mm 2) example, at night or during severe water stress. and V. labrusca and V. cinerea at the higher The tissue between the two epidermal layers end (>300 mm 2) of this range (Du¨ ring, 1980; is called the mesophyll (see Figure 1.14). As the Liu et al., 1978; Scienza and Boselli, 1981). Inter- leaf expands, the mesophyll cells stop growing estingly, mature leaves seem to be able to before the epidermis cells do. This pulls the detect the conditions around them and send a mesophyll cells apart and leads to the forma- (still unknown) signal to the newly developing tion of a large system of intercellular spaces leaves that adjusts the number of stomata that that facilitate the diffusion of gases such as will form on those leaves. Stomata are responsi- water vapor and CO2 (Van Volkenburgh, ble for regulating the gas exchange (mainly 1999). The mesophyll consists of a single layer CO2 for photosynthesis, oxygen for respiration, of elongated cells termed palisade parenchyma and water vapor from transpiration) between and four to six layers of irregularly shaped, the leaf and the atmosphere. Their major func- loosely packed cells termed spongy paren- tion is to balance the uptake of CO2 with the chyma (Galet, 2000; Pratt, 1974; Viala and loss of water (see Chapter 3.2). Stomata open Vermorel, 1909). Both cell types contain large when the guard cells take up water and swell numbers of chloroplasts (Greek chloros ¼ light (their volume more than doubles); because they green, yellow), which are the actual centers of are attached to each other at both ends, they photosynthesis and assimilation. 1.3. MORPHOLOGY AND ANATOMY 41
The palisade parenchyma lies below the Mullins, 1978, 1980, 1981). Accordingly, the adaxial epidermis and has small intercellular absence of gibberellin production or supp- spaces. It also contains crystals of calcium ression of its activity at the site of organ initiation oxalate, which are arranged inside large cells gives rise to an inflorescence (flower cluster), (typically 50–100 mm long and 20–25 mm wide) whereas the presence of gibberellin results in a in bundles of needle-like crystals called raphides tendril. The cell-division hormone cytokinin, in and, in older leaves inside very small cells ( 10 contrast, promotes inflorescence formation mm in diameter) lining the veins, in aggregates over tendril formation (Srinivasan and Mullins, of octahedral (star-shaped) crystals called 1978, 1980, 1981). It seems likely that the druses (Evert, 2006; Viala and Vermorel, 1909). response to these hormones is quantitative The spongy parenchyma follows toward the because there is a continuum of transitional or abaxial side and has large intercellular spaces. intermediary structures that are partly tendril- Many of the mesophyll cell walls are thus like and partly inflorescence-like. Moreover, exposed to the air inside the leaf (internal the degree of differentiation depends largely on leaf atmosphere) so that the internal leaf surface the cultivar and environmental conditions is approximately 10–40 times the surface area during the time of cluster initiation and diffe- of the leaf. Water evaporates from the meso- rentiation (see Chapter 2.3). Both tendrils and phyll cell walls into these intercellular spaces, inflorescences, in turn, are modified shoots which are continuous with the outside air when (lateral branches), and there is also a variety of the stomata are open so that water vapor is dis- intermediary forms between inflorescence and charged (transpired) into the atmosphere (see shoot (Figure 1.15). The individual flowers on Figure 1.14; see also Chapter 3.2). The CO2 an inflorescence, again, are modified shoots, follows the reverse diffusion route into the leaf with the separate floral organs having evolved (see Chapter 4.2). from highly modified leaves. Similar to leaves, tendrils grow in intercalary 1.3.5. Tendrils and Clusters fashion in addition to growth in their own api- cal meristem (Tucker and Hoefert, 1968). They Tendrils and fruiting clusters of the grapevine also differentiate an epidermis with numerous are generally considered homologous on the stomata, a spongy parenchyma, and a large basis of anatomical, morphological, and physio- but short-lived hydathode at each of their tips, logical similarities. Darwin (1875) concluded which soon become suberized (Gerrath and from observations of grapevines growing in his Posluszny, 1988a; Tucker and Hoefert, 1968). backyard that "there can be no doubt that the Although two tips are most common, vigorous tendril is a modified flower-peduncle." Indeed, vines often form tendrils with three or more studies of gene expression—that is, of the manu- tips (Figure 1.16). Tendrils enable wild vines facture of RNA and protein from the segment of to access sunlight at the top of tree canopies DNA making up a gene—also suggest that ten- with a relatively small investment in shoot bio- drils are modified reproductive organs that have mass per unit height gain. The tendril’s tips been recruited (i.e., adapted during evolution) search for surrounding objects by making as climbing organs (Calonje et al., 2004; Dı´az- sweeping, rotating movements during growth Riquelme et al., 2009). As sterile reproductive (Darwin, 1875; Galet, 2000; Viala and Vermorel, structures, they are prevented from completing 1909); this oscillatory growth pattern is termed floral development by the cell-elongation circumnutation. When one of the tips detects a hormone gibberellin (Alleweldt, 1961; Boss and support (via contact-sensitive epidermis cells), Thomas, 2002; Boss et al., 2003; Srinivasan and the arms rapidly coil around the support in 42 1. BOTANY AND ANATOMY
FIGURE 1.15 Intermediary forms of Syrah inflorescence/tendril/shoot (left; inset: inflorescence that would rather be a tendril; photos by M. Keller) and structure of a grape cluster with berries removed (right; illustration by A. Mills; inset reproduced from Viala and Vermorel, 1909).
FIGURE 1.16 Vitis shoot tip showing tendrils with two and three tips (left) and tips of a tendril coiling around a trellis wire (right). Photos by M. Keller. 1.3. MORPHOLOGY AND ANATOMY 43 two directions (see Figure 1.16). Following this rays and to the inside of the vascular bundles so-called thigmotropic (Greek thigmos ¼ touch, accumulate starch and may serve as a tran- trope ¼ turn) movement, the entire tendril ligni- sient nutrient storage compartment. The epi- fies and stiffens to prevent unwinding (Braam, dermis on the exterior contains numerous 2005). Tendrils that fail to find a support die stomata and is covered with a cuticle and are abscised at the point of attachment to (Theiler, 1970). Clusters vary widely in length the shoot. from 3–5 cm in wild grape species to more The fruiting structures of the grapevine are than 50 cm in some table grape cultivars, and panicles and are known as inflorescences or this is accompanied by varying degrees of flower clusters. They are always inserted branching. The cluster architecture resulting opposite a leaf and are initially protected by from this branching is probably determined bracts covered with trichomes. After fruit set by auxin released from developing flower the inflorescence is called a cluster or bunch. meristems. It is composed of the peduncle (stalk; i.e., the entire branched axis apart from the pedicels 1.3.6. Flowers and Grape Berries of individual flowers) carrying the flowers or, following fruit set, the berries. Each flower or As discussed in Chapter 1.1, wild grapes are berry is attached to the peduncle via a pedicel dioecious, but most V. vinifera cultivars have (final branch or flower stalk) that widens into perfect (hermaphroditic) flowers. Physiologi- thereceptaclethatcarriestheflowerorberry. cally, female flowers are rare among V. vinifera The main stem or axis consists of the hypo- cultivars (e.g., Madeleine Angevine and Picolit) clade (between the shoot and the first branch- but common among M. rotundifolia cultivars. ing) and the rachis (central axis) that is made Many rootstocks have male flowers, whereas up of a main axis (also called inner arm) and some (e.g., Ramsey, 101-14 Mgt, 5BB, and 41B) a lateral wing or shoulder (also called outer have female flowers (Meneghetti et al., 2006). arm). Both arms carry multiple secondary The formation and development of grape flow- branches (and many of these tertiary branches, ers is considered in Chapter 2.3. The flowers some of which in turn carry quaternary (Figure 1.17) are green, inconspicuous, and branches), although the shoulder may be miss- small; their size varies, depending on species, ing or may be a tendril or, occasionally, a from 2 mm (V. berlandieri)to6or7mm(V. lab- shoot. Just as in tendrils, each of the arms rusca). They are made up of a receptacle (some- and branches is subtended by a bract. The times called torus) carrying five sepals (fused flowers usually occur in groups of three (with to form the calyx), five petals (with their bor- two, often slightly smaller, flowers flanking a dering cells interlocked to form a protective central, terminal flower) or five; these basic cap termed calyptra or corolla), five stamens, floral units are termed a triad, a dichasium, and the pistil (May, 2004; Meneghetti et al., or a cyme (Gerrath and Posluszny, 1988b; 2006; Srinivasan and Mullins, 1981; Swanepoel May, 2004; Mullins et al., 1992). The hypo- and Archer, 1988). The stamens are the male clade, which varies in length from 2 to 10 cm reproductive organs with bilobed anthers depending on cultivar, connects the cluster (microsporangia), each lobe carrying two pol- with the shoot, similar to the function of the len sacs containing more than 1000 pollen petiole for the leaf, and therefore also contains grains (each composed of a large vegetative cell multiple (typically approximately 30) vascular and two sperm cells), and their stems termed bundles that are separated by rays (Viala and anther filaments. The pistil is the female repro- Vermorel, 1909). The parenchyma cells of the ductive organ with a superior ovary 44 1. BOTANY AND ANATOMY
FIGURE 1.17 Diagrammatic longitudinal section of a Muscat Hamburg flower (left; from Rafei, 1941) and longitudinal section of a Vitis seed (right; illustration by A. Mills). surmounted by a short style and a papillate with undeveloped style and stigma (Caporali stigma (Considine and Knox, 1979a; Hardie et al., 2003; see also Figure 1.1). et al., 1996b; Pratt, 1971). The set of stamens of A network of vascular bundles, or traces, a flower are collectively termed the androe- supplies each of the floral organs (see Fig- cium, and the female parts together are named ure 1.17). The base of the ovary is encircled by the gynoecium. The gynoecium of grape flow- a whorl of odor glands termed osmophors but ers generally consists of two seed pockets sometimes referred to as nectaries. The ovary called carpels (megasporangia) that are fused comprises two, rarely three, cavities termed to form the pistil; the central region in the locules, each normally containing two seed pri- ovary where the two inner carpel walls meet mordia called ovules (Nitsch et al., 1960) that is termed septum, which divides the ovary, are formed by a placental wall to which they and later the fruit, into two locules. Female remain attached via a funiculus (Latin funiculus flowers are difficult to distinguish from perfect ¼ thin rope), which serves the same function as flowers; they have stamens that look normal the umbilical cord in mammals. Both the calyp- but produce sterile pollen, whereas in male tra and the pistil (after capfall) have a cuticle- flowers the pistil is reduced to a tiny ovary covered epidermis containing chloroplasts 1.3. MORPHOLOGY AND ANATOMY 45 and, through their stomata (10–20 per mm2), epidermis and later forms a cuticle on the out- are capable of their own gas exchange (Blanke side. The protective testa contains inclusions and Leyhe, 1987, 1988, 1989a). The pistil tissues of calcium oxalate crystals and is rich in pheno- are rich in calcium oxalate crystals, starch, and lic compounds such as tannins that protect the phenolic compounds, although the starch dis- seed against premature “consumption” by appears during berry development (Hardie microorganisms or insects (Dixon et al., 2005). et al., 1996a; Sartorius, 1926). The apparent folding of the integuments leaves The ovules (female gametophytes) with their a small opening (pore) termed micropyle fertilized egg (zygote) and remaining embryo through which the pollen tube enters during sac (composed of seven cells including the egg pollination (Sartorius, 1926; see also Chapter cell flanked by two synergid cells and the cen- 2.3), and water will enter and the radicle will tral cell) can develop into heart- or pear-shaped exit during seed germination. The endosperm seeds whose main role is to protect and nurture progressively expands at the expense of the the developing embryo (Roberts et al., 2002). nucellus that may serve as a food supply for Thus, a grape berry with two locules can have the embryo and endosperm early during seed a maximum of four seeds, and one with three development (Nitsch et al., 1960). locules can have six seeds. In practice, the seed The vascular bundles that previously served number is usually one or two in V. vinifera cul- the ovary give rise to a complex network of tivars but two or three in many Chinese Vitis vascular traces that supply the seed (ovular species (Wan et al., 2008c). Depending on culti- bundles) and the pericarp (ventral and dorsal var and seed number, the seeds make up bundles). The dorsal, or peripheral, vascular between 2 and 7% of the fresh weight of mature bundles of most V. vinifera cultivars lie approx- grapes (Viala and Vermorel, 1909). After fertili- imately 200–300 mm beneath a mature berry’s zation, the pistil develops into the fruit, with surface and can give some berries a “chicken- the ovary wall becoming the pericarp—that is, wire” appearance (see Figure 1.18), whereas in the flesh and skin (Figure 1.18); thus, the fruit V. labrusca the bundles are approximately twice can be viewed as a swollen or mature ovary as far from the surface (Alleweldt et al., 1981). (Gillaspy et al., 1993). Grapes are classified as The ventral, or central, vascular bundles and a berry fruit because the thick pericarp encloses their associated parenchyma cells are termed the seeds. the “brush” and remain attached to the pedicel The ovule is made up of the outer and inner when a ripe grape is plucked from the cluster. integument and the nucellus (Pratt, 1971) and Individual tracheary elements (mostly vessel the embryo sac. After fertilization, the integu- elements with some tracheids) in the berry are ments will form the seed coat termed testa, approximately 250–300 mm long with an aver- and the nucellus will surround the developing age diameter of approximately 3.5–5 mm (Cha- endosperm and the embryo (see Figure 1.17) telet et al., 2008a). Although existing tracheary with an epicotyl (embryonic shoot tip), two elements are stretched at the beginning of rip- cotyledons (embryonic or seed leaves), a hypo- ening, most of them remain intact, and new ele- cotyl (embryonic shoot base), and a radicle ments are added throughout berry (embryonic root). The epicotyl contains at its development. The pericarp consists of three tip the shoot apical meristem, and the radicle anatomically distinct tissues: the exocarp (or contains at its tip the root apical meristem, epicarp), mesocarp, and endocarp (Galet, 2000; which defines the apical–basal axis of the Viala and Vermorel, 1909). embryo and future seedling. The outer integu- The exocarp forms the grape’s dermal sys- ment consists of a palisade layer and an tem or “skin,” which, depending on thickness 46 1. BOTANY AND ANATOMY
FIGURE 1.18 Structure of a ripe grape berry (left; reproduced from Coombe, 1987, reprinted by permission of AJEV), scanning electron micrograph detail of ruptured Syrah berry surface with epicuticular wax and raphide crystals of calcium oxalate (top right; photo courtesy of S. Rogiers), and stylar end of Syrah berry covered with epicuticular wax (bottom right; photo by M. Keller).
(which varies by species and cultivar) and structure of the epicuticular wax changes with berry size, makes up between 5 and 18% of age of the berry. Although the wax is normally the fresh weight of mature berries. It is com- crystalline, the crystals appear to degrade posed of the epidermis covered with a waxy slowly over time, fusing into amorphous cuticle and the underlying outer hypodermis masses and eventually into a more-or-less (Bessis, 1972a; Considine and Knox, 1979b; continuous layer on top of the normal, uninter- Fouge`re-Rifot et al., 1995; Hardie et al., 1996b). rupted wax layer (Casado and Heredia, 2001; Epicuticular wax, often referred to as “bloom,” Rogiers et al., 2004a). The wax material is rather covers the surface of the cuticle as overlapping soft and can be altered or removed by the platelets forming a strongly hydrophobic layer impact of rain, by abrasion from wind-blown that protects the berry from water loss and particles, or by contact with other berries, forms the first barrier against pathogen inva- leaves, or other adjacent objects (Shepherd sion (Rosenquist and Morrison, 1988). The and Griffiths, 2006). However, because of its 1.3. MORPHOLOGY AND ANATOMY 47 vital importance as a desiccation protectant, a “boundary” between hypodermis and meso- damaged or removed wax layer is quickly carp, are now situated within the mesocarp regenerated, especially in young berries. The (Chatelet et al., 2008a; Fouge`re-Rifot et al., cuticle is approximately 1.5–4 mm thick, and 1995). The relative thickness of the skin the epidermis consists of a single cell layer that declines from approximately one-eighth to is approximately 6–10 mm thick (Alleweldt one-hundredth of the total berry diameter et al., 1981; Bessis, 1972a; Hardie et al., 1996b). between fruit set and maturity. Unlike the hypodermis and mesocarp cells, The mesocarp, which is commonly called the the cells of the epidermis divide in a strictly “flesh” or “pulp” of the grape berry, forms anticlinal manner (i.e., normal to the surface), soon after fruit set and consists of 25–30 layers thus maintaining their cell “lineage” (Consi- of large (100–400 mm), thin-walled, and highly dine and Knox, 1981). The initially single-lay- vacuolated storage parenchyma cells. The ered outer hypodermis (100–250 mm thick) vacuoles can make up as much as 99% of the increases to 6–10 layers of thick-walled cells cell volume in ripe grape berries (Diakou and after fruit set and then transiently even to Carde, 2001) and serve as an internal reservoir 15–20 layers. These cells contain chloroplasts storing sugars, organic acids, and nutrients and tannin-rich vacuoles, interspersed with (see Chapter 6.2). Mesocarp cells outside the amyloplasts, which form within a few days network of peripheral vascular bundles of the after fruit set but disappear during berry devel- pericarp are termed the outer mesocarp; those opment, and so-called idioblast (Greek idios ¼ toward the inside are the inner mesocarp, unique, blastos ¼ germ, sprout) cells rich in which, at maturity, makes up almost two-thirds bundles of needle-like calcium oxalate crystals of the berry volume. At maturity, the mesocarp called raphides (Evert, 2006; Fouge`re-Rifot cells are approximately 75 times larger than the et al., 1995; see also Figure 1.18). As an aside, skin cells, and their radial width is approxi- a cell may contain more than one kind of vacu- mately 10 times that of the skin cells, whereas ole, so it is possible that different components the cell walls in the skin are almost 20 times may be stored in specialized vacuoles within thicker than those in the mesocarp. It is no the same cell (Marty, 1999; Mu¨ ntz, 2007). The wonder, then, that the skin sets the upper limit outermost hypodermal cells become stretched to which a berry can expand. and radially compressed, and their cell walls The innermost tissue of the pericarp, the thicken approximately 10-fold as the berry endocarp, surrounds the seeds and consists of expands (Considine and Knox, 1979b; Hardie the inner hypodermis with cells rich in star- et al., 1996b). The innermost cells expand and shaped calcium oxalate crystals (druses) and seem to be progressively converted to meso- also the inner epidermis (Fouge`re-Rifot et al., carp cells. Thus, the skin becomes thinner (4 1995; Hardie et al., 1996b). The endocarp is or 5 cell layers) in the ripening berry and the often difficult to distinguish from the mesocarp vascular bundles, which initially marked the in grapes (Mullins et al., 1992). CHAPTER 2
Phenology and Growth Cycle
OUTLINE
2.1. Seasons and Day Length 49 2.3. Reproductive Cycle 68 2.2. Vegetative Cycle 53
2.1. SEASONS AND DAY LENGTH of cultural practices as part of vineyard man- agement (Dry and Coombe, 2004). Phenology (Greek phainesthai ¼ to appear, Grapevines, like other plants, monitor the logos ¼ knowledge, teaching) is the study of seasons by means of an endogenous “clock” natural phenomena that recur periodically in (Greek endon ¼ inside, genes ¼ causing) in plants and animals and of the relationship of order to prevent damage by unfavorable envir- these phenomena to climate and changes in onments. Every plant cell contains a clock (per- season. In other words, it is the study of the haps even several clocks), and the clocks of sequence of plant development. Its aim is to different cells, tissues, and organs act autono- describe causes of variation in timing by seek- mously (McClung, 2001, 2008). The clock is ing correlations between weather indices and driven by a self-sustaining oscillator consisting the dates of particular growth events and the of proteins that oscillate with a 24-h rhythm in intervals between them. Phenology investigates response to light detected by phytochromes a plant’s reaction to the environment and (see Chapter 5.2) and other light-sensitive pro- attempts to predict its behavior in new environ- teins that translate the light signal into a clock ments. In viticulture, phenology is mainly input that uses this time-encoded information concerned with the timing of specific stages of to regulate physiological functions (Fankhauser growth and development in the annual cycle. and Staiger, 2002; Spalding and Folta, 2005). Such knowledge can be used for site and culti- Additional input is provided through feedback var selection, vineyard design, planning of loops that arise from various compounds pro- labor and equipment requirements, and timing duced by the plant’s metabolism (McClung, 2008).
The Science of Grapevines 49 Copyright # 2010 Markus Keller. Published by Elsevier Inc. All rights reserved. 50 2. PHENOLOGY AND GROWTH CYCLE
Just how plants manage to integrate and drought (see Chapter 7.2), nutrient deficiency or transform this environmental and metabolic excess (see Chapter 7.3), or infection by patho- information into daily and seasonal functions gens (see Chapter 7.5). In general, higher and how they avoid being “fooled” by bright temperatures accelerate plant development moonlight is still mysterious, but it could and advance grapevine phenology. Therefore, involve fluctuations in calcium concentration one of the most conspicuous consequences of þ of the cytosol. Cytosolic Ca2 also oscillates climate change associated with the current and with a 24-h period with a peak before dusk, future increase in temperature due to the man- and it participates in the “translation” of many made rise of atmospheric CO2 is a shift of signals, both internal and external (Hotta et al., phenological stages to earlier times during the 2007; McClung, 2001; Salome´ and McClung, season (Chuine et al., 2004; Wolfe et al., 2005). 2005). In contrast to the temperature depen- Because day length during the growing sea- dence of most biochemical processes, the son is longer at higher latitudes, the effects of period of the rhythm is temperature compen- day length and latitude on the internal clock sated so that it remains constant at 24 h over are very similar (Salome´ and McClung, 2005). a wide range of temperatures. The gradual In tropical climates, with little or no change in shift in time of sunrise (actually the change day length (and temperature at lower eleva- between dawn and dusk) as the season pro- tions) during the year, grapevines behave as gresses serves to reset the phase of the clock evergreens with continuous growth and strong every day and enables plants to “know” when apical dominance, and often continuous fruit the sun will rise even before dawn (Fankhauser production, throughout the year (Dry and and Staiger, 2002). This synchronization of the Coombe, 2004; Mullins et al., 1992). With appro- circadian (Latin circa ¼ about, dies ¼ day) clock priate pruning strategies (and sometimes delib- to the external environment is termed entrain- erate defoliation by hand, often in combination ment. It translates day length (i.e., photoperiod) with water stress and sprays of ethephon or into a vine’s internal time as an estimate of both other chemicals to induce near dormancy, fol- the time of day and the time of year, enabling lowed by applications of hydrogen cyanamide it to anticipate and prepare for daily and sea- to induce uniform budbreak), two crops can sonal fluctuations in light and temperature. For often be harvested each year in the tropics. In example, the genes responsible for “building” temperate climates (and at high elevations in enzymes involved in the production of ultravio- the tropics), where winter conditions prevent let (UV)-protecting phenolics are most active the survival of leaves, grapevines have a dis- before dawn, photosynthesis-related genes continuous cycle with alternating periods of peak during the day (and opening and closing growth and dormancy. Under these conditions, of the stomata can anticipate dawn and dusk), the time of active growth generally occurs from whereas genes associated with stress are March/April to October/November (Northern induced in late afternoon (Fankhauser and Hemisphere) or September/October to April/ Staiger, 2002; Hotta et al., 2007). May (Southern Hemisphere). Several distinct On a seasonal scale, flowering, onset of developmental stages or key events have fruit ripening, bud dormancy, leaf senescence, been identified (Figures 2.1 and 2.2) and have and cold acclimation are typical responses to been given a variety of names. These stages day length (so-called photoperiodism), include dormancy, budbreak (budburst), bloom although each of these developmental pro- (anthesis or flowering), fruit set (berry set or cesses is also modulated by temperature, and setting), veraison (French ve´rer ¼ to change: some can be altered by stress factors such as color change or onset of ripening), harvest 2.1. SEASONS AND DAY LENGTH 51
FIGURE 2.1 Grapevine growth stages according to Baillod and Baggiolini (1993). 52 2. PHENOLOGY AND GROWTH CYCLE
FIGURE 2.2 Grapevine growth stages according to Eichhorn and Lorenz. Reproduced from Coombe, B. G. (1995). Adoption of a system for identifying grapevine growth stages. Aust. J. Grape Wine Res. 1, 104-110, with permission by Wiley-Blackwell. 2.2. VEGETATIVE CYCLE 53
(ripeness or maturity), and leaf senescence moisture, and rootstock, but on average it and subsequent fall (abscission). Although by seems to begin when the soil temperature rises definition no visible growth occurs during dor- above approximately 7 C (Alleweldt, 1965). mancy, metabolism does not rest completely, Bleeding can last for a few days or several but high concentrations of the dormancy hor- weeks (probably depending on whether or not mone abscisic acid keep it at a minimum neces- air temperatures are conducive to budbreak); sary for survival of the buds and woody it can also be a stop-and-go process that fluctu- tissues. Although the division of meristematic ates with changes in soil (root-zone) tempera- cells in the buds and cambium is blocked, chro- ture (Andersen and Brodbeck, 1989b; Reuther mosome duplication and protein synthesis and Reichardt, 1963). A vine can exude bleed- resume during the later stages of dormancy in ing sap at rates of less than 0.1 L to more than preparation for the activation of growth when 1 L per day, with the highest rates occurring temperature and soil moisture become favor- on warm and moist soils (Alleweldt, 1965; able in spring. Currle et al., 1983). Bleeding is probably caused The annual growth cycle of mature, fruiting by root pressure (Fischer et al., 1997; Priestly grapevines is often divided into a vegetative and Wormall 1925; Sperry et al., 1987), whose and a reproductive cycle. existence in grapevines was first demonstrated by the English cleric Stephen Hales (Hales, 2.2. VEGETATIVE CYCLE 1727). Root pressure arises from the remobiliza- tion of nutrient reserves from starch and pro- teins and pumping of sugars (especially In late winter or early spring, grapevines glucose and fructose) and amino acids (espe- often exude xylem sap from pruning surfaces cially glutamine) into the xylem (Reuther and and other wounds that have not yet been Reichardt, 1963; Roubelakis-Angelakis and suberized (Figure 2.3). Such sap flow or “bleed- Kliewer, 1979). The concentration of sugars ing” marks the transition from dormancy to apparently increases exponentially as the active growth. Initiation of bleeding is related temperature declines close to or below 0 C; to the restoration of metabolic activity in the this response occurs within a few hours and is roots and is influenced by soil temperature, equally rapidly reversible when the temperature
FIGURE 2.3 Bleeding grapevine cane (left); swelling, woolly bud just before budbreak (right). Photos by M. Keller. 54 2. PHENOLOGY AND GROWTH CYCLE rises again (Moreau and Vinet, 1923; Reuther cases root pressure seems to remain active for and Reichardt, 1963). In addition, organic acids more than 1 month after budbreak (Reuther (especially malate, tartrate, and citrate) and min- and Reichardt, 1963). Both bleeding and root eral nutrient ions (especially potassium and respiration, which may serve as a proxy for calcium) are also present in the bleeding sap metabolic activity in the roots, decline as soil (Andersen and Brodbeck, 1989a; Glad et al., moisture decreases, and membrane leakage 1992; Marangoni et al., 1986; Priestley and increases sharply once soil moisture falls below Wormall, 1925). The resulting increase in the approximately 5% (v/v) (Currle et al., 1983; osmotic pressure of the xylem sap provides Huang et al., 2005b). This may be the reason the driving force for osmotic water uptake by why bleeding is minimal or even absent in the roots from the soil, which generates a positive vines growing in dry soil, which can cause hydrostatic pressure in the entire xylem of 0.2–0.4 stunted shoot growth (possibly due to cavita- MPa at the trunk base and declining at a rate of tion) when shoots begin to grow with stored 0.01 MPa m 1 as the sap rises (Scholander et al., water from the trunk and root parenchyma in 1955; Sperry et al., 1987). This pressure in turn the absence of sufficient soil water to supply lifts water up the vine (see Chapter 3.3) and to the unfolding leaves. Nevertheless, the emerging the swelling buds up to several meters of height. shoots quickly become dependent on the Root pressure serves to dissolve and push delivery in the reactivated phloem of sugar out air bubbles that have formed during the and other nutrients remobilized from the winter in the xylem and thus restores xylem permanent vine organs. function (Scholander et al., 1955; Sperry et al., Cell division and auxin production in buds, 1987). This is necessary because when xylem beginning in the proximal leaf primordia of sap freezes, the gas dissolved in the sap is the distal buds, starts 1–3 weeks before bud- forced out of solution because gases are practi- break (see Figure 2.3), which marks the onset cally insoluble in ice. This can lead to cavitation of vegetative growth in spring. At the same (see Chapter 3.3) upon thawing if the bubbles time, abscisic acid (ABA) declines strongly, expand instead of going back into solution. relieving the growth inhibition that keeps the Given that bleeding typically starts before the vine dormant while low temperatures would buds begin to swell, root pressure may also be threaten its survival. The release of auxin and necessary to rehydrate the dormant buds and its basipetal (i.e., from the apex to the base) trans- reactivate them by cytokinin that is delivered port in the cambial region begins suddenly and from the roots (Field et al., 2009; Lilov and rushes like a continuous wave down the vine, Andonova, 1976). In fact, the buds’ water con- stimulating the cambium cells to resume divi- tent quickly rises from approximately 40% to sion and differentiation into new phloem and 80% during the budswell phase (Fuller and xylem cells (Aloni, 2001). The cambium cells Telli, 1999). The delivery of sugars in the become sensitive to auxin when they are absence of phloem flow may enable the buds exposed to cytokinin, which is supplied from to resume growth (i.e., cell division in the api- the roots as a result of root pressure. Although cal meristem), produce the growth hormone it remains to be demonstrated whether the root auxin (or release it from storage pools), and tips are capable of cytokinin production this break. Accordingly, root pressure-driven water early in spring, cytokinins have indeed been ascent may stimulate bud swelling, and the found in the bleeding sap of grapevines and, in maximum rate of bleeding occurs a few days addition to being influenced by the rootstock before budbreak and generally stops 10–14 (Skene and Antcliff, 1972), higher root tempera- days later (Alleweldt, 1965), although in some tures favor cytokinin delivery via the xylem 2.2. VEGETATIVE CYCLE 55
(Field et al., 2009; Zelleke and Kliewer, 1981). As discussed in Chapter 7) but may take 10 times an aside, the same combination of the actions of longer in areas with mild winters. As a conse- bud-derived auxin and root-derived cytokinin quence, budbreak is often erratic in warm and probably also induces the cambium on either (sub)tropical regions, where the majority of the side of new grafts to produce undifferentiated world’stablegrapesaregrown.Thisiswhy cells termed callus and to form a union between hydrogen cyanamide (H2CN2), usually in the form scion and rootstock. The differentiation of cells of calcium cyanamide (CaCN2), is commonly produced in the previous growing season to applied 3 or 4 weeks before the intended time of form the first xylem vessels of the new shoot budbreak to end dormancy (so-called dormancy begins even before the cambium starts produc- release) and promote budbreak of many table ing new cells (i.e., before budbreak), although grapes (Dokoozlian et al., 1995; Lavee and these vessels only mature (i.e., lignify) after bud- May, 1997). The higher winter temperatures in break. Warmer temperatures during vessel dif- such warmer areas lead to higher respiration ferentiation, perhaps by enhancing the supply rates (see Chapter 4), and this might be asso- of remobilized nutrient reserves to the cambium, ciated with an increase in oxidative stress (e.g., result in wider vessels that may be more effec- hydrogen peroxide release by the mitochondria) tive at supplying water to the developing shoots that, under extreme conditions, could result in later in the season (Fonti et al., 2007); this temper- bud damage called bud necrosis (Pe´rez et al., ature effect could be an important driving vari- 2007). Hydrogen cyanamide also leads to hydro- able for seasonal shoot vigor. Because cambium gen peroxide accumulation by inhibiting the reactivation occurs well before the new leaves antioxidant enzyme catalase, but the subsequent þ will begin supplying assimilates, the energy movement of calcium ions (Ca2 )fromthecell and carbon requirements of the dividing cells walls to the cytoplasm, similar to the movement have to be met by the remobilization of stored induced by chilling, then somehow seems to starch and supply of sugar from the permanent initiate cellular mechanisms culminating in structure of the vine. This starch breakdown dormancy release of the buds (Pang et al., 2007). may by induced by gibberellin, whose produc- In cooler, temperate climates, mean daily tion is stimulated by the auxin emanating from temperatures above approximately 8–10 C the buds and shoot tips (Frigerio et al., 2006; induce budbreak and shoot growth, but the Yamaguchi, 2008). precise base temperature requirement depends Although some grape cultivars (e.g., Thompson on the species (e.g., V. berlandieri > V. rupestris > Seedless) are believed not to require chilling (i.e., V. vinifera > V. riparia) and cultivar (e.g., a period of low temperatures) to resume growth Riesling > Cabernet Sauvignon, Viognier > from dormant buds, budbreak generally occurs Cabernet Franc, Syrah, Se´millon > Merlot, earlier, more rapidly, and more uniformly on Pinot, Sauvignon blanc > Chardonnay, Chenin vines that have experienced a period of cool blanc > Gewu¨ rztraminer). Cultivars with a weather (Antcliff and May, 1961; Currle et al., lower base temperature break buds earlier than 1983; Kliewer and Soleimani, 1972). Moreover, those with a higher temperature threshold longer duration of chilling and lower tempera- (Pouget, 1972). The rate of budbreak increases tures typically accelerate the rate of budbreak as the temperature rises above the threshold once warmer temperatures return (Dokoozlian, temperature up to approximately 30 C and then 1999; Lavee and May, 1997). The buds on a declines again. However, it is the temperature single vine usually break within a few days in of the buds themselves, rather than the areas with cold winters (provided the winter is surrounding air temperature, that determines not so cold as to damage or kill the buds, as when and how rapidly those buds break. 56 2. PHENOLOGY AND GROWTH CYCLE
During the day, the temperature of buds can be buds of the cane break early and grow vigorously, several degrees above the ambient temperature, whereas those toward the base of the cane grow especially when they are heated by the sun and weakly or not at all (Figure 2.4). This inhibition when there is little or no wind, yet even at wind of basal bud outgrowth is especially prevalent speeds greater than 5 m s 1 buds may still be in warm climates, where it can lead to erratic approximately 2 C warmer than air (Grace, and nonuniform budbreak. It also varies among 2006). Conversely, at night the bud temperature cultivars; for instance, Carignan and Mourve`dre is often cooler than the air temperature. In con- (synonyms Monastrell and Mataro) show partic- trast to the temperature of the buds, and despite ularly strong correlative inhibition, whereas its impact on the rate of bleeding sap flow, root Syrah and Cinsaut are at the other end of the temperature may have little influence on the spectrum. Like the evolution of tendrils, this time and rate of budbreak (Field et al., 2009; behavior is an adaptation to the habitat of wild May, 2004). vines, which enables them to access sunlight at Shoot growth starts in the distal (topmost or the top of tree canopies with a small investment apical) buds of a cane or vine, proceeding basi- in shoot biomass per unit height gain. This petally to the roots. Yet preferential budbreak adaptive behavior creates both a viticultural pro- of distal buds often inhibits the growth of the blem (mainly in warm climates) and an oppor- proximal (basal) buds (Galet, 2000). This phe- tunity (mainly in cool climates) for cultivated nomenon is often attributed to apical domi- grapevines. The problem arises from the inconsis- nance, although it may actually be a case of tent budbreak of cane-pruned vines. Viticultural correlative inhibition. In correlative inhibition, practices aimed at overcoming correlative inhibi- growing shoots may inhibit each other and tion and promoting uniform budbreak include prevent buds on the same cane from breaking, spur pruning, bending (“arching”) and partial whereas in apical dominance the shoot tip inhi- cracking of canes, and application of hydrogen bits the outgrowth of lateral buds (Beveridge, cyanamide. Spur-pruned vines are typically 2006; Ferguson and Beveridge, 2009). Correlative unaffected by correlative inhibition, and where inhibition is most conspicuous in early spring mixed “populations” of spurs and canes are when, in cane-pruned vines, the most distal retained on the same vine, the buds on the spurs
FIGURE 2.4 Cane-pruned Cabernet Sauvignon vines displaying strong correlative inhibition in spring. Notice that bud- break is not inhibited on spurs adjacent to the canes. Photos by M. Keller. 2.2. VEGETATIVE CYCLE 57 break at the same time as the distal buds of the growth of both shoots and roots is entirely canes (see Figure 2.4; Clingeleffer, 1989). The dependent on nutrient reserves (mainly carbo- opportunity stems from the delay in budbreak hydrates and proteins or amino acids) stored on vines that are spur-pruned only after the in the permanent structure of the vine, until canes’ distal buds have already broken, which new leaves become photosynthetically compe- may enable vines grown in frost-prone locations tent and begin to produce and export carbohy- to escape spring-frost damage (see Chapter 7.4). drates. These reserves are thus depleted and Budbreak is followed by a period of very reach a minimum around bloom time or even rapid shoot growth, with a new leaf appearing later, making vines vulnerable to stress around every few days. In many cultivars, the unfold- the time of bloom (Lebon et al., 2008; Weyand ing leaves, and often the developing inflores- and Schultz, 2006b; Williams, 1996; Zapata cences as well, are reddish before they turn et al., 2004). Carbohydrate reserves are used green (Figure 2.5). This transient accumulation not only as building supplies to construct of anthocyanin pigments in or just below the new plant material in the absence of photosyn- epidermis probably protects the developing thesis but also to generate the energy (by respi- photosynthetic machinery (e.g., protochloro- ration) required to fuel this construction (see phyll) from excess light (Chalker-Scott, 1999; Chapter 4.4). Consequently, grapevines literally Liakopoulos et al., 2006; Rosenheim, 1920). The lose (dry) weight during the initial spring anthocyanins usually disappear by the time growth phase, before the photosynthesizing the leaf has produced roughly half of its maxi- leaves begin to add new biomass from about mum (i.e., mature) chlorophyll and carotenoid the five- to six-leaf stage onward (Eifert et al., content and the palisade cells have expanded 1961; Hale and Weaver, 1962; Koblet, 1969). to occupy approximately 50% of the mesophyll Nitrogen reserves, in addition to carbohy- (Hughes et al., 2007). In addition, the dense drates, are very important for new growth in covering of “hair,” termed pubescence, of the spring; early season growth and especially leaf young leaves also serves the same photoprotec- chlorophyll correlate closely with the amount tive purpose (Liakopoulos et al., 2006). Early of nitrogen reserves stored in the perennial
FIGURE 2.5 Emerging Merlot shoot with pink unfolding leaves and inflorescences days after budbreak (left); Muscat Ottonel shoot with red young leaves (right). Photos by M. Keller. 58 2. PHENOLOGY AND GROWTH CYCLE parts of the vine (Cheng et al., 2004b; Keller and the agricultural Green Revolution in the second Koblet, 1995a; Treeby and Wheatley, 2006). half of the 20th century (Hartweck, 2008). Higher soil temperatures are associated with Meanwhile, the thick, cellulose-reinforced more rapid starch remobilization in the roots, outer cell walls and rigid cuticle of the epider- which may allow the shoots to grow more mis act like a corset, forcing the internode rapidly on vines in warm than in cool soil cells to elongate rather than simply expanding (Field et al., 2009; Skene and Kerridge, 1967; radially in all directions (Kutschera, 2008a,b; Woodham and Alexander, 1966; Zelleke and Schopfer, 2006). The suppression of axillary Kliewer, 1979). bud outgrowth by auxin is indirect because The initial growth phase is characterized by the hormone does not enter these buds, which, strong apical dominance, whereby slow but of course, begin producing and exporting auxin active (i.e., mediated by ATP-consuming pumps) themselves as soon as they begin to grow basipetal flow of auxin released by the leaf (Ongaro and Leyser, 2008). In addition to at primordia and young leaves near the shoot least one unknown mechanism, auxin may apex inhibits lateral shoot growth from prompt exert its remote control by inhibiting cytokinin buds and stimulates extension of internodes production in the roots and in the shoot tissues (Woodward and Bartel, 2005). Intriguingly, adjacent to the axillary buds (Beveridge, 2006; however, if auxin flows too rapidly, such as in Dun et al., 2006; Ferguson and Beveridge, overly vigorous shoots, apical dominance is 2009; McSteen and Leyser, 2005). Damage to compromised and lateral buds are able to grow the shoot tip or its removal (e.g., by summer out (Berleth et al., 2007; Dun et al., 2006; Lazar pruning) releases the repression of cytokinin and Goodman, 2006). Gibberellins produced synthesis and delivery from the roots via the upon stimulation by auxin in or near the shoot xylem (Nordstro¨m et al., 2004; Tanaka et al., tip and brassinosteroids produced in the inter- 2006). Cytokinins that enter the buds stimulate node epidermis promote cell elongation in the growth of lateral shoots, whose leaves are ordi- shoot’s internodes, perhaps by activating the narily smaller than the leaves of the main shoot wall-loosening protein expansin (Cosgrove, and which often stop growing after they have 2000; Frigerio et al., 2006; Olszewski et al., 2002; formed just a few leaves. Incidentally, removal Savaldi-Goldstein et al., 2007; Weiss and Ori, of the shoot tip also depletes auxin in the roots 2007; Yamaguchi, 2008), in addition to promot- and inhibits their growth, at least temporarily ing production of the membrane water channel (Ferguson and Beveridge, 2009; Fu and protein aquaporin (Maurel et al., 2008; see also Harberd, 2003). Growers employ continued Chapter 3.1). Gibberellins act, it seems, by shoot tipping to encourage lateral growth in binding to so-called DELLA proteins in the cell the tropics, where strong apical dominance is nucleus and inducing the degradation of these notorious. In temperate climates, the influence growth-suppressing proteins; however, just of apical dominance decreases once the main how DELLA proteins put the breaks on growth shoot has formed approximately 18–20 leaves remains to be discovered (Achard and Genschik, (around the time of bloom) and its growth 2009; Hartweck, 2008). Mutations that render slows somewhat (Alleweldt and Istar, 1969). DELLA proteins unable to bind gibberellins Cytokinins arriving with the transpiration and are thus immune to degradation result in stream and produced in the nodes then seem dwarf phenotypes that look like bonsai vines to be able to break auxin-dependent apical (Boss and Thomas, 2002; Boss et al., 2003); such dominance, and lateral shoots begin to grow, mutations formed the basis for the short and even when the main apex is intact. Lateral out- high-yielding cereal cultivars that gave rise to growth is especially pronounced (and may 2.2. VEGETATIVE CYCLE 59 occur much earlier than the 18-leaf stage) if summer so that growth can be controlled by warm temperatures stimulate rapid shoot appropriate deficit irrigation strategies (see growth, possibly due to cytokinin production Chapter 7.2). Decreasing day length, perhaps within the lateral buds. In addition, rapid aided by declining temperatures, shortly before transpiration at high temperatures would also the grapes begin to ripen seems to be the trig- stimulate cytokinin delivery from the roots. ger that halts cell division in the apical meri- Abundant nitrogen availability also leads to stem and in the cambium, thus inducing strong lateral shoot growth (Keller and Koblet, dormancy (Davis and Evert, 1970; Esau, 1948; 1995a), possibly because nitrate stimulates Fracheboud et al., 2009; Rohde and Bhalerao, cytokinin production in the roots and export 2007). The cessation of growth and shoot matu- in the xylem (Sakakibara, 2006; Sakakibara ration (i.e., periderm formation) are brought et al., 2006). about by a rapid decline in apex-derived auxin The most distal lateral shoot often gradually in the cambial region and a concomitant assumes the function of the main shoot, and the increase in the root-derived abscisic acid in more basal laterals soon stop growing (Currle the xylem and phloem, where it remains et al., 1983). If water and nutrient availability high during the subsequent dormant period permits, shoot growth often resumes after fruit (Mwange et al., 2005). set and more or less stops in midsummer. The shoot growth cycle is completed with Shortly before or while the grapes begin to leaf senescence associated with the recycling ripen, the shoots begin to form a periderm, of nutrients from the leaves to the canes and hence becoming canes and turning from green permanent parts of the vine (Conradie, 1986; to yellowish-reddish brown. This shoot “matu- Dintscheff et al., 1964; Lo¨hnertz et al., 1989; ration” proceeds acropetally from the base Schaller et al., 1990), abscission (shedding) of toward the tip and is accompanied by replen- leaves, and, finally, dehydration and cold accli- ishment of storage reserves in preparation for mation of all woody parts in fall. Before they the following growing season (Eifert et al., are shed, the senescing leaves develop their 1961; Lo¨hnertz et al., 1989). This process is characteristic and often spectacular yellow delayed by abundant water and nutrient and/or red autumn color. The coloration supply, which also encourages continued or occurs because the green chlorophylls are renewed shoot growth beyond veraison. This degraded more rapidly than the yellow-orange can cause problems in regions with significant carotenoids, which are thus “unmasked” in the summer rainfall, especially on fertile sites that process. The carotenoids also undergo various have high water and nutrient storage capacity, chemical modifications that influence their color. where late-season shoot growth can compete In addition, newly produced red anthocyanins with the ripening clusters and storage reserve are accumulated once a leaf has degraded pools for assimilate supply (Sartorius, 1973). roughly half of its original chlorophyll. Neverthe- Unlike the shoots of many other woody per- less, whereas white-skinned grape cultivars are ennials, grapevine shoots do not form terminal unable to produce anthocyanins, even the leaves buds but continue to grow as long as environ- of most dark-skinned cultivars normally turn mental conditions permit. When this is no lon- yellow (sometimes orange) rather than red. ger possible, the shoot tip with the apical Anthocyanin production is triggered by and meristem dies. The sensitivity of the shoot apex consequently a sign of sugar accumulation to adverse environmental factors (e.g., drought (Gollop et al., 2002; Pirie and Mullins, 1976), or low temperature) may be exploited in areas which can be caused by restricted phloem efflux where the soil dries down sufficiently in early (e.g., due to virus infection, wind-induced injury, 60 2. PHENOLOGY AND GROWTH CYCLE
FIGURE 2.6 Red leaves above the point where a tendril has wrapped itself around a Cabernet franc shoot (left), on undercropped shoots growing from a twisted Cabernet Sauvignon cane (center), and pale yellow leaves on a partly broken Muscat blanc shoot (right). Photos by M. Keller.
or girdling; Figure 2.6) or nutrient deficiency (see interaction with temperature likely vary by Chapter 7.3). The anthocyanins, which seem to species and cultivar. During the carefully be dominated by the glucosides of peonidin orchestrated senescence program, which nor- and cyanidin, and other phenolic components mally starts in the mesophyll cells at the leaf that are accumulated in the leaves may be margins and then progresses inward and released into the soil, where they might be able toward the base, the leaves transition from to suppress the germination of seeds from other photosynthetic carbon assimilation to the plant species the following spring, at least until breakdown (catabolism) of chlorophyll, pro- they are themselves degraded by soil microbes teins, lipids, and nucleic acids and subsequent (Weir et al., 2004), thus acting as preemer- export of valuable nutrients to other plant gence bioherbicides that reduce interspecific parts (Jones and Dangl, 1996; Lim et al., competition. 2007). Before they die, the leaves lose about It is thought that decreasing day length triggers half of their carbon, much of it due to respira- leaf senescence. Temperature, by contrast, has tion to generate the energy required for the no effect on the onset of senescence, although degradation and export metabolism in the face lower temperature may accelerate the rate of of declining photosynthesis (Keskitalo et al., senescence once initiated (Fracheboud et al., 2005; Williams and Smith, 1985). The high 2009). In other words, for a given cultivar respiratory costs of senescence are increasingly grown at a particular latitude, leaf senescence met by the degradation of cell membranes and should start on approximately the same day use of their lipids to provide energy through each year but progress more slowly in warm respiration (Buchanan-Wollaston, 1997). Not than in cool years or on warm than on cool surprisingly, the nucleus with its complement sites. At the other end of the temperature of genes and the mitochondria with their res- spectrum, heat stress ( 45 C) may also accel- piration machinery are the last cell compo- erate the senescence rate (Thomas and Stod- nents to be dismantled in the process (Lim dart, 1980). Of course, the critical day length et al., 2007). Moreover, chlorophyll is degraded necessary to induce senescence and the not because its products are reusable but, rather, 2.2. VEGETATIVE CYCLE 61 because its light-absorbing properties make it a intermediate (e.g., Cabernet Sauvignon, Syrah, strong phytotoxin in the absence of photosyn- Chardonnay, Teleki 5BB, SO-4, or 3309 Couderc) thesis and to facilitate access to more valuable or weak (e.g., Gamay, 101–14 Mgt, or Riparia materials such as chloroplast proteins and Gloire). Growth typically accelerates with lipids. The nitrogen and much of the carbon of increasing temperature up to an optimum of the chlorophyll molecules, but not their magne- 25–30 C but slows as temperature rises sium, remain in the leaf (Ho¨rtensteiner and further and ceases at approximately 35–38 C Feller, 2002), at least until soil microbes decom- (Buttrose, 1969b; Currle et al., 1983). Tempera- pose the leaf once it has fallen to the ground. ture has a strong effect on photosynthesis and Considering that chlorophyll contains only respiration (see Chapter 5.2): Warm (25–30 C) approximately 2% of the leaf’s nitrogen, this loss daytime temperatures promote photosynthetic is a rather modest price the plant pays for recy- CO2 fixation, whereas somewhat cooler cling of proteins and lipids (Ho¨rtensteiner, (15–20 C) nighttime temperatures limit respi- 2006). Leaves thus manage to remobilize up to ratory CO2 loss. Consequently, a sufficient 50–80% of their nitrogen and phosphorus, 50% (but not excessive) difference between day of their sulfur, and 20% of their iron before their and night temperatures, rather than simply cell membranes collapse, marking the end of a warm daily average temperature, is likely their life (Keskitalo et al., 2005; Niklas, 2006). to promote growth, a phenomenon termed Following this resorption process, the phloem thermoperiodic growth (Went, 1944). Nonethe- is plugged and sealed, and the leaves are shed less, the assumption of a linear increase in at predetermined positions of cell separation at growth rate with rising mean temperature the base of the petiole, termed abscission zones, forms the basis for the calculation of the which are put in place (i.e., differentiated) soon so-called growing degree days (GDD), thermal after the leaves have fully expanded (Roberts time, or heat units. Amerine and Winkler (1944) et al., 2002). The formation of abscission zones proposed that grape-growing regions could is induced by the plant hormone ethylene, be classified by heat summation or cumulative whereas their position is dictated by auxin, GDD for a “standard” growing season (April– which disappears completely during the October in the Northern Hemisphere) as subsequent dormancy period. Each year, grape- follows: vines invest most of their aboveground carbon Xn and nitrogen in leaf production, which cycles GDD ¼ ðTi T Þ i¼1 b as resorbed material in the perennial plant parts and as dead organic material, so-called detritus where Ti is the mean daily air temperature for or humus, through the soil. the period April 1 to October 31 (n ¼ 214 days), T ¼ In viticultural terminology, the rate of shoot and b 10 C is the base temperature. growth or shoot elongation over time is Normally, Ti is calculated as the mean of the referred to as vigor. In addition to genetic daily maximum and minimum temperature. T T > effects (species, cultivar, and rootstock), it is Only daily values of i b 0 are added to strongly influenced by temperature, soil mois- the cumulative GDD, whereas values 0are ture, nutrient availability, vine reserve status, set to zero. Despite some limitations (Dry and pruning level, and even vine age. Some culti- Coombe, 2004) and although variations on this vars (e.g., Thompson Seedless, Cabernet franc, theme have been proposed, including a or Grenache) or rootstocks (e.g., Richter 140, “heliothermal index” (Huglin and Schneider, 1103 Paulsen, or St. George) are considered to 1998), a “latitude-temperature index” (Jackson be vigorous, whereas others are thought to be and Cherry, 1988), and “biologically effective 62 2. PHENOLOGY AND GROWTH CYCLE day degrees” (Gladstones, 1992), the “Winkler also favor high vigor (due to more abundant regions” (Winkler et al., 1974) are still used as a resource availability), as does increasing prun- standard today. Each 1 C increment in tempera- ing severity, because plants in general try to ture averaged over the period April–October reestablish a balanced root:shoot ratio (Buttrose adds 214 GDD to the standard growing and Mullins, 1968; Clingeleffer, 1984; Poorter season so that vineyards at lower latitude or and Nagel, 2000). Nevertheless, the canopy of lower elevation will tend to be classified into lightly pruned vines (e.g., after mechanical or higher (warmer) Winkler regions. The projected minimal pruning) develops earlier and more rise in temperature associated with global rapidly and grows larger than that of heavily climate change (Intergovernmental Panel on pruned vines (e.g., after manual spur or cane Climate Change, 2007) will likely shift many pruning) because a higher number of buds vineyard sites into the next higher Winkler leads to more shoots growing simultaneously region within the next 50 years. The predicted in spring, using a larger pool of reserves increase in average spring and autumn tem- accumulated in the previous growing season peratures will also lead to longer actual (as (Araujo and Williams, 1988; Possingham, 1994; opposed to standard) growing seasons, which Sommer et al., 1995; Weyand and Schultz, are determined by the frost-free period (days 2006a,b; Winkler, 1929). On the other hand, between the last spring and first fall frosts). the presence of fruit tends to depress vegetative Moreover, in addition to its influence on growth of both shoots, especially lateral growth rates, a rise in temperature also typi- shoots, and roots so that the total biomass cally accelerates phenological development, production per vine remains approximately with a consistent trend toward earlier bloom, constant (Eibach and Alleweldt, 1985; Pallas veraison, and harvest (Alleweldt et al. 1984b; et al., 2008; Petrie et al., 2000b). The direction Chuine et al., 2004; Jones and Davis, 2000; of growth also affects growth rate (Currle Webb et al., 2007; Wolfe et al., 2005). et al., 1983; Kliewer et al., 1989); upright- Nevertheless, although the duration of the growing shoots, such as those in a typical phase between budbreak and the time the fruit vertical shoot positioning trellis system, gener- of a given cultivar attains a specific sugar ally grow more vigorously than shoots forced concentrationisshorterinawarmerthanina to grow horizontally, which in turn grow more cooler climate, the cumulative GDD during rapidly than shoots pointing downward, such this period is higher in the warmer climate as those in Geneva double-curtain or Scott– (McIntyre et al., 1987). Henry systems. This effect on growth of the The temperature effect on shoot growth is shoot angle relative to vertical is termed gravi- caused, at least in part, by a stimulation of morphism (Wareing and Nasr, 1961) and is auxin production by high temperature, which associated with variations in xylem vessels, in turn promotes cell elongation. Vigor declines whose number and diameter seem to be tailored with increasing vine size; therefore, shoots of to optimize water and nutrient supply to the small, young vines usually grow more vigor- shoot (Lovisolo and Schubert, 2000; Schubert ously than those of large, old vines. This et al., 1999). Accumulation of auxin near the decline arises from the increased mechanical shoot tip is thought to be responsible for the cost of maintaining an older vine’s permanent production of more but narrower vessels in structure, from the hydraulics of water trans- downward-pointing shoots (Lovisolo et al., 2002b). port, and from the higher number of buds that The time of rapid shoot growth coincides result in more shoots. Increasing soil moisture with secondary growth (increase in diameter) and nutrient (especially nitrogen) availability of the shoots, cordon, trunk, and roots. This 2.2. VEGETATIVE CYCLE 63 secondary growth ends when the decreasing coinciding with shoot browning (Alleweldt, day length after midsummer leads to cessation 1960) and cold acclimation (see Chapter 7.4). of cell division in the cambium, well before The buds’ water content decreases from the temperature becomes too low to limit cell approximately 80% to 50%, and around verai- division (Druart et al., 2007). The shortening son they enter into a phase of endodormancy, day length also seem to decrease gibberellin in which is also called true, inherent, organic, the shoot elongation zone so that growth rates deep, or winter dormancy or rest (Currle et al., decline. The rapid growth period also coincides 1983; Mullins et al., 1992). This transition is with the formation of prompt (lateral) and dor- characterized by a continuous, rapid increase mant (compound) buds in the leaf axils. How- in the temperature and time required for the ever, development of the first compound buds buds to break and by minimal respiration rates. in the basal nodes begins as early as 3 or 4 The shift from paradormancy to endodor- weeks before budbreak or more than 1 year mancy is induced by decreasing day length before these buds will give rise to shoots (i.e., shorter photoperiod) and cooler tempera- (Morrison, 1991). Dormancy in the case of these tures, and it is associated with a strong increase buds means that cell division is blocked, in ABA in the buds and nodes (Currle et al., whereas metabolic processes (e.g., respiration) 1983; Rohde and Bhalerao, 2007). This transi- continue, albeit at a reduced level. Although tion occurs later in cultivars that break buds the buds are under apical dominance from the early (i.e., at lower temperature) the following growing shoot tip, and thus do not normally spring (Pouget, 1972). The genetic programs break, the prompt buds do not become dor- for bud dormancy and cold hardiness are mant and can break relatively readily to form superimposed; shorter photoperiods are suffi- lateral shoots, which contrasts with the com- cient to induce dormancy and leaf senescence pound buds. Nonetheless, following loss of (but not cold hardiness), whereas cooler tem- the shoot tip and removal of lateral shoots or peratures are required to develop cold hardi- severe defoliation (e.g., due to water stress or ness (Fennell and Hoover, 1991). Nonetheless, physical damage caused by pests), even dor- entryintoendodormancyalsoappearstobe mant buds are initially able to resume growth delayed under warm conditions. Endodor- (Alleweldt, 1960). This phase is variously mancy, which is defined as the inhibition of referred to as paradormancy, predormancy, growth by internal bud signals, prevents bud- conditional dormancy, or summer dormancy, break and renewed shoot growth during the defined as the inhibition of growth by distal favorable growth conditions of late summer organs or, specifically, by auxin released by and early autumn, which would lead to cer- the shoot tip and young leaves (Galet, 2000; tain death of the emerging shoots due to fall Horvath et al., 2003; Lang et al., 1987). However, and winter frosts (Horvath et al., 2003; Lang the same buds that are arrested by incoming et al., 1987). (exogenous) auxin will start producing and Approximately 1 week at low but above- releasing auxin as soon as they are released freezing temperatures (chilling) in combination from dormancy and begin to grow. With the with short days appears to break endodor- slowing down of shoot growth (which starts mancy and to stimulate the renewed ability of earlier on less severely pruned vines due to dormant buds to resume growth when favor- their higher shoot number), the buds progres- able (i.e., warm) conditions return (Currle sively (over 2 or 3 weeks) lose the ability to et al., 1983; Galet, 2000; Pouget, 1972). This break even in the absence of apical dominance, period ordinarily coincides approximately with beginning with the most basal buds and leaf fall or soon thereafter in temperate climates 64 2. PHENOLOGY AND GROWTH CYCLE but may extend into midwinter in regions with of a given cultivar is determined by thermal mild winters. The transition seems to be related time (cumulative GDD) during the ecodor- to a decline in ABA in the buds and nodes mancy period. Incidentally, the cambium of (Currle et al., 1983). The upper temperature woody organs also seems to experience similar threshold required to break endodormancy is periods of rest (endodormancy) and quiescence approximately 10 C (mean daily temperature), (ecodormancy) and is reactivated in spring but it is lower for “late” cultivars with a higher when the cambial cells resume cell division base temperature for budbreak and shoot (Begum et al., 2007). growth in spring (and higher for “early” culti- Root growth, especially the formation of lat- vars with lower base temperature). Thus, early eral roots, is stimulated by auxin that has been cultivars enter into dormancy later than late delivered acropetally (from its main place of cultivars, and their dormancy is less intense production in the shoot tips and unfolding and more easily lifted (Pouget, 1972). Although leaves to the root tip) in the phloem (passive chilling is probably not an absolute require- transport, dominant) and parenchyma (active ment to break dormancy in grapevines (at least cell-to-cell transport, which is less important not in all cultivars), and its effect seems to be except during seed germination) and by gibber- quantitative rather than qualitative, no chilling ellins (promoting cell elongation) produced in (especially in combination with an extended the root tips (Friml, 2003; Kramer and Bennett, growing period in late fall) usually leads to 2006). Changes in auxin flow usually lead to delayed, limited, and uneven budbreak. This changes in root growth, and interrupting the can be a problem in warm subtropical and auxin flow (e.g., by cutting off the trunk) will tropical climates, where budbreak is often arti- inhibit root growth completely. Lateral roots ficially induced by application of H2CN2 after grow from the pericycle (and not the apical pruning. Once endodormancy has been lifted, meristem as in the shoot) following the initia- the main factor that prevents budbreak is low tion of renewed cell division by basipetal (from temperature (but lack of water can also inhibit the tip to the base) auxin transport in the cortex budbreak); thus, this last phase is called and epidermis. Following initiation, further ecodormancy (also termed enforced, relative, division of the “founder cells” leads to the for- imposed, or postdormancy or quiescence), mation of a lateral root primordium, which which is defined as the inhibition of growth pushes radially out of the parent root via cell by temporary, adverse environmental condi- expansion, again stimulated by auxin (and tions (Galet, 2000; Horvath et al., 2003; Lang repressed by ABA, e.g., under drought or et al., 1987). In other words, unfavorable grow- osmotic stress from excessive nutrient supply ing conditions simultaneously break endodor- or salinity). After emergence, the apical meri- mancy and impose ecodormancy. The latter is stem is activated and the new lateral root related to cold-induced (or drought-induced) begins to grow autonomously. Note that accumulation of ABA and coincides with maxi- whereas dormant buds seem to inhibit adventi- mal cold hardiness (see Chapter 7). The return tious root formation on hardwood cuttings of warmer temperatures in spring (or unsea- (Smart et al., 2003), the lack of auxin that is nor- sonably warm temperatures during winter) mally produced and released by swelling buds eventually leads to the release from dormancy, may be the reason disbudded cuttings are often which is associated with a rapid (within days) unable or struggle to form roots, unless they gain in the buds’ moisture content back to are dipped in an auxin solution. Similar to 80% and a concomitant loss of cold hardiness shoots, roots also display a kind of apical dom- (Lavee and May, 1997). The date of budbreak inance, whereby the growing tip of a root 2.2. VEGETATIVE CYCLE 65 inhibits the outgrowth of lateral roots from its shortly after budbreak until leaf fall in both pericycle. In contrast to the shoots, however, young and mature vines (Araujo and Williams, the roots use cytokinin rather than auxin to 1988; Williams and Biscay, 1991). However, inhibit lateral root initiation too close to the most of the fine roots (<1 mm in diameter), root tip (Aloni et al., 2006b). This enables the which are the vine’s major water- and nutrient- root to continue growing down (or across) the absorbing structures, are short-lived, so they soil profile in search of water and nutrients. If have to be replaced continuously, probably the root fails to detect nutrients such as nitrate because the soil around them quickly or phosphate, its tip ceases to produce cytoki- becomes exhausted (Comas et al., 2000; Viala nin so that lateral roots can proliferate to and Vermorel, 1909). Only a few of them live explore new territory in the upper soil layers. longer and may eventually become large Although root tips can extend at rates of sev- structural roots. It is estimated that vines allocate eral centimeters per day, it seems that root between 30 and 60% of total net photosynthesis growth generally lags behind shoot growth, products (see Chapter 5.1) to the roots under peaks around the time of bloom and early fruit favorable soil conditions, and this portion is development, and tapers off during fruit ripen- even greater in poor soils or during drought ing (Reimers et al., 1994). Early canopy develop- (Anderson et al., 2003). New roots are initially ment, such as occurs with lightly or minimally white, but they turn brown after approximately pruned grapevines, is associated with more pro- 5 weeks and black after an additional 3–6 nounced early root growth (Comas et al., 2005). weeks. The onset of browning generally indi- The supply of photoassimilates from the leaves cates cessation of root metabolic activity and is evidently critical to sustain root growth, and thus death of the root as a “functional” organ loss of leaf area inhibits root growth far more in terms of water and nutrient absorption and than it does shoot or fruit growth (Buttrose, as a source for new laterals (Comas et al., 1966a). The majority of new roots are pro- 2000). Black roots shrivel and eventually disap- duced while the shoots are growing vigor- pear altogether. Fine roots produced before ously, but root growth can continue as long bloom have an extremely short life span com- as water and nutrient availability permit and pared with those produced later in the growing the soil temperature remains high enough season (Anderson et al., 2003). Fine roots also (Williams and Biscay, 1991). In warm climates, die sooner in dry soil and at shallower soil or in warm growing seasons in cooler cli- depth, and so do roots of heavily pruned vines mates, that permit continued assimilate pro- supporting only few shoots. Such an increase in duction by the leaves, there may be a second root mortality can be considered a response to flush of new root growth around or just after herbivory, whereby plants compensate for the harvest (Conradie, 1980; Lehnart et al., 2008; loss of shoot biomass (i.e., carbon shortage) Mohr, 1996). Surprisingly, root growth has and attempt to restore their root:shoot ratio to even been recorded through the winter season, a balance that is more favorable for growth presumably sustained by nutrient reserves (Bloom et al., 1985). Therefore, mechanical (Bauerle et al., 2008b). Early canopy develop- and, particularly, minimal pruning tend to ment (i.e., larger leaf area) in spring promotes result in more and longer-lived fine roots than early root growth; thus, root growth of lightly spur or cane pruning. Lightly pruned vines pruned vines peaks earlier than that of heavily also produce more fine roots (to balance their pruned vines. The total root biomass increases larger canopy) than heavily pruned vines, even from growth of permanent roots in length and though the former generally support a larger diameter throughout the growing season from crop. In addition, infection with mycorrhizal 66 2. PHENOLOGY AND GROWTH CYCLE fungi greatly prolongs the life span of fine roots, cytokinin, which is then distributed throughout whereas nitrogen addition shortens it. the vine via the transpiration stream in the Grapevines, like other plants, are stationary: xylem, mainly to rapidly transpiring organs Unlike animals, they cannot move from place (Aloni et al., 2005). Depending on soil condi- to place to find better food sources. Because tions, the majority of roots are usually concen- uptake of water and nutrients by the roots trated in the nutrient- and oxygen-rich surface from the soil quickly depletes readily available soil, typically in the top 50 cm, with some roots resources, the roots must keep elongating penetrating to several meters of depth (Celette throughout the growing season and throughout et al., 2008; Morlat and Jacquet, 1993; Smart the vine’s life to maintain the supply of these et al., 2006). Vineyard floor vegetation, such as raw materials. This leads to the formation of a cover crops or permanent swards, tends to shift branched root system that is elaborated with the maximum root density of grapevines to root hairs and symbiotic mycorrhiza (Gebbing 50–100 cm, probably because of competition et al., 1977; Possingham and Groot Obbink, for water and nutrients (Lehnart et al., 2008; 1971; Stahl, 1900). Root density increases as a Reimers et al., 1994). Somewhat paradoxically, vine grows older but appears to settle and however, the average rooting depth tends to remain more or less constant after the vine be much greater in shallow soils (over bedrock) reaches 5–10 years of age. Although 40–50% of than in deep soils because the roots in shallow the volume of a typical soil consists of pores, soils can grow down cracks in the rock, root exploration of the soil ordinarily follows following the path of water infiltration. the path of least resistance; roots predomi- Root growth and distribution is influenced by nantly grow along natural fracture lines of soil species, rootstock, and cultivar. Scion cultivars and reinvade macropores created by old root also alter the root growth of different rootstocks debris forming microniches of high organic (Erlenwein, 1965a). Such genetic influences matter. The decaying dead roots additionally and interactions may be partly responsible for supply nutrients to the growing new roots, the variation in shoot vigor observed with provided these mineralized nutrients are not different scion/rootstock combinations because leached by excessive rainfall or irrigation. Of grapevines try to maintain an equilibrium in course, the general direction of root growth is root:shoot ratio. In fact, total root length nor- downward; that is, the root perceives and mally correlates closely with leaf area, and root responds to gravity (a process termed gravi- dry weight correlates with aboveground dry tropism). The “gravity sensors” are specialized weight in grapevines (Petrie et al., 2000b). amyloplasts (starch-containing plastids) called However, whereas root density may be depen- statoliths that settle at the bottom of the root dent on the genotype, it appears that root cap cells. When a root is forced away from ver- distribution is dictated by soil properties tical (by rocks or other obstacles), the resettling (Mullins et al., 1992). Even genetically identical of these statoliths triggers a redistribution of plants (i.e., clones) can develop vastly different calcium and movement of auxin to the new root systems in terms of both size and archi- lower side of the root back in the elongation tecture, depending on microscale variations in zone so that the cells on that side elongate more soil conditions that the roots encounter during than on the upper side, allowing the root to growth. Such differences result from changes navigate around the obstacle (Kramer and in the number, growth rate, distribution, and Bennett, 2006). Note that the gravity-sensing orientation of lateral roots. Thus, temperature, cells in the root cap are also the ones that pro- soil texture, moisture, and nutrient availability duce most of the plant’s cell division hormone also strongly and rapidly affect root growth 2.2. VEGETATIVE CYCLE 67
FIGURE 2.7 Riesling roots grown at 15 C (left) or 25 C (center; reproduced with permission from Erlenwein, 1965b), and Cabernet Sauvignon roots growing down fractures in limestone bedrock (right; photo by M. Keller). and distribution. In some cases, the effect of down a moisture gradient in a process called temperature on root growth may be even more hydrotropism, and this ability is so strong that pronounced than its influence on shoot growth it can override the roots’ gravity response (Skene and Kerridge, 1967; Figure 2.7); a mini- (Eapen et al., 2005). How the roots “sense” the mum soil temperature of 6 C seems to be neces- presence of water and how they integrate the sary for roots to grow (Richards, 1983), and a resulting signal with the gravity signal is not 10 C increase in temperature can double or known, but hydrotropism partly seems to result triple the rate of root extension. Although the from the roots’ propensity to grow away from optimum is close to 30 C, temperatures above regions of high osmotic pressure (Takahashi that threshold can kill at least the fine roots et al., 2003). Their growth clearly responds to within days (Huang et al., 2005b). This can result the presence of water tables, the amount and in problems for grapevines grown in pots, where frequency of water application (whether from the soil temperature during the day often rainfall or irrigation), and the type of irrigation exceeds the ambient temperature by several system (e.g., drip vs. furrow). Therefore, grape- degrees. Although under field conditions soil vines can tolerate considerable drought due to temperatures usually fluctuate much less than their ability to selectively grow roots in soil aboveground temperatures due to the large heat patches with available water (Bauerle et al., capacity of wet soils, heavy rainfall or applica- 2008b). This is why the roots of drip-irrigated tion of cold irrigation water can lead to short- vines (especially in sandy soils and dry term changes in soil temperature. Moreover, climates) are often concentrated beneath the dry and sandy soils are not as well buffered drip emitters, whereas in rainfed vineyards or against temperature fluctuations as are wet those irrigated with other methods (e.g., flood- and loamy soils. Mild water deficit has little ing, furrows, and overhead sprinklers) the root effect on lateral root growth, but more severe system is typically much more widespread stress suppresses root growth, although not (Stevens and Douglas, 1994), unless root as severely as it inhibits shoot growth (see growth in the midrows is prevented by soil Chapter 7.2). compaction from heavy machinery. Roots of Roots tend to search out moist soil regions. mature vines also change their growth pattern In other words, they grow toward water or accordingly following conversion from flood 68 2. PHENOLOGY AND GROWTH CYCLE or sprinkler to drip irrigation (Soar and Loveys, growing season under optimal conditions. 2007). Similarly, permanent interrow cover Another trait typical of perennial woody spe- crops tend to lead to a concentration of vine cies is that they require two consecutive grow- roots in the bare, herbicide-treated vine rows ing seasons for flower and fruit production: (Celette et al., 2008; Morlat and Jacquet, 2003). Buds formed in the first year give rise to shoots The roots’ ability to find water even below carrying fruit in the second year. Reproductive the bedrock underlying the “useful” soil hori- growth is very similar in the different Vitis spe- zons (see Figure 2.7) can also greatly expand cies and begins with flower formation, which the apparent rootzone or rooting depth as esti- can be divided into three separate processes: mated from soil pits. This is why grapevines inflorescence initiation (or induction), flower planted on apparently shallow soil are not initiation, and flower differentiation. The first always less vigorous than those planted on step involves the formation of lateral meristems deeper soil, and why their growth is not neces- as uncommitted primordia or initials by the sarily easier to control using deficit irrigation apical meristem inside the shoot’s axial com- strategies. pound buds from spring through early summer In contrast to its effect on shoot growth, (Alleweldt and Ilter, 1969; Carmona et al., 2008; increasing nutrient (especially nitrogen) avail- Gerrath, 1992; Gerrath and Posluszny, 1988b; ability decreases root growth. Soil compaction, Meneghetti et al., 2006; Morrison, 1991; Pratt, leading to a decrease in the soil’s pore volume, 1971; Srinivasan and Mullins, 1981). Primordia also reduces root (and consequently shoot) are first formed in the buds at the shoot base, growth; pressurizing the soil by 1 bar due to and thereafter they also appear progressively compaction may decrease root elongation by in the buds toward the shoot tip. These primor- 90%. Nonetheless, localized soil compaction dia (i.e., meristems) can develop (differentiate) (e.g., due to heavy machinery or plow soles) into inflorescence primordia (i.e., they form reduces shoot growth only when it decreases several inflorescence branch primordia or mer- the total root length (Montagu et al., 2001). When istems), tendril primordia, or sometimes even a single root grows through loose soil into com- shoot primordia. Initiation of uncommitted pri- pact subsoil, compensatory root growth in the mordia occurs after four or five leaf primordia loose soil maintains total root length and thus have been initiated in the buds (i.e., from shoot growth remains unaffected. Such plastic- budbreak to bloom), and subsequent inflo- ity in root system development in response to rescence differentiation begins around anthesis heterogeneous soil conditions enables grape- of the current season’s inflorescences and vines to acquire soil resources with a conserva- may continue until the buds enter dormancy tive investment in root biomass. (Alleweldt and Ilter, 1969; Morrison, 1991; Snyder, 1933; Swanepoel and Archer, 1988). In 2.3. REPRODUCTIVE CYCLE other words, inflorescences are formed during the year that precedes the flowering and fruit- ing year, and the fate of the following year’s Grapevines grown from seeds, like other primordia is determined as early as bloom time woody perennials, have a vegetative phase of of the current year. The number of inflores- several years before they reach their reproduc- cences is lowest at the shoot base and increases tive phase and become able to produce fruit. with higher bud position (Sartorius, 1968). Yet This ensures that the vine has ample resources in most cultivars only the basal six to eight to support fruit production. By contrast, vines buds of a shoot (i.e., those preformed in the grown from cuttings are able to fruit in the first previous growing season) are able to form 2.3. REPRODUCTIVE CYCLE 69 inflorescences (Sa´nchez and Dokoozlian, 2005); unfruitful. Buds that develop inside dense the uncommitted primordia initiated at youn- canopies are less fruitful than the buds at the ger node positions (i.e., those produced by canopy exterior (May et al., 1976; Shaulis et al., the actively growing shoot) always become 1966). However, bud fruitfulness appears to tendrils, never inflorescences (Morrison, 1991). be independent of the light exposure of the Thompson Seedless (synonym Sultana) is an buds but depends strongly on the supply of exception: Fruitful buds are initiated over assimilates to the buds (Keller and Koblet, the entire length of its shoots (Sa´nchez and 1995a; May and Antcliff, 1963; Sa´nchez and Dokoozlian, 2005). The origin of the shoot is Dokoozlian, 2005). These assimilates may come unimportant for the fruitfulness of its buds; from stored reserves or from leaf photosynthesis, shoots arising from 1-year-old spurs or canes but they are mostly supplied by the leaves on produce equally fruitful buds as those growing the same side of the shoot, perhaps especially from older cordons or trunks (Mu¨ ller-Thurgau, from a bud’s subtending leaf (Hale and 1883). Weaver, 1962). This has implications for the The maximum number of future inflores- design of trellis and training systems in viti- cences per bud, the so-called bud fruitfulness, culture. A renewal zone (the zone where buds appears to be determined by approximately are located that produce next year’s shoots 3 months after budbreak—that is, before verai- and crop) with well-exposed leaves, ideally son (Alleweldt and Ilter, 1969). Depending on situated close to the top of the canopy, and a species and cultivar, several inflorescence pri- canopy that is not too dense will maximize mordia may be formed in the primary buds of the potential crop because the number of pri- most Vitis species. Most V. vinifera cultivars mordia determines the vine’s yield potential typically initiate two such primordia, but for for the following year (see Chapter 6.1). In some, such as Se´millon, Muscadelle, Gamais, addition, the growth direction of the shoot or Carignan, three is more common, whereas also appears to influence bud fruitfulness, the fruitful Gouais blanc forms four, V. labrus- at least in some cultivars: Upright-growing cana cultivars form three or four, and V. riparia shoots produce buds with more inflorescence and V. rupestris buds form seven or eight inflo- primordia than horizontally or downward- rescence primordia (Viala and Vermorel, 1909). growing shoots (Alleweldt and Ilter, 1969). Moreover, environmental conditions strongly The commonly accepted view (Carmona modulate the number of primordia. Conditions et al., 2002, 2008; Gerrath, 1992; Gerrath and required for the formation of the maximum Posluszny, 1988b; May, 2004; Meneghetti et al., number of inflorescence primordia include 2006; Morrison, 1991; Mullins et al., 1992; high irradiance, warm temperature (25–30 C, Scholefield and Ward, 1975; Snyder, 1933; especially during the 3 weeks prior to bloom), Srinivasan and Mullins, 1981) of inflorescence and adequate water and nutrient supply and flower differentiation is that the formation (Currle et al., 1983; Meneghetti et al., 2006; of inflorescence meristems and that of flower Mullins et al., 1992; Sartorius, 1968; Sommer meristems are seasonally separated (Figure 2.8). et al., 2000; Srinivasan and Mullins, 1981). According to this view, inflorescence primordia Warm temperatures promote the induction of grow rapidly during the summer, producing inflorescences, whereas cool temperatures several branch meristems before the buds enter (<20 C) promote tendril formation (Buttrose, dormancy in late summer. Further inflores- 1969a, 1970b). Unfavorable conditions (e.g., cence development is arrested with the onset temperatures >35 C) may reduce the primor- of dormancy and does not resume until the dia number to zero, rendering the buds buds begin to swell in late winter or early 70 2. PHENOLOGY AND GROWTH CYCLE
FIGURE 2.8 Cluster initiation (1) and differentiation (2–6: numbers indicate increasing orders and positions of branch- ing) and flower development (8: initiation of sepal lobes; 9: initiation of petal lobes; 10: initiation of stamen primordia; 11: initiation of carpel primordia; 12: appearance of locules; 13: ovule initiation; 14: pollen formation; 15: anthesis) of Concord grapevines (left; reproduced with permission of M. C. Goffinet); inflorescences of the rootstock 3309C emerging after bud- break (right; photo by M. Keller). spring. When the buds are reactivated, the inflo- The pistil with its reproductive structures rescence branch meristems produce additional does not develop until the individual flowers meristems that give rise to clusters of three or become visibly separated on the inflorescence. four flower meristems or floral primordia during Separate genes or groups of genes specify the the period of budswell that precedes budbreak identity of the individual floral organs, although by several days to weeks. This process is termed the mechanism by which they exert their influ- flower initiation, and within each of these clus- ence is not known (Dı´az-Riquelme et al., 2009). ters the floral primordia are initiated “from the The dissenting minority opinion of this top down”; that is, the terminal flower is initiated developmental process differs mainly in the first and the basal flower last. Flower initiation timing of flower initiation (Agaoglu, 1971; happens simultaneously in all parts of the inflo- Alleweldt, 1966; Alleweldt and Ilter, 1969; rescence primordium during the budswell phase Sartorius, 1968). This view holds that a portion and culminates in the formation of a calyx (fused of the calyxes are formed before bud dor- sepals) for each flower. Branching and flower mancy, some during winter dormancy (albeit initiation ostensibly cease once a shoot starts very slowly), and the remainder during the growing out of the bud. The subsequent flower subsequent budswell and budbreak period. differentiation, or the development of the indi- The corolla and stamen also appear before bud- vidual organs of each flower, also called floral break, followed by the pistil, whose formation organogenesis, occurs over approximately 5 is complete within 10–15 days of the emergence weeks during and after budbreak. The corolla of the inflorescence after budbreak. Intrigu- (joined petals) and stamens appear successively ingly, evidence supporting both of these views under the stimulating influence of gibberellin comes from both warm and cool climates, so (Cheng et al., 2004a), while the inflorescences it is not clear what might cause the disparity become visible on the shoot and then separate. in observations. Inflorescences that arise on 2.3. REPRODUCTIVE CYCLE 71 lateral shoots, especially on those laterals on tissues. The anthers produce and release particu- the free-growth portion of a shoot that could larly high amounts of auxin, as well as gibberel- not possibly have been initiated in the previous lin, and thereby synchronize (i.e., delay) the growing season, obviously run through their development of the other flower parts so that entire developmental program from inflores- they become functional just before anthesis cence initiation to fruiting within the same (Aloni et al., 2006a; Olszewski et al., 2002; growing season. Yamaguchi, 2008). Gibberellins help maintain The extent of branching before and after the the inflorescence meristems and favor inflores- dormancy period and the degree to which the cence elongation (Alleweldt, 1959). This is why individual flowers develop in spring determine table grape growers sometimes apply gibberel- the number of flowers that form on each inflo- lin sprays before bloom to lengthen the rachis rescence. The flower number is highly variable, and make clusters less compact (Mullins et al., even on the same shoot: The basal inflorescence 1992). Application during bloom may lead to of a shoot typically produces the most flowers, lower fruit set, which tends to further promote and numbers decline with increasing height of loose clusters. However, gibberellins act anta- insertion on the shoot (May, 2004). Branching gonistically to cytokinins and favor tendril for- of the primordia is induced and controlled by mation by inhibiting the branching process so two groups of plant hormones, namely auxins that bud fruitfulness may decrease. This inhibi- and cytokinins. The production of cytokinins, tory role appears to be somewhat unique to which occurs predominantly in the root tips, grapes because gibberellins are usually involved and their transport to the shoots in the xylem in establishing floral primordia in many plants. are regulated by developmental (genetic) and Approximately 2 weeks after the ovules environmental signals (e.g., water and nitrogen have been formed and 5–10 weeks after bud- availability). Auxin is produced and released break, anthesis marks the beginning of bloom, by the floral meristems (i.e., the tip of each floral exposing the male and female floral organs organ) and controls both formation and differen- (Figures 2.9 and 2.10). Anthesis theoretically tiation (i.e., morphogenesis) of the flowers; it refers to the release of pollen but is commonly also induces development of their vascular regarded as the shedding of the calyptra, or
FIGURE 2.9 Grape flowers immediately before anthesis (left, with cross section center left), during anthesis (center right), and immediately after anthesis (right). Illustrations by A. Mills. 72 2. PHENOLOGY AND GROWTH CYCLE
FIGURE 2.10 Chardonnay inflorescence at the beginning of bloom (left) and in full bloom (right, with insert showing a close-up of a group of flowers). Photos by M. Keller. capfall. Similar to budbreak, anthesis generally flowers are termed star flowers and can appear begins in the shoots growing from distal buds in a range of cultivars (Longbottom et al., 2008; and progresses basipetally toward the trunk. Pratt, 1971). On the other hand, anthesis within an individ- When the anthers burst open, they release ual inflorescence often starts close to its base, their pollen into the air. The production of pollen and all flowers normally open within 5–7 days, is stimulated by the plant hormone cytokinin. although this can be delayed during cold and The transfer of pollen grains (microgameto- rainy conditions. Only few flowers open at phytes) from the anthers to the stigma is called 15 C, and the optimum temperature seems to pollination and is followed by abscission of the be in the range 20–25 C; the rate of cap fall stamens. During anthesis, the stigma releases a slows greatly at 35 C (Galet, 2000; May, 2004). sticky sap that retains and rehydrates the pollen Each day opening typically begins at approxi- for pollination (Considine and Knox, 1979a; mately dawn and ends at approximately noon Meneghetti et al., 2006). This sap is supplied via (Staudt, 1999). Different inflorescences on the the newly formed stylar xylem that has previ- same vine, or on different vines in the same ously been induced by auxin released by the vineyard, follow the same rhythm but may stigma’s papillae. Because the tiny, dry pollen start it on different days so that the entire grains are highly exposed during this process, bloom period in a particular vineyard may last the pollen wall contains high concentrations of 2 or 3 weeks. Capfall occurs when the petals phenolic compounds termed flavonols that act detach from the receptacle and are lifted off as “sunscreen” (Downey et al., 2003b; see also by the anthers. It is thought that it is the rapid Chapter 5.2) protecting the genetic information elongation of the anther filaments that pushes (DNA) contained in the chromosomes from the calyptra away from its base once the basal destruction by UV radiation (Caldwell et al., cells have dehisced, which is a rare mechanism 1998). Nevertheless, excessive UVB radiation, among angiosperms. Occasionally, the petals such as that resulting from sudden exposure to instead open at the top and are spread open sunlight of previously shaded inflorescences, and pushed downward by the anthers. Such especially in conjunction with heat (>33 C), can 2.3. REPRODUCTIVE CYCLE 73 sometimes decrease pollen germination or pol- sperm cells (male gametes; Greek gamete ¼ len tube growth (Torabinejad et al., 1998). In spouse), penetrates the stigma and grows down other words, stress reduces pollen viability and through the intercellular spaces of the style like leads to male sterility. a hydraulic drill, delivering its package of chro- In contrast to their wild relatives and the few mosomes into the egg inside the ovule (Zonia cultivars with female flowers, cultivated grape- and Munnik, 2007). The male and female vines are typically self-pollinated, whereby pol- gametes (sex cells) meet when a pollen tube len originates from the flower’s own anthers enters the embryo sac and releases the sperm. (Mullins et al., 1992; Sartorius, 1926). Wind Their fusion in the ovule is termed fertilization appears to be responsible only for occasional and results in the zygote, which develops into cross-pollination, where pollen originates from the embryo and leads to fruit set. In a typical a flower on a different plant than the polli- grape flower, the pollen tubes cover a distance nated one (Scherz, 1939). In fact, the anthers of approximately 2 mm. Both pollen germina- often burst shortly (1–24 h) before the calyptra tion and pollen tube growth are induced by is shed, releasing the pollen on the flower’s auxin, with which the pollen has previously own stigma, a process termed cleistogamy been “loaded” in the anthers, and stimulated (Heazlewood and Wilson, 2004; May, 2004; by the growth hormone gibberellin that may Staudt, 1999). Despite this self-pollination strat- be produced by the pollen (Gillaspy et al., egy, the osmophors on the ovary may serve to 1993). Growth of the pollen tube represents a attract insects by emitting monoterpenes (simi- more than 1000-fold elongation of the pollen lar to citrus and lavender flowers), although cell and is very rapid; rates of 0.3 mm h 1 are they ostensibly do not offer the insects any nec- not uncommon (Staudt, 1982). Not surpris- tar. An intriguing possibility is that the release ingly, this process is very energy-consuming, of volatiles might be a defense strategy to deter and the growing pollen tubes have high respi- herbivorous insects or fungi that would feed on ration rates. Respiration is fuelled by the break- the highly vulnerable flowers or to attract down of starch reserves that are accumulated enemies of these herbivores (Gang, 2005). Insect in the pollen grains from imported sucrose pollination appears to be much more common, prior to anthesis. and is perhaps required for commercially The speed at which the pollen tubes grow acceptable fruit set, in Muscadinia than in Vitis toward the ovules is critical for fruit set because species (Olien, 1990; Sampson et al., 2001). the ovules are only receptive for a limited However, even in the latter, berries originating period after anthesis, after which they can no from cross-pollinated flowers apparently grow longer be fertilized. The rate of pollen tube larger than those from self-pollinated flowers, growth depends on temperature; at 25–30 C, which suggests that cross-pollination improves fertilization can occur within 12 h, at 20 C after seed formation. In any case, the pollen grains 24 h, and at 15 C after 48 h, whereas even absorb water from the moist stigma to rehy- lower temperatures do not permit fertilization drate (a process that requires approximately because the pollen tubes stop growing before 30 min), germinate on the stigma whose surface they reach the ovules or they simply reach cell layer subsequently suberizes (Considine them too late (Staudt, 1982). Eggs that are not and Knox, 1979a), and form a pollen tube. The fertilized within 3 or 4 days after anthesis will optimum temperature for pollen germination degenerate (Callis, 1995; Kassemeyer and is between 25 and 30 C, whereas temperatures Staudt, 1981). Therefore, slow pollen tube below 10 C and above 35 C inhibit germina- growth (e.g., due to low pollen “fuel status” tion. The pollen tube, which carries two fused resulting from stress interfering with starch 74 2. PHENOLOGY AND GROWTH CYCLE accumulation prior to pollination) leads to poor Similar to the placenta in mammals, the endo- fruit set (Koblet, 1966). Nonetheless, only one sperm protects the embryo developing from pollen tube from the many pollen grains land- the zygote, controls nutrient transfer from the ing on the stigma needs to reach one of up to berry, and serves as a nutrient store for the four receptive ovules to make fertilization developing embryo and the emerging plant likely. How the pollen tubes find the ovules is organs during seed germination. It seems that not well-known, but it seems that they are double fertilization is an all-or-nothing process: attracted by the synergid cells (which die after Either both the egg cell and the central cell are they have served their purpose) flanking the fertilized or neither is fertilized (Kassemeyer egg cell and follow gradients of g-amino buty- and Staudt, 1981). The resulting diploid embryo rate (GABA) (Berger et al., 2008; Palanivelu and (initially) triploid endosperm together form et al., 2003). GABA is chemically an amino acid, the filial (daughter) generation, whereas all but it is not referred to as such because it is not other berry tissues (including the seed coat and incorporated into proteins. In plant cells, it is the nucellus) are part of the maternal (mother) synthesized from glutamate, and in the human generation or mother plant. Yet the contribution brain GABA released from nerve cells is of the central cell’s maternally derived cyto- the predominant inhibitory neurotransmitter. plasm to the developing endosperm is quite sig- In any case, once the pollen tube reaches the nificant (Berger et al., 2006). Incidentally, the ovule, it stops growing, and its volume and embryo should more accurately be called the pressure rapidly increase. This leads to an zygotic embryo because somatic plant cells can explosive rupture of the tip, which ejects be forced through tissue culture (i.e., in vitro)to- the sperm cells into the ovule (Staudt, 1982). The form so-called somatic embryos, which can four ovules in a flower do not all reach develop into whole plants that are clones of their the same stage of development simultaneously: mother plant (Franks et al., 2002; Martinelli and Some embryo sacs may fail to develop or not Gribaudo, 2001; Mullins and Srinivasan, 1976). mature in time to be fertilized, others are not This somatic embryogenesis is exploited in “found” by a pollen tube, and yet others are vegetative propagation, which is used to help fertilized but are slow to develop (Kassemeyer eliminate viruses and bacteria by so-called and Staudt, 1982; Nitsch et al., 1960; Pratt, micropropagation or meristem culture, whereby 1971). This is why grape berries rarely contain shoot tips less than 0.5 mm in length are excised four seeds. under aseptic conditions and grown into full Fertilization in angiosperms is more com- plants (Barlass et al., 1982; Bass and Vuittenez, plex than in other plants or in animals. During 1977). Although this practice may occasionally the process of double fertilization, one haploid lead to the expression of juvenile leaf characters (i.e., having one set of chromosomes) male in the resulting plant, this is normally confined gamete fuses with the haploid female gamete to the lower portion of the new shoots, and (the egg cell or simply egg) to form the diploid cuttings taken from the distal portion will (i.e., having two sets of chromosomes—one give rise to true-to-type plants (Grenan, 1984). from the mother via her egg and one from the However, such propagation methods are belie- father via his pollen) zygote. The other male ved to encourage the activation of normally gamete (delivered by the same pollen tube) silent jumping genes (see Chapter 1.2), which almost simultaneously fuses with the other could induce occasional somatic mutations female gamete, the diploid central cell of the and thus introduce additional genetic variation embryo sac to give rise to the endosperm (Ber- in some of the tissue-cultured vines (Benjak ger et al., 2008; May, 2004; Raghavan, 2003). et al., 2008). 2.3. REPRODUCTIVE CYCLE 75
In contrast to somatic embryogenesis, two separate sexes and prohibiting them from although they are also derived from somatic mixing prevents this problem (i.e., limits the cells (and hence may carry somatic mutations), damage) while still permitting the (nuclear) but due to genetic recombination in addition to chromosomes to take advantage of the costly occasional mutations, each egg cell is geneti- sexual recombination. cally different from all other egg cells on the As a result of the chromosome mixing, every same vine, and each sperm cell is genetically single seed, even within the same berry, will different from all other sperm cells on the same give rise to a genetically unique plant, slightly vine. The recombination is basically a random different from all others, even though it inher- shuffling of paternal and maternal genes, just ited all its plastids and mitochondria from the like the shuffling of a deck of cards, and occurs mother plant. This production of new geno- approximately 1–3 weeks before anthesis types by mixing of genes (more accurately, (depending on cultivar; Lebon et al., 2008) dur- alleles; i.e., slightly different versions of the ing the production of gametes in a special cell same gene, one of which may or may not be division termed meiosis (Greek meioun ¼ to dominant over the other) in each generation in make small). Just before the cell divides, the addition to occasional mutations (which gener- chromosomes line up alongside each other ate the different alleles in the first place) pro- and randomly exchange pieces of DNA so that vides the endless genetic variety that under the new chromosomes that end up in the natural conditions allows evolution (specifi- gametes constitute a novel and distinctive mix cally, natural selection) to act by “weeding of genes from the mother and the father. This out” those individuals (phenotypes) less well shuffling is the true essence of sex and thus adapted to the prevailing environmental condi- only applies to genes contained in the cell tions (Mayr, 2001). As Charles Darwin (2004) nucleus—that is, on the chromosomes. The stated, "Natural selection acts solely through other cell organelles with their own genes (i.e., the preservation of variations in some way the mitochondria and plastids such as chloro- advantageous." Because “environment” also plasts) are reproduced asexually (by simple comprises other species, including pathogens division), and the male organelles are normally such as fungi, bacteria, and viruses, the major excluded from or left behind in the pollen. benefit of recombination is the decreased risk Only very few of them occasionally “slip” for a plant’s offspring to be wiped out by path- through (Birky, 1995; Timmis et al., 2004), and ogen attack (see Chapter 1). This is the meaning some of their genes may escape destruction of sex: By producing genetic variation (i.e., by and be “absorbed” by the nucleus (Sheppard “not putting all its eggs in the same basket”), et al., 2008). Indeed, this exclusion of the a species greatly increases the chance that at father’s organelle genes is the fundamental rea- least some individuals will resist and thus sur- son why there are two, and only two, sexes in vive pathogen attack (or drought, etc.). Of all sexually reproducing plants (and animals). course, such genetic variation is possible only Building and running an organism as complex in wild grapevines but not in cultivated popu- as a grapevine (or an animal) requires a great lations of vegetatively propagated (i.e., clonal) deal of cooperation among the genes. The plants. Nevertheless, even the various cultivars genes of the plastids and mitochondria have originally arose this way (whether by deliber- never “learned” how to cooperate, and these ate crossing or by natural pollination), and organelles would often attack and kill each recombination accounts for the vast differences other when two cells fuse (which, in fact, they among cultivars in terms of growth, yield for- do in single-celled algae). Assigning them to mation, fruit composition, pest and disease 76 2. PHENOLOGY AND GROWTH CYCLE susceptibility, and reaction to environmental is exploited in table grape and raisin produc- influences. Somatic mutations (bud sports), tion; growers routinely apply gibberellin sprays however, can arise in vegetatively propagated at bloom and fruit set to increase berry size grapevines and may give rise to clonal varia- (Mullins et al., 1992). In contrast to early seed tion that can be subject to selection by humans formation, the seeds later take charge of their (artificial as opposed to natural selection). own maturation by switching on the embryo’s Within days after fertilization, the endo- genes; in other words, there is a change from sperm nucleus begins to divide, whereas the maternal to filial control of seed development zygote (i.e., the future embryo) does not start (Gutierrez et al., 2007). These genes initiate dividing until approximately 2 or 3 weeks and coordinate the import of nutrients and after fertilization (Nitsch et al., 1960; Pratt, accumulation of storage reserves such as starch 1971); by this time, the cells of the nucellus and proteins in the endosperm and the have reached their maximum size, and the subsequent desiccation of the seeds. endosperm, whose nuclei have by now become Although pollination, rather than fertiliza- polyploid, begins to accumulate starch and tion, triggers the initial stages of ovary develop- lipids (Kassemeyer and Staudt, 1983). The first ment, subsequent fruit formation depends on a few divisions of both tissues seem to be mainly supply of seed-derived auxin (Kassemeyer and under the control of maternal genes, which Staudt, 1983; O’Neill, 1997). Therefore, the may give the mother plant some control over development of grape berries is usually depen- which embryos to develop (Raghavan, 2003). dent on pollination, fertilization, and develop- The fertilized ovule develops into the seed, ment of at least one seed (Pratt, 1971). Ovaries whose embryo starts to produce and release that do not contain at least one fertilized ovule auxin to pericarp tissues, where it stimulates or whose seed development is aborted early on gibberellin synthesis (O’Neill, 1997; Pandolfini are abscised from the fruit cluster (Nitsch et al., et al., 2007; Serrani et al., 2008). Gibberellin 1960). Arrested ovule development prior to may also be produced in and released by the anthesis prohibits fertilization and also results seeds upon stimulation by embryo-derived in abortion if all four ovules fail to develop auxin (Dorcey et al., 2009). Auxin reactivates (Kassemeyer and Staudt, 1982). Later degenera- cell division, and gibberellin induces cell tion of the embryo and endosperm can result in expansion (Serrani et al., 2007). Thus, fruit set apparently normal-looking but hollow (and is dependent on both auxin and gibberellin: thus nonviable) seeds termed “floaters” These two hormones induce the pistil to because they float in water, unlike normal develop into the fruit and to differentiate an seeds, which are thus named “sinkers” (Ebadi exocarp (skin) and mesocarp (flesh or pulp). et al., 1996). Although auxin induces gibberellin produc- Nonetheless, some cultivars can produce tion, auxin synthesis might be stimulated by apparently normal seedless berries. This is pollen-derived gibberellins and is necessary called parthenocarpic (Greek parthenos ¼ vir- for embryo formation and fruit tissue produc- gin, karpos ¼ fruit) fruit development, when tion and continues until the embryo is mature. the berries develop without fertilization Therefore, the interaction, also called cross talk, (although typically not without pollination) between auxin and gibberellin plays a key role and lack seeds completely, or stenospermo- in the ovary’s commitment to initiate fruit carpy, when the berries contain at least one fer- development and growth (Gillaspy et al., 1993; tilized but subsequently aborted seed so that in Ozga and Reinecke, 2003). The promoting effect many cases all that remains is a small, soft, white of gibberellin on cell volume in seedless berries rudimentary seed or seed trace (Barritt, 1970; 2.3. REPRODUCTIVE CYCLE 77
Ledbetter and Ramming, 1989; Olmo, 1936; intensity, warm temperature, and adequate soil Pearson, 1932; Roytchev, 1998, 2000; Staudt moisture and nutrient availability. This is the and Kassemeyer, 1984). Stenospermocarpy is a likely reason why years in which the shoots quantitative trait: The seed traces vary in size carry more clusters than usual often follow and hardness over a wide range (Cabezas et al., years with above-average fruit set. Adverse 2006; Ledbetter and Shonnard, 1991). Cultivars environmental conditions (e.g., cloudy, wet, with parthenocarpic fruit (e.g., Black Corinth) cool, or hot bloom period and water or nutrient are often used to produce seedless raisin stress), insufficient or inefficient leaf area (e.g., grapes, whereas most so-called seedless table due to hail or insect or disease attack), shade grapes (e.g., Sultana; synonym Thompson created by dense canopies, or excessively vigor- Seedless) are stenospermocarpic. Intriguingly, ous shoot growth (competing with inflores- the embryos of stenospermocarpic grapes cences for assimilates) often result in poor remain alive and can be rescued and grown fruit set and loose clusters (see Chapter 6.1). into full plants using in vitro techniques, which Because, as a perennial species, the grapevine permits crossing of seedless table grapes to does not depend on producing seeds each year, breed new cultivars (Cain et al., 1983; Emershad such “self-thinning” enables it to adjust its and Ramming, 1984; Emershad et al., 1989). reproductive output to available resources Because the lack of growth hormone production without jeopardizing survival of the plant. By in stenospermocarpic fruit limits cell expan- culling a portion of its reproductive structures, sion and hence berry size (Coombe, 1960; the vine may enable the remaining seeds to Nitsch et al., 1960), gibberellin sprays are often attain adequate size and maturity. applied after fruit set to increase the size of Poor fruit set due to excessive abortion of table grapes with the added benefit of decreased flowers and ovaries is also termed shatter, cluster compactness due to the elongation of shedding, or “coulure” (French couler ¼ to leak, the rachis. to fall off). Loss of such ovaries can occur for The proportion of flowers that develop into up to 4 weeks after anthesis (Kassemeyer and berries following anthesis is typically in the Staudt, 1983; Staudt and Kassrawi, 1973). Fertil- range of 20–50% and is inversely related to ization of at least one of up to four ovules is the number of flowers per inflorescence. Fruit required for the ovary to develop into a berry; set is highly variable between cultivars and is hence, the abscised ovaries typically do not strongly modulated by environmental condi- contain any fertilized ovules, although the tions and rootstocks (Alleweldt and Hofa¨cker, vast majority of them have been pollinated 1975; Keller et al., 2001a; Schneider and Staudt, (Kassemeyer and Staudt, 1982; Staudt and 1978). Cultivars that are susceptible to environ- Kassemeyer, 1984). Normal fruit set despite mental conditions leading to poor fruit set inadequate seed development in a portion of include Merlot, Grenache (synonym Garnacha), the fertilized berries, on the other hand, is and Gewu¨ rztraminer, whereas cultivars such as called “millerandage” and results in clusters the Pinots, Chardonnay, or Sylvaner seem to be having the appearance of “hens and chickens” much less susceptible. The process of meiosis (Fouge`re-Rifot et al., 1995; Galet, 2000; May, appears to be particularly sensitive to stress 2004). The term “hens” refers to large, normal conditions that curtail the supply of carbohy- berries with at least one viable seed, and drates to the inflorescences. Optimum condi- “chickens” describes small berries with tiny, tions for flowering and fruit set are thought to degenerated seeds (i.e., seed traces) that often be similar to those governing inflorescence ini- lack endosperm (Ebadi et al., 1996; Staudt and tiation (see Chapter 6.1), namely high light Kassemeyer, 1984). Although the expansion of 78 2. PHENOLOGY AND GROWTH CYCLE
“chickens” is arrested within approximately with naturally tight clusters can be manipu- 3 weeks after anthesis due to seed abortion, lated using viticultural practices that tend to these berries apparently ripen normally. By decrease cluster compactness, although such contrast, so-called “shot berries” are essentially practices usually come at the cost of decreased arrested at the ovary stage; remain small, yield. Potential strategies include light or no green, and hard; and are high in tannins pruning resulting in large numbers of clusters (Fouge`re-Rifot et al., 1995). The term is unfortu- and, therefore, relatively poor fruit set (see nate: Because these ovaries fail to differentiate a Chapter 5.1); removing a portion of the leaves discernible mesocarp and exocarp, shot berries before or soon after bloom, which also are technically not berries at all. They usually diminishes fruit set (Candolfi-Vasconcelos and contain at least one sterile ovule (e.g., due to Koblet, 1990; Poni et al., 2006); or cutting incomplete or failed meiosis) in addition to through inflorescences around bloom time degenerated fertile but unfertilized ovules. Per- (i.e., removing the distal cluster portion), which haps shot berries are prevented from abscising leads to elongation of the remaining rachis. by the initial, although very limited, growth of In seeded grapes, fertilization is followed by the nucellus of the sterile ovule (Kassemeyer a second period of cell division (the first period and Staudt, 1982). Coulure results in very is before bloom and determines the cell number loose clusters, and millerandage leads to highly of the ovary), which is probably stimulated by variable seed numbers and berry sizes, a brief increase in cytokinin in addition including seedless berries, on the same cluster. to auxin produced in the developing seeds. Variations in the degree of fruit set together The berry cytokinin concentration, it seems, with the extent of inflorescence branching and declines after anthesis throughout berry devel- variations in the rate and/or extent of rachis opment (Chacko et al., 1976). During the first 3 cell expansion during cluster development weeks after anthesis, the polar nucleus of the before and after bloom lead to wide variation seeds’ endosperm divides several times before in cluster size and architecture, both between the endosperm becomes cellular so that endo- species and cultivars and within them (e.g., sperm nuclei become polyploid (Kassemeyer due to environmental influences). Such differ- and Staudt, 1983). By contrast, the zygote cells ences in morphological features have implica- giving rise to the embryo only begin dividing tions not only for yield formation (see Chapter approximately 2–4 weeks after anthesis. The 6.1) but also for the clusters’ susceptibility to rate and duration of cell division in the berry fungal attack. For instance, cultivars with tight pericarp are controlled by the seeds (i.e., the clusters (e.g., Chardonnay, Riesling, and Pinot) embryos) so that berries containing more seeds have small rachis cells and a short, compact will become larger than berries with fewer branching pattern. They are much more sus- seeds (Gillaspy et al., 1993). Whereas the cell ceptible to infection by Botrytis cinerea than number successively doubles approximately those with loose clusters (e.g., Cabernet Sau- 17 times before anthesis, only one or two more vignon, Syrah, and Thompson Seedless). This doublings happen after anthesis (Coombe, is probably because tight clusters have their 1976). The mesocarp cells cease dividing 3 or berries packed into a small volume and hence 4 weeks after anthesis (the innermost cells stop retain more rain (or irrigation) water and dry dividing within 2 weeks), whereas skin cells down more slowly than loose clusters. In addi- may continue to divide for up to 5 weeks after tion, berries touching and rubbing on each anthesis (Considine and Knox, 1981; Harris other may lose epicuticular wax and hinder et al., 1968; Nakagawa and Nanjo, 1965; Pratt, uniform fungicide coverage. Yet even cultivars 1971). Cell division is completed before the 2.3. REPRODUCTIVE CYCLE 79 berry enters a lag phase of slow or no fresh Chapter 6.2). In ripe grape berries, cell wall weight accumulation. Although it is frequently thickness ranges from >1 mm in the exocarp stated that berry growth is divided into a phase to <0.1 mm in the mesocarp and, if grapes of cell division and one of cell expansion, this is behave like tomatoes, the resistance of both not accurate. Growth is always by cell expan- the epidermis and the cuticle against defor- sion because plants (like any other organism) mation, their so-called modulus of elasticity or cannot add fully grown cells like bricks to a elastic modulus increases strongly during fruit wall. Although dividing cells usually remain ripening (Bargel et al., 2006). As first described small and have small vacuoles, cell division is by Winkler and Williams (1935), the growth or accompanied and followed by cell expansion increase in fresh weight of seeded berries traces (Gillaspy et al., 1993; Pratt, 1971). As the cells a double-sigmoid pattern; that is, the growth expand, the cell wall becomes thinner and the curve is bent in two directions, like the letter vacuole becomes predominant. Therefore, the S, and can be divided into three more or less vast majority of the volume gain of growing distinct phases or stages (Figure 2.11): berries is due to cell expansion (i.e., an increase in vacuole volume; see Chapter 3.1), especially Stage I: The first rapid increase in size of the in the mesocarp, which is stimulated by seed- seeds and pericarp. The first part of seed produced auxin. The lag phase is followed by development is termed morphogenesis and another period of cell expansion so that the vol- consists of embryo formation ume of mesocarp cells may increase more than (embryogenesis) and endosperm growth, 300-fold from anthesis to maturity (Coombe, but there is little growth of the embryo 1976) or 15-fold between fruit set and maturity. (Ebadi et al., 1996; Nitsch et al., 1960). The At the same time, the fruit surface area may seeds are green, and the berry is green and increase approximately 400-fold from anthesis hard and accumulates organic acids (mainly to veraison and almost doubles again from ver- tartrate and malate) but little sugar. The aison to maturity (Considine and Knox, 1981). growth hormone auxin, the cell division Cell division seems to be mostly under hormone cytokinin, and the cell expansion genetic control (i.e., fruit from different culti- hormone gibberellin (GA) produced by the vars should have different cell numbers), embryos and released into the pericarp reach whereas cell expansion is predominantly high concentrations during this period driven by environmental factors (i.e., berries (Iwahori et al., 1968; Sakakibara, 2006; of the same cultivar grown under different con- Scienza et al., 1978). Auxin and gibberellin ditions should have different cell sizes). This are also exported in the phloem to the means that fruit size can vary due to variations pedicel and peduncle, where gibberellin in both cell number and cell size. It also means induces cell elongation and auxin induces that cell division is comparatively immune to the production and differentiation of environmental influences so that viticultural vascular bundles to ensure that the transport attempts to manipulate berry size will have to capacity of the vascular system does not aim mostly at cell expansion, whereas breeding limit growth of the developing berry. In efforts should address cell division. The maxi- addition, import from the “mother” vine of mum extent of cell expansion, and hence berry the germination-inhibiting hormone ABA size, is limited by the elastic properties of the prevents seed abortion and promotes normal skin, which in turn may be related to the thick- embryo development (Nambara and ness and/or stiffness of the exocarp cell walls Marion-Poll, 2005). This phase lasts 6–9 or the cuticle (Matthews et al., 1987; see also weeks and ends when the skin cells stop 80 2. PHENOLOGY AND GROWTH CYCLE
FIGURE 2.11 Cabernet Sauvignon berry growth over 3 years (left; note 50% anthesis occurred on day 160 in year 1 and on day 150 in years 2 and 3; M. Keller, unpublished data), and beginning of ripening (veraison) in Concord clusters with berries displaying various stages of color change (right; photo by M. Keller).
dividing. By this time, the berry has attained gibberellin production in the seeds at least half its maximal size and fresh (Gutierrez et al., 2007; Nambara and Marion- weight. Poll, 2005). At the same time, there seems to Stage II: The lag phase during which the seed be a fleeting increase in ethylene (though far enters its maturation stage (just after the less pronounced than in so-called climacteric embryo’s genes take over control from the fruits; Chervin et al., 2004) that is mother vine), the embryo grows rapidly, and accompanied by heightened sensitivity to by 10–15 days before veraison the seeds ethylene (Chervin et al., 2008) and a dramatic reach their final size, maximum fresh and persistent rise in brassinosteroids (Pilati weight, and maximum tannin content et al., 2007; Symons et al., 2006), a class of (Adams, 2006; Niimi and Torikata, 1979; growth-promoting hormones probably Pratt, 1971; Ristic and Iland, 2005, Viala and produced in the epidermis. These two Vermorel, 1909). Moreover, the innermost hormones may trigger the changes in cell cells of the outer integument lignify, and wall chemistry that lead to berry softening the outermost cell walls thicken to form a and expansion. Berry turgor pressure also cuticle; the integuments now restrict further declines approximately 10-fold during this seed expansion. Although the pericarp period (Thomas et al., 2006). The length of the grows only insignificantly, the concentration lag phase (1–6 weeks) largely depends on of the growth hormone auxin (and ripening the cultivar and is important in determining inhibitor; Davies et al., 1997) peaks during the time of fruit maturity; late-ripening this phase and then declines sharply cultivars seem to have a long lag phase (Alleweldt and Hifny, 1972; Alleweldt et al., (Alleweldt and Hifny, 1972; Currle et al., 1975; Nitsch et al., 1960). The influx of ABA 1983). Seedless grapes tend to have a less increases toward the end of this phase, pronounced lag phase than seeded grapes suppressing embryo growth by blocking (Iwahori et al., 1968; Nitsch et al., 1960). 2.3. REPRODUCTIVE CYCLE 81
Stage III: The ripening period, which lasts expansion (Matthews et al., 1987). Red, 5–10 weeks. Its onset, which typically occurs purple, or blue pigments called over a period of 7–10 days within a grape anthocyanins accumulate in the exocarp cluster, is termed veraison and signals a (and in “teinturier” cultivars, such as fundamental shift from partly Alicante Bouschet, also in the pericarp), and photosynthetic to wholly heterotrophic sugars (glucose and fructose) accumulate in (Greek heteros ¼ other, different; trophe´ ¼ the pericarp, whereas organic acids (malate) nutrition) metabolism. Veraison is marked and chlorophyll are degraded. Chloroplasts by berry softening and an increase in sugar are converted into chromoplasts. Unlike in content, followed by a rapid change in skin climacteric fruits, the “fruit-ripening” color from green to red, purple or blue in hormone ethylene does not appear to play a dark-skinned cultivars (see Figure 2.11), and prominent role in grape ripening, and berry to more or less yellow in some white respiration apparently declines during this cultivars. Softening (i.e., a decrease in fruit period (Coombe and Hale, 1973; Harris et al., firmness or increase in deformability) 1971). Embryo growth stops early in this mainly arises from the disassembly of the phase (Nitsch et al., 1960; Pratt, 1971), the mesocarp cell walls (Huang and Huang, seeds turn from green to yellow and finally 2001). Although the cell walls remain brown due to oxidation of tannins in the functionally intact, they become more open, parenchyma cells of the outer integument, more hydrated, less flexible, and weaker so and they become hard and desiccated that they can no longer hold the cells (Adams, 2006; Ristic and Iland, 2005). Storage together very well (Brummell, 2006). Water reserves,suchasstarchandlipids,and import through the xylem declines gradually mineral nutrients are accumulated in the while the berry changes color, and water endosperm until the seeds become dormant import via the phloem increases rapidly (Rogiers et al., 2006). Thus, the seed fresh (water being the solvent for the incoming weight declines while the dry weight may sugar). The xylem now probably serves continue to increase until the berry reaches mostly to recycle excess phloem water back its maximum size and weight; seeds that out of the berry (Keller et al., 2006; Patrick, are hard and brown and have attained 1997). Berry turgor pressure is low (<0.5 bar) their maximal dry weight are called but remains positive and relatively constant “mature.” The concentration of ABA throughout ripening (Lang and Du¨ ring, increases rapidly and peaks in the seed 1990; Thomas et al., 2006). This last phase is and berry flesh during seed maturation, characterized by a further increase in berry decreases due to degradation during seed size, which is somewhat unusual because desiccation, and is relatively low in mature the fruit of most other species reach their seeds (Inaba et al., 1976; Lund et al., 2008; final size before ripening begins (Gillaspy Scienza et al., 1978). ABA, which now et al., 1993). The volume increase is initially accumulates mainly as a result of glucose- very rapid but slows progressively toward induced production by the embryo, helps fruit maturity; berry size may plateau or coordinate seed maturation and is necessary decrease due to evaporative water loss to induce seed dormancy, in which during later stages of ripening. The metabolism ceases until the seed is extensibility of the skin temporarily rehydrated for germination (Gutierrez et al., increases at veraison to accommodate berry 2007; Nambara and Marion-Poll, 2005; Wobus expansion, but the skin later restricts further and Weber, 1999). 82 2. PHENOLOGY AND GROWTH CYCLE
In contrast to seeded grapes, seedless grape degradation, facilitates breakage of the seed berries often do not show clearly discernible coat, and induces remobilization of nutrients phases of growth, and their volume increases from the endosperm (Nambara and Marion- much more “smoothly” (Nitsch et al., 1960; Poll, 2005; Olszewski et al., 2002). Winkler and Williams, 1935). The length of The evolution of a fleshy, sweet, and dark- each stage of fruit growth and the final berry skinned berry during the late Cretaceous size depend on the cultivar but are strongly period approximately 100 to 65 million years modified by rootstocks and environmental ago was probably an adaptation to the simul- conditions. Optimum conditions for rapid taneous evolution of birds and mammals berry development are similar to those gov- (Hardie, 2000). Birds are the predominant seed erning the other phases of reproductive dispersers of European wine grapes, whereas development: high light intensity, warm tem- bears are major dispersers of American juice perature, and adequate soil moisture and grapes, whose distinctive aroma birds find nutrient availability (see Chapter 6). unattractive. The only role grape berries play Harvest of grapes by birds or mammals, in the life of a vine is to spread its genes including humans, normally completes the (DNA). The ability to advertise seed maturity reproductive cycle. If grapes remain on the by visual and olfactory cues and reward seed vine for an extended period, the continued cell vectors with energy-rich sugar greatly wall disassembly leading to diminished cell increases the chances for gene dispersal. As wall strength and loss of intercellular adhesion Charles Darwin (2004) remarked, a fruit’s eventually results in tissue failure, cell separa- "beauty serves merely as a guide to birds and tion, and senescence (Considine and Knox, beasts in order that the fruit may be devoured 1979a), which sets the seeds “free”—often and the manured seeds disseminated." Berries aided by infection by opportunistic fungi such accumulate pigments and aroma volatiles to as B. cinerea and others (Brummell, 2006). advertise to potential seed vectors their con- Seeds are ready to germinate at or soon after tent of nutritionally valuable components, veraison (Pratt, 1971; Winkler and Williams, such as sugars, amino and fatty acids, vita- 1935); in fact, it is thought that fruit ripening mins, and antioxidants (Goff and Klee, 2006). is not initiated until seed maturation has been These compounds, in turn, constitute the completed (Cawthon and Morris, 1982; “ticket price” to cover the costs of transporta- Gillaspy et al., 1993). Whereas seeds may ger- tion by means of wings and legs of seed minate immediately if postveraison berries dispersers. Low concentrations of (volatile) drop in a moist environment, they lose this ethanol produced by yeasts, which have been ability within a day upon removal from a living on and in fruits since the Cretaceous, berry, when their water content drops from may also form part of the aroma bouquet of approximately 50% to 20% (Currle et al., ripe berries (Dudley, 2004). Although fruit 1983). Once dehydrated, the seeds require a color is certainly an important attractant dur- stratification period of 4–12 weeks at chilling ing the day, volatiles may be more important temperatures (0–5 C) and 40–50% relative at night, when colors are difficult to see. At humidity to overcome embryo dormancy. thesametime,tanninsandothersecondary Light (especially red light) induces water compoundsmaydetermicrobesthatalso imbibition and subsequent germination by would like to consume the fruit but do not dis- stimulating the embryo cells to produce gib- perse the seeds (Levey, 2004). Likewise, the berellin and expand rapidly, which overcomes high acidity and unpleasant aroma (caused the ABA-related dormancy by accelerating ABA by methoxypyrazines and other volatile 2.3. REPRODUCTIVE CYCLE 83 compounds) of unripe berries prevents the characteristic accumulation of sugars and berries from being eaten before the seeds are other compounds also means that fruit ripen- mature. This strategy enables grapevines to ing is different from the senescence process explore new habitats beyond the small area in leaves that involves disassembly of proteins around an existing immobile plant that simply (called proteolysis) and recycling of nutrients drops its fruit (seed) to the ground. The to other plant parts (Gillaspy et al., 1993). CHAPTER 3
Water Relations and Nutrient Uptake
OUTLINE
3.1. Osmosis, Water Potential, and Cell 3.3. Water and Nutrient Uptake and Expansion 85 Transport 92 3.2. Transpiration and Stomatal Action 89
3.1. OSMOSIS, WATER POTENTIAL, AND CELL EXPANSION channel proteins, called aquaporins, in the mem- branes (Katsuhara et al., 2008; Maurel et al., 2008; Tyerman et al., 1999). In contrast to the diffusion Between 70 and 95% of a grapevine’s fresh pathway, the rate (but not the direction) of water mass consists of water (H2O). Even the woody tis- transport across aquaporins can be fine-tuned by sues of trunk and roots have a water content of active regulation because these proteins are able approximately 60%. Most of this water serves as to open or close their pores in a process called gat- a solvent for ions and organic molecules in the ing. Additional regulation is provided by varying interior of the vine’s cells. Water can diffuse freely, the number of aquaporin proteins in the mem- albeit slowly, across the phospholipid bilayer of branes. Aquaporins are relatively temperature cell membranes, which are approximately 5–10 insensitive and can speed up water diffusion nm thick, and diffusion accelerates with rising across membranes more than 10-fold; in other temperature. More important, however, H2O words, they strongly raise the membranes’ molecules are small enough (0.3 nm diameter) to (osmotic) water permeability (also called hydrau- be able to just barely (i.e., in single file), but lic conductivity). At the same time, most of them extremely rapidly (up to 1 billion H2Omolecules are impermeable to mineral nutrient ions and per second), penetrate pores formed by water even to protons. In fact, cells use protons, in
The Science of Grapevines 85 Copyright # 2010 Markus Keller. Published by Elsevier Inc. All rights reserved. 86 3. WATER RELATIONS AND NUTRIENT UPTAKE
þ addition to calcium ions (Ca2 ) and removal of thermodynamic (or absolute) temperature in phosphate groups (so-called dephosphorylation) Kelvin (K ¼ C þ 273). from the protein, to close aquaporins (Johansson Sugars such as sucrose, organic acids such as et al., 2000; Luu and Maurel; 2005, Maurel et al., malate, and inorganic ions such as potassium þ 2008; Tournaire-Roux et al., 2003). In addition to (K ) and chloride (Cl ) are the major osmotic size exclusion (i.e., filtration), dissolved ions are solutes (osmolytes) of plant cells. Thus, solutes also repelled by charged residues at the entrance have an osmotic function in addition to their þ to the pores; thus, cations cannot pass through metabolic functions. For instance, cells use K a “gate” guarded by a positive charge, and anions to neutralize the negative charge of inorganic cannot pass through a gate guarded by a negative and organic anions (i.e., organic acids) and charge. In other words, membranes are selec- anionic groups of macromolecules because the tively permeable or semipermeable (Latin semi total charge of all anions and cations inside a ¼ half). Specialized gates consisting of transport cell must balance (Clarkson and Hanson, proteins and ion channels are required for 1980). When solutes attract water into a cell, molecules other than H2Otopassthrough the cell swells and causes the cell membrane membranes (see Chapter 3.3). to exert force on the cell wall. This, in turn, Because the concentration of dissolved ions leads to a balancing wall pressure, which raises (from dissociated salts) and organic molecules the energy of the water inside the cell until it inside a cell is almost always higher than that of equals that of water outside. At this point, the the exterior (i.e., cell walls and intercellular space), cell’s internal hydrostatic pressure (more there is relatively less water inside the cell than out- accurately, the pressure difference between side (i.e., the cellular water is “diluted” by the inside and outside of the cell), termed turgor solutes). This water concentration gradient causes pressure (P) but often abbreviated as turgor water to move passively through aquaporins and (Latin turgere ¼ to be swollen), is equal to the across the membrane into the cell, and this move- difference in osmotic pressure between the cell P ¼ p p ment is termed osmosis (Greek osmos ¼ impulse and its surroundings (i.e., inside outside), or thrust). In other words, the presence of solutes and net water influx stops (Figure 3.1). Thus, in a cell exerts a “pull” on the water molecules sur- osmotic pressure is defined as the hydrostatic rounding the cell, and this tension is termed pressure required to stop the net flow of water osmotic potential. The greater the concentration across a membrane separating solutions of of solutes, the more water will move into the cell differing concentration. In other words, osmo- to restore (hydraulic) equilibrium. This leads to a sis can only work as a driving force for water buildup of pressure due to the incompressibility flow (see Chapter 3.3) when two compartments of water and the rigidity of plant cell walls resisting are separated by a semipermeable membrane. an increase in volume. Osmotic pressure (p, Turgor pressure, in turn, is the osmotically expressed in MPa) of an aqueous solution is thus caused by dissolved solutes and increases linearly as the solute concentration (c, expressed in mol L 1; i.e., the number of dissolved molecules per unit volume) increases, and also as the temperature (T) increases, according to the following equation: FIGURE 3.1 During diffusion, solutes (red) and water p ¼ RTc (blue) move in opposite directions (left), whereas during osmosis water moves across a semipermeable membrane R ¼ 1 where is the universal gas constant ( 8.31 J K (brown) to balance the water concentration deficit created 1 1 1 mol ¼ 0.00831 MPa L K mol ), and T is the by the solutes (right). 3.1. OSMOSIS, WATER POTENTIAL, AND CELL EXPANSION 87 maintained hydrostatic pressure inside living as their natural trellis system (see Figure 1.2). plant cells. The turgor pressure of growing Pure water has the highest possible water cells typically reaches approximately 0.3–1 potential, which by convention equals zero MPa—that is, 3–10 times atmospheric pressure (C ¼ 0) at atmospheric pressure and 25 C. P ( a 0.1 MPa) (Cosgrove, 1997). Opening sto- Therefore, the water potential of aqueous matal guard cells can generate over 4 MPa; solutions is always negative (C < 0). In other nevertheless, the deformation of these cells words, C is a measure of water concentration, and their cell walls is completely reversible and p is a measure of how much the dissolved (Franks and Farquhar, 2007; McQueen-Mason, molecules, termed solutes, decrease C. The C 2005; Roelfsema and Hedrich, 2005). of a cell describes the potential tension (nega- The solute and pressure forces in plants and tive pressure or suction) that the cell solution soils are conveniently described as free energy exerts on surrounding pure water. In the per unit volume (i.e., equivalent to force per unit absence of other forces, the movement of water area, or pressure) that is generated by the during osmosis is always from a region of lower random movement and collision of molecules. solute concentration (i.e., greater water poten- The free energy of water or, more accurately, of tial) to one of higher solute concentration (i.e., an aqueous solution, per unit molar volume lesser water potential). In other words, osmosis of liquid water ( 18 mL mol 1), is termed water is driven by a water potential gradient. potential (C, expressed in J m 3 or as its pressure Plant cells use osmotic solutes, especially equivalent Pa ¼ 10 5 bar) and is the sum of the sucrose and potassium, although in ripening component potentials arising from the effects of grape berries it is predominantly the hexoses C ¼ P turgor pressure (i.e., turgor potential, P ) glucose and fructose with smaller amounts of and solutes (i.e., solute potential or osmotic potassium, to lower their C in order to attract potential, Cp ¼ p) in addition to interactions water into the cells. Therefore, solutes have a with matrices of solids (e.g., cell walls) and major role in building P, which is necessary C macromolecules (i.e., matrix potential, M) and for cell expansion and hence plant and organ effects of gravity (i.e., gravitational potential, growth. In fact, cell growth is mainly a function C G), as described by the following equation of turgor and the mechanical properties of the (Boyer, 1969): cell walls. Cell expansion is made possible by cell wall loosening and accumulation of solutes C ¼ P p þ C þ C ¼ C þ Cp þ C þ C M G P M G inside the cell vacuole, which draws water into Although the matrix potential is very impor- the vacuole to create turgor pressure. Thus, tant in soils (due to the adhesion of H2O mole- water movement into cells (i.e., across mem- C þ C cules to soil particles) and cell walls, it is very branes) is determined by P p (i.e., it is close to zero inside the cells of well-watered driven by both pressure and osmotic compo- plant tissues and is therefore negligible unless nents). In contrast, water movement outside of the tissue is severely dehydrated (e.g., loss of cells (i.e., through the apoplast; see Chapter 50% of tissue water; see Hsiao, 1973). The 3.3) is driven solely by pressure differences gravitational potential increases with height at (because water does not cross membranes in a rate of 0.01 MPa m 1. Although this may be the apoplast). The P of growing cells is typi- of little significance in the relatively short- cally approximately 0.3–1 MPa, and this pres- statured cultivated grapevines that rarely sure irreversibly extends (i.e., stretches) and exceed a height of approximately 2 m, the “pushes” the plastic cell walls outward C P downward pull of G is a factor in their taller (Cosgrove, 1997). Nonetheless, although causes wild relatives, as well as in the trees serving tension, it does not induce extension of the cell 88 3. WATER RELATIONS AND NUTRIENT UPTAKE walls, which is controlled by the cross-linking which are then converted by the enzyme of xylem-delivered calcium with pectic acids peroxidase into hydroxyl radicals (•OH), which produced by the expanding cells (Boyer, 2009; in turn may attack the bonds between cell wall Schopfer, 2006). Although P must be high polysaccharides (M ller et al., 2007). Slipping of enough to expand the cell, it must also be the parallel microfibrils and production and low enough to sustain water influx (i.e., maintain insertion of new wall material is necessary for a gradient of C). Continued water influx is the expanding cell to enclose its increasing necessary in expanding cells because the incom- volume; cell walls usually maintain their thick- pressibility of water means that P would ness during expansion (Boyer, 2009; Cosgrove, otherwise drop to zero as soon as the cell walls 2005). Because cells must increase in volume become stretched. Indeed, a decrease in P by before they can divide, cell expansion, in turn, is as little as 0.02 MPa can stop cell expansion but also a prerequisite for cell division. As much as not reverse it (Taiz, 1984). 90% of the entire volume gain during plant Cell expansion also requires loosening of the growth is due to the expansion of cell vacuoles cell walls of turgid cells before they can be and concomitant stretching and addition of cell extended (Boyer, 2009; Schopfer, 2006). Wall wall material. Hence, most of the increase in loosening, and thus wall stress relaxation, is volume during cell expansion is due to water thought to be achieved by the “relaxing” activ- uptake by the vacuole, whereas the amount of ity of proteins known as expansins that are cytoplasm increases only insignificantly. activated at low pH (Cosgrove, 1997, 2000, The term water potential is convenient 2005; Li et al., 2003). According to the “acid because it can be applied not only to (cell) growth theory,” the cell wall is acidified by solutions but also to soil solution and air—that þ protons (H ) pumped from the cell interior to is, to any medium with variable water content. the cell wall in exchange for potassium Water potential indicates the availability of þ ions (K ) that maintain the electrical charge water in any aqueous system. The water poten- balance inside the cell (Rayle and Cleland, tial of air can be estimated from the air’s rela- 1992; Stiles and Van Volkenburgh, 2004; Van tive humidity (RH; expressed in percentage). Volkenburgh, 1999). As a consequence, the cell Relative humidity is the amount of water vapor wall pH can temporarily decrease from the in air at a particular temperature in relation to normal 5.5–6 to as low as 4.5 (compared with the total amount the air could hold at that tem- a cytosol pH of 7.5; Kurkdjian and Guern, perature (i.e., the saturation vapor density or 1989), which equates to a more than 10-fold saturation vapor pressure). The water-holding þ increase in H in the cell wall. The ATP- capacity of air (and thus its saturation vapor þ powered H pumps in turn are activated pressure) increases exponentially with incre- by the growth hormone auxin and light (via asing air temperature. In the 10 to 35 C range, phytochromes and other photoreceptors; see an increase in air temperature of 12 C doubles Chapter 5.2). Expansins may temporarily the water vapor concentration of saturated air; disrupt the chemical associations (hydrogen that is, air at 22 C can hold twice as much bonds) between the parallel cellulose micro- water vapor as air at 10 C, and air at 34 C can fibrils and the cross-linking xyloglucans hold four times as much. In other words, if (hemicelluloses) in the cell walls, which makes saturated air (100% RH) is heated by 12 C, its the walls flexible enough to allow expansion RH drops to 50% (i.e., the air develops a satu- C and insertion of new cell wall material. In addi- ration deficit). The relationship between air tion, auxin also seems to stimulate the production (in MPa) and RH (in %) can be approximated of reactive oxygen species (see Chapter 7.1), by the following equation: 3.2. TRANSPIRATION AND STOMATAL ACTION 89
TABLE 3.1 Typical Values of RH and C for Leaves leaf depends on two major factors: the differ- and Atmosphere at 25 C ence in absolute water vapor concentration, or water potential, between the leaf air spaces RH (%) C (MPa) C C ( leaf) and the external air ( air) and the resis- r Air spaces in leaf 99 1.4 tance ( ) of this pathway to diffusion. As water evaporates from the mesophyll cell walls and is Boundary layer 95 7.0 discharged as water vapor into the atmosphere Atmosphere 100 0 (see Figure 1.14, Chapter 1.3), it meets a series 90 14.5 of resistances that work in the opposite 50 95.1 direction of its movement. A resistance slows 10 315.9 the H2O molecules down and thus limits their rate of diffusion. The major ones are the resis- tance at the stomatal pore and the resistance C : T ð = Þ due to the thin film of still and relatively moist air 0 46 ln RH 100 air at the surface of the leaf (the so-called where T is the thermodynamic (or absolute) boundary layer), but the stomatal resistance r temperature in K. ( s) is usually much greater than the boundary- C r As this equation shows, of saturated air is layer resistance ( b). The boundary layer (also always zero (ln1 ¼ 0), regardless of temperature, known as the unstirred layer) acts like an whereas C of drying air drops dramatically additional (cuticular) membrane in series, but (Table 3.1). Even relatively humid air exerts an wind increases turbulence in the boundary r enormous tension on the water inside a plant layer, making it thinner and decreasing b, which r or in the soil. This tension is responsible for increases transpiration dramatically even as s transpiration (see Chapter 3.2). In other words, declines (Freeman et al., 1982). On the other r transpiration is also driven by a water potential hand, b can be very high in dense canopies so gradient. Incidentally, the relationship between that transpiration of such canopies may be C r temperature and air is also responsible for the determined mainly by b and is not easily so-called rain shadow effect of the Cascades influenced by atmospheric conditions. Trans- mountain range in eastern Washington and piration rate per unit leaf area (E; expressed in Oregon and the Andes range in Mendoza mol m 2 s 1) increases as the leaf-to-air vapor (Argentina). As moist air from the Pacific Ocean concentration difference (i.e., the driving force) rises up the western slopes of the mountains, increases, and it decreases as the total resistance it cools and drops excessive moisture as rain- increases. This relationship can be described by fall or snow so that the saturation deficit (or the following equation, which is analogous to evaporative demand) of the now drier air Ohm’s law (I ¼ Vr 1,whereI is the current, increases further as the air warms down the and V is the electrical potential difference or eastern slopes of the mountains. voltage) used in electricity and thermodynamics: E ¼ DCðr þ r Þ 1 s b 3.2. TRANSPIRATION AND DC ¼ C C STOMATAL ACTION where leaf air. Water vapor concentration is proportional to water vapor pressure, and the difference in The evaporation of water from plants is water vapor pressure (actually the vapor termed transpiration (Latin trans ¼ on the other pressure difference divided by the atmospheric side, spirare ¼ to breathe). Transpiration from a pressure) between the inside and the outside 90 3. WATER RELATIONS AND NUTRIENT UPTAKE of the leaf (i.e., the water potential gradient, DC) causes stomata to close partially to protect the is often called the vapor pressure deficit. As dis- xylem conduits from cavitation (see Chapter cussed in Chapter 3.1, an increase in air tempera- 3.3). This effect can override the influence of ture strongly increases the air’s water-holding high light intensity on stomatal opening so that capacity and thus results in a greater vapor even relatively mild soil water deficit in a vine- pressure deficit. The steeper concentration gradi- yard often leads to a midday depression in ent stimulates water evaporation from the leaf stomatal conductance (Du¨ring and Loveys, g mesophyll. Therefore, temperature is an impor- 1982). This temporary decline in s is usually tant determinant of the transpiration rate, and not observed in well-watered vines, except for transpiration increases linearly with increasing vines growing in hot climates (Downton et al., temperature. 1987; Williams et al., 1994). The aperture of the In a typical grapevine leaf, only 5–10% of the stomatal pores also decreases as air humidity total transpiration occurs across the cuticle, declines (i.e., as the vapor pressure deficit although the cuticular portion can reach up to between leaf and air increases), although this 30% in old leaves that have low overall transpi- could be a response to transpiration rather than ration rates compared with younger leaves to humidity per se. As would be expected from (Boyer et al., 1997). The remainder of the water the preceding “transpiration equation,” this vapor escapes through the stomatal pores, even stomatal closure can compensate for variations C though they cover less than 5% of the leaf sur- in humidity, holding transpiration and leaf face. Stomata are thus very important for regu- constant. Nevertheless, a decrease in relative lating diffusional water loss from a leaf. They humidity from 85 to 45% (at constant tempera- function like a pressure regulator, limiting ture) can decrease the transpiration rate by C pressure changes (i.e., leaf) by controlling flow more than 80% via partial stomatal closure. rate (transpiration), and respond very sensi- A rise in temperature initially increases stomatal tively to environmental variables (Sperry et al., aperture due to its stimulating effect on photo- 2002). The most conspicuous of these responses synthesis, but very high temperature leads to is the light-induced opening of stomata at sun- stomatal closure, possibly due to its stimulating rise and their closing at sunset in response to effect on respiration (see Chapter 5.2). An photosynthetically driven changes in the sub- increase in leaf nitrogen also leads to a wider stomatal CO2 concentration (Raschke, 1975; stomatal aperture because nitrogen also stimu- Roelfsema and Hedrich, 2005). This minimizes lates photosynthesis (see Chapter 5.3), but this water loss at night, when there is insufficient effect levels off at high nitrogen content. light energy to fuel photosynthesis (see Chapter Stomata are hydraulically driven gas valves 4.1) and hence no need for CO2 diffusion into because, as discussed in Chapter 1.2, they open the leaf (see Chapter 4.2). During the course of when the guard cells take up water and swell, the day, stomatal opening (i.e., stomatal con- causing their turgor pressure to rise (Franks g ¼ r 1 ductance, s s ) normally follows the daily and Farquhar, 2007). This clever mechanism is change in light intensity, which peaks around possible because the guard cells lack plasmo- midday (Du¨ ring and Loveys, 1982). Similarly, desmata connecting them to neighboring stomatal aperture is wider in bright light than epidermis and mesophyll cells; that is, they in dim light such as prevails under cloudy con- are symplastically isolated (Wille and Lucas, ditions or in the interior of dense canopies 1984). The osmotic uptake of water is driven (Shimazaki et al., 2007). Conversely, stomata close by a DC (see Chapter 3.1) caused by movement þ when a leaf runs out of water. Therefore, low of potassium ions (K ) from surrounding cells C þ leaf (indicating water stress; see Chapter 7.2) into the guard cells across K channels in 3.2. TRANSPIRATION AND STOMATAL ACTION 91
þ exchange for hydronium ions (H3O ) or pro- 2009). In other words, ABA modifies the per- þ tons (H ), which are driven out by a light- meability of the guard cell membrane by way þ þ activated pump called H -ATPase embedded of a Ca2 signal. in the plasma membrane (Roelfsema and The reaction of stomata to ABA is rapid Hedrich, 2005; Shimazaki et al., 2007). The (within approximately 10 min) and indepen- þ þ C reverse exchange (K out/H in) causes water dent of leaf (although ABA is also produced C to diffuse out of the guard cells so that the in the leaves in response to decreasing leaf) decrease in turgor causes the stomata to close. and vapor pressure deficit, and it does not þ The H may come from the organic acid involve starch degradation (Roelfsema and malate, which is produced from starch stored Hedrich, 2005). However, it is related to an in the guard cell chloroplasts and, courtesy of increase in the pH of the xylem sap (from the its double-negative charge, also serves as the “normal” pH 5–6 to 6.5–7), which enhances þ counterion to K . The starch, in turn, is pro- ABA delivery to the guard cells by reducing duced by guard cell photosynthesis or from uptake of ABA by intervening tissues (Bacon sucrose imported across the cell walls from et al., 1998; Hartung and Slovik, 1991; Wilkinson the underlying mesophyll cells (Lawson, and Davies, 2002). In other words, at higher 2008). In addition, sucrose derived from starch pH, less ABA becomes “trapped” inside the usu- þ breakdown also supplements K as an osmoti- ally more alkaline parenchyma and phloem cells cum, especially in the afternoon (Roelfsema lining the xylem vessels. It appears that a and Hedrich, 2005; Shimazaki et al., 2007; decrease in the flow rate of xylem sap leads to Talbott and Zeiger, 1996). Thus, guard cells an increase in the sap’s pH, perhaps because actively, and rapidly, regulate stomatal opening there is more time for the adjacent cells to and closing via a dynamic balance of starch/ remove protons from the sap (Jia and Davies, þ þ malate and H /K . The hydraulic effect of 2007). Even when only a portion of the roots C leaf on stomatal opening or closing can be experience dry soil, the ABA produced by that amplified by abscisic acid (ABA), which is portion is sufficient to trigger strong stomatal produced from carotenoids in dehydrated cells, closure despite the remainder of the roots C vascular tissues, and guard cells (Nambara and taking up enough water to maintain high leaf Marion-Poll, 2005; Wilkinson and Davies, 2002). (Comstock, 2002; Lovisolo et al., 2002a; In addition to the ABA produced in drying Stoll et al., 2000). When rainfall or irrigation leaves, the amount arriving from the roots can replenishes soil moisture, the roots stop increase considerably in response to reduced producing ABA, whose content in the xylem soil moisture (Bray, 1997; Davies and Zhang, therefore decreases rapidly. Nonetheless, the 1991; Davies et al., 2002). Thus, ABA acts as a transpiration-driven rise in water flow up the messenger from the roots, indicating water xylem after such water supply may “flush” stress when the soil dries out (see Chapter 7). residual ABA to the leaves, where it may keep C As soil declines, the roots produce increasing the stomata partially closed for 1 or 2 days, per- amounts of ABA, which is transported in the haps to enable any embolisms that may have xylem sap to the leaves’ guard cells (Loveys, formed in the xylem to be repaired (Lovisolo 1984). ABA triggers stomatal closure by induc- et al., 2008; see Chapter 3.3). Because the ABA ing an increase in reactive oxygen species that, that is present in the leaves is quickly degraded þ in turn, lead to a brief rise in calcium (Ca2 ) (mainly to phaseic acid; Nambara and in the cytosol, which blocks influx into and Marion-Poll, 2005), the stomata can then reopen. þ enhances efflux from the guard cells of K The turnover of water in a rapidly transpiring (Allen et al., 2001; McAinsh and Pittman, leaf occurs within 10–20 min; that is, the leaf 92 3. WATER RELATIONS AND NUTRIENT UPTAKE loses the equivalent of its entire water content than the leaf temperature. Nevertheless, every 10–20 min (Boyer, 1985; Canny, 1993). even in bright sunlight, leaves can radiate Although transpiration is sometimes described 50–80% of the energy they absorb. as a necessary evil (due to a plant’s dependence Convection: Heating or cooling of ambient air on gas exchange), it has some useful side effects. via sensible heat—that is, heat that can be For instance, the negative pressure (tension) in felt and measured with a thermometer. The the xylem created by transpiration aids in energy of molecules on the leaf surface is extracting water from the soil (see Chapter 3.3) exchanged with that of air molecules with that can be used for plant growth. This, in turn, which they are in direct contact. The attracts nutrients, which can then be taken up by warmed air becomes lighter and rises, which the roots. Although the increased flow of water leads to cooling. Convection is driven by during transpiration leads to a corresponding the temperature difference between the air increase in the rate at which dissolved nutrients and the leaf, and the boundary layer is the move upward in the xylem conduits, trans- main resistance to this process. Wind piration is not necessary for this nutrient move- decreases the thickness of the boundary ment (Tanner and Beevers, 2001). In fact, plants layer, which increases convective heat can absorb and transport nutrients just as well transfer and thus convective cooling. at night when transpiration is minimal as they Transpiration: Release of latent heat during do during the day. More important, evaporation the vaporization of water into the air, also of water is a powerful cooling process, which referred to as transpirational cooling. Water is why this principle is also used in air evaporation consumes energy called the conditioning. This heat loss prevents leaves latent heat of vaporization (2.45 kJ g 1 at exposed to sunlight from overheating. The 20 C) because water vapor contains more temperature inside actively transpiring leaves energy than liquid water. In other words, is ordinarily no more than 2 or 3 C higher than evaporation removes heat that cannot be felt the ambient temperature. Nevertheless, leaf or measured with a thermometer. temperature can rapidly fluctuate by more than Transpiration is driven by the water 10 C in direct sunlight (Sharkey et al., 2008): potential difference between the inside of a Water evaporation cannot completely prevent leaf and the outside. The water potential “heat spikes.” Grapevines have several cooling depends on temperature (which in turn tactics to lose (dissipate) excessive heat gained depends on solar radiation) and relative from absorbed solar radiation. The three humidity. main pathways are as follows (Mullins et al., 1992; Taiz and Zeiger, 2006): Radiation 3.3. WATER AND NUTRIENT : Direct transfer (re-radiation) of heat UPTAKE AND TRANSPORT as long-wavelength (infrared) radiation to surrounding objects or the sky. The percentage of incoming solar radiation Although the main force driving water reflected by a surface is termed albedo. movement through plants is water pressure, Leaves absorb ultraviolet, visible, and water uptake by the roots and transport to the infrared radiation and radiate infrared leaves is largely brought on by passive forces energy, which is a cooling process. This because the hydraulic system of plants has no happens mostly during the night, when the moving parts (Tyree and Zimmermann, 2002). temperature of the surrounding air is cooler The direction of movement is always from high 3.3. WATER AND NUTRIENT UPTAKE AND TRANSPORT 93 to low pressure or, if the movement is across grapevines oscillates with a minimum at night membranes, from a region with low solute con- and a maximum at approximately midday centration to one with high solute concentra- and why vines with a large canopy sustain tion. As discussed in Chapter 2.2, water higher flow rates than do small vines (Tarara absorption and movement in early spring is and Ferguson, 2006). The transpirational pull driven by positive root pressure induced by is possible because water forms a continuous remobilization of stored nutrients and starch system from evaporating surfaces in the leaves, and release of osmotically active ions and shoots, tendrils, and flower and fruit clusters to organic molecules into the xylem sap. Under absorbing surfaces in the roots and out into the these conditions, water moves from the soil soil. This system is termed the soil–plant–air into the root by an osmotically generated gradi- continuum. The chemical and physical proper- ent (difference per unit distance) of water ties of water (i.e., high surface tension, cohesion potential and is then pushed up the vine or the strong tendency of H2O molecules to through the xylem. Although an osmotic gradi- “stick” together by forming hydrogen bonds ent (Dp)—that is, a gradient in solute concentra- between the negatively charged oxygen and tion—has virtually no effect on the flow rate of the two positively charged hydrogen atoms, water inside the xylem because vessels have no and strong adhesion of H2O molecules to the membranes, it drives water inflow into the surfaces of xylem conduits) keep it in this con- xylem from the soil across the root cell mem- tinuum so that water exiting the leaves by tran- branes (Wegner and Zimmermann, 2009). This spiration is continually replaced by water being Dp usually collapses after leaf expansion pulled up the xylem from the roots (Steudle, (although it can remain important at night or 2001; Tyree and Zimmermann, 2002). The tran- during rainfall or overhead irrigation), and spiration–cohesion–tension theory of xylem water loss to the atmosphere by transpiration sap flow states that the evaporation of water from the unfolding leaves begins to hydrauli- from cell walls inside the leaf creates tension, cally produce a negative pressure (tension) in and adhesion of water to the walls of the xylem P the xylem ( x), which maintains the upward helps counteract gravity (i.e., the gain in gravi- flow of water in the vine. At night or under tational potential energy with increasing light conditions low enough to lead to stomatal height). Due to cohesion, the tension extends closure, such as during overcast periods, and down the xylem, into the roots, and out into P with plentiful soil water supply x can remain the soil. Therefore, plants directly utilize solar positive, albeit below atmospheric pressure energy (which causes water to evaporate from P ( a 0.1 MPa at sea level), but drops below the leaves) to drive water uptake and distribu- P < zero (i.e., x 0 MPa) as soon as the stomata tion. Again, the water moves along a gradient open even slightly to let water vapor escape of water potential (from high to low). This and declines as light intensity and transpiration gradient is intermediate within the root rise (Wegner and Zimmermann, 2009). xylem (0.3–2.5 MPa m 1 with lower values in Approximately 95–98% of all water absorbed well-watered and higher values in water- by the roots is lost to the atmosphere in the pro- stressed vines), gentle within a shoot (0.1–0.5 cess of transpiration; very little water is MPa m 1), and quite steep within the petiole required to build the plant body through cell (1–5 MPa m 1) of a transpiring leaf (Lovisolo expansion. Thus, transpiration is the main et al., 2008). driving force for water uptake and movement For xylem sap to sustain the tension required in the xylem up the vine to the leaves. This is in a grapevine to pull water from the soil, the why water flow through the trunk of hydraulic system has to be air-tight. The water’s 94 3. WATER RELATIONS AND NUTRIENT UPTAKE surface tension acts like a seal, keeping water removal), or if it receives some other physical inside the xylem and air out. However, the xylem shock (e.g., a bump by machinery), it snaps is a vulnerable pipe, and water is saturated with back like a piece of elastic in a rigid tube. This air at atmospheric pressure (Steudle, 2001; Tyree phenomenon is termed cavitation, in which the and Zimmermann, 2002). If the water column water column breaks and fills with H2O vapor inside the xylem is put under too much tension or air bubbles (Figure 3.2). The air can be sucked (e.g., excessive transpiration due to drying wind to adjacent vessels through the holes of the or heat combined with dry soil), if the tension is interconnecting perforation plates. Such gas suddenly relieved (e.g., by shoot, leaf, or cluster blockages are called embolisms and render the
End wall of vessel element with bordered pits
Gas-filled cavitated vessel Scalariform perforation plate
Pit
Liquid water
FIGURE 3.2 Cavitation of xylem vessel by water vapor bubble (left; reproduced from Taiz and Zeiger, 2006), bleeding of xylem sap during budbreak (top right); and guttation at night during the growing season (bottom right; photos by M. Keller). 3.3. WATER AND NUTRIENT UPTAKE AND TRANSPORT 95 vessels nonfunctional; in severe cases, this can Trifilo` et al., 2003). Moreover, the vessels and lead to leaf wilting and even canopy collapse tracheids in leaves and petioles are narrower (Schultz and Matthews, 1988a; Tyree and and much shorter than those in the rest of the Zimmermann, 2002). For the gases to dissolve plant, and the pits in the adjoining end walls P and conduits to refill with liquid water, x must of these tracheary elements normally do not rise to within approximately 0.1 MPa of atmo- allow air to pass into the next vessel—that is, P spheric pressure or above (i.e., x 0 MPa) pits function as hydraulic safety valves (Choat (De Boer and Volkov, 2003; Sperry et al., 2002, et al., 2008). This prevents spreading of embo- 2003). More accurately, gas emboli dissolve at lisms in the xylem network following leaf P > tr 1 t x 2 (where is the surface tension of damage and sacrifice, a protective mechanism water, and r is the radius of curvature of the dubbed the “lizard’s tail” strategy (Tyree and air–water interface). If cavitation occurs in the Zimmermann, 2002). Finally, the high number trunk or shoots, repair normally requires of parallel tracheary elements makes it unlikely positive root pressure or starch remobilization that the entire transport pathway is interrupted in the surrounding xylem parenchyma cells at any one time. The fact that roots, trunks, and loading of sugar into the vessels to decrease and shoots are not simply hollow pipes ensures the osmotic potential (Cp) of the xylem sap. that some residual water flow persists under These metabolic processes require an input of all but the most severe drought conditions energy to proceed. In addition, the parenchyma (see Chapter 7.2), and this flow can rapidly cells may also release water into the vessels by rehydrate the leaves upon restoration of soil opening the aquaporins in their cell membranes moisture (Lovisolo et al., 2008). (Lovisolo and Schubert, 2006). Vitis species are The root xylem of grapevines is much more thought to be able to generate root pressure vulnerable to cavitation than their shoot xylem >0.1 MPa, which is sufficient to push water to (Lovisolo and Schubert, 2006; Lovisolo et al., more than 10 m above ground level (Tibbetts 2008). This can be a problem with shallow root and Ewers, 2000). Nonetheless, embolism repair systems because roots in the surface soil are the C in grapevines may require both a rise in leaf and first to experience dry soil, and their failure can a cessation of xylem flow so that it typically trigger leaf wilting and canopy collapse. If dee- occurs only at night, provided soil moisture is per roots are present, cavitation in the surface high enough, or during rainfall (Holbrook et al., roots will simply shift water uptake down to 2001). The trunk and shoots, at least after wetter soil layers so that shoot damage can be they have produced some lignified secondary avoided (Sperry et al., 2002). Moreover, the xylem, are much less prone to cavitation than expansive mycelium network created by sym- the petioles and minor veins in the leaves biotic mycorrhizal fungi also facilitates water (Choat et al., 2005; Lovisolo et al., 2008; Nardini uptake into the roots (Stahl, 1900) and helps et al C ., 2001). However, leaf recovers very avoid cavitation by osmotic water “lifting,” rapidly upon water supply, which implies that which is possible due to the high carbohydrate only a fraction of the vessels are cavitated at concentration in the fungi supplied from any one time. the plants. Plants, it seems, can also reciprocate Although the minor veins can suffer from the “favor” at night by transferring water from embolisms even during an average sunny day, the roots to the symbionts (Querejeta et al., the high concentration of inorganic ions and 2003). In a “compensation game” fungal coloni- organic molecules in the leaves can generate zation of the roots tends to increase as soil the positive pressure needed to refill the veins moisture, and hence root growth, decreases relatively easily (De Boer and Volkov, 2003; and vice versa (Schreiner et al., 2007). The fungi 96 3. WATER RELATIONS AND NUTRIENT UPTAKE
DC ¼ C C r can even transfer water, as well as sugar and where soil leaf, h is the hydraulic l nutrients, between neighboring plants (includ- resistance, and h is the hydraulic conductance. ing grapevine–weed transfer) either directly Inside the xylem conduits (vessels and trac- (through the mycelium connecting the roots of heids), the osmotic component of C is insigni- two plants) or indirectly (through uptake of ficant because there are no membranes in the water that was released by the roots of another pathway; hence, water flow follows a pressure et al DC DP plant) (Egerton-Warburton ., 2007). difference over distance (i.e., x). Some- Grapevines growing under conditions of times a negative sign is added on the right side moderate water deficit develop narrower xylem of the equation to indicate that water flows in vessels than vines growing with abundant the opposite direction to that of DC—that is, water supply (Lovisolo and Schubert, 1998; from high to low C (Tyree and Zimmermann, Mapfumo et al., 1994b). Narrower vessels might 2002). Although this equation is surely an over- be a consequence of curtailed supply of assi- simplification, it is quite convenient to explain milates to the cambium resulting from lower most phenomena involved in water flow r photosynthesis due to higher s. Although the through plants. The hydraulic conductance is r smaller vessel diameter increases h, which a measure of transport capacity or water may lead to reduced water loss and increased permeability. At the whole-plant level, it deter- C cavitation resistance, this effect cannot be very mines leaf at any specific transpiration rate. large because the few large vessels typically The term “conductance” should not be confused L K present in grapevines account for the majority with “conductivity” ( h or ), which refers to r DP of the s (Tibbetts and Ewers, 2000). The vulnera- the water flow normalized to over a specific x l ¼ L x 1 bility to cavitation also varies among species length ( ) of the flow path (thus, h h ). L and cultivars (Schultz, 2003); less vulnerable In addition, “specific conductivity” ( s) applies cultivars (e.g., Syrah) probably have denser and to the cross-sectional area (A) of a unit flow L ¼ L A 1 stronger wood and less permeable pits (i.e., path ( s h ) and is roughly proportional more resistant xylem) than more vulnerable to the number of xylem conduits passing cultivars (e.g., Grenache, which nevertheless through that cross section and the fourth thrives in hot, arid, and windy conditions). power of their radius (r4; the rate of water The flow of water from the soil toward the flow per unit cross-sectional area is termed roots is maintained by suction at the root flux, J ¼ FA1) (Tyree and Zimmermann, L L surface, which is caused by root pressure (i.e., 2002). Therefore, h and s are independent of l osmotically) or transpiration (i.e., hydrostati- the total length of the pathway, whereas h r cally). As transpiration removes water, the soil decreases (and h increases) with increasing adjacent to the roots (rhizosphere) becomes path length (e.g., increasing plant height). The r drier than the more distant (bulk) soil. Thus, term h reflects the resistance to water flow water flow to the roots, and from the roots to due to friction between water and conduit the leaves, is caused by a water potential walls and between the H2O molecules. It is a gradient. In analogy to Ohm’s law (and to the measure of how much a component of the flow transpiration equation discussed previously), path hinders the free flow of water and thus the flow rate (F; volume per unit time) of water slows water flow. There are both short-distance from the soil to the leaves can be described by radial and long-distance axial or longitudinal r the following equation: components of h; the axial resistance to water flow is orders of magnitude lower than the F ¼ DC r 1 ¼ DC l ðbecause l ¼ r 1Þ r h h h h radial resistance. The radial h is the resistance 3.3. WATER AND NUTRIENT UPTAKE AND TRANSPORT 97 water encounters when it flows across the root given vine but varies for several reasons. In r into the xylem and out of the xylem into the long term, h usually increases with increas- r leaf and fruit cells, much of it imposed by cell ing plant size and age (as does s), but it membranes and regulated by aquaporins decreases with increasing shoot number on oth- (Maurel et al., 2008; Vandeleur et al., 2009). erwise similar vines. In the short term, a rise in r r Most of the axial h is imposed by the vine’s temperature decreases h (by approximately “plumbing layout,” termed hydraulic architec- 2–2.5%/ C due the decrease in water viscosity), ture, which is mainly determined by the num- which enables more rapid water delivery for ber, shape, size, and arrangement of xylem transpirational cooling (see Chapter 3.2) during conduits and their interconnections (e.g., pits), heat periods. Moreover, transport of nutrient þ as well as the total length of the flow pathway ions, especially of cations such as K , may r and the number and shape of bends in the decrease h by increasing pit permeability, pathway (i.e., plant size and shape) (Tyree possibly due to ion-induced shrinking of the and Zimmermann, 2002). Due to their small gel-like pectin matrix (“hydrogel”) in the so- size and fine pore meshwork, however, the pits called pit membranes, which are specialized connecting individual xylem conduits usually modifications of the primary cell wall, rather r account for more than 50% of a plant’s total h than proper cell membranes, with pores et al r < m et al (Choat ., 2008). The increase in h with 0.2 m (van Ieperen ., 2000; Zwieniecki et al increasing plant height partly accounts for the ., 2001). Nitrate (NO3 ), but not other tendency of photosynthesis to decrease as anions such as phosphate (H2PO4 ) or sulfate 2 r plants grow taller because their leaves must (SO4 ), also decreases h, perhaps by altering close their stomata earlier (i.e., at less severe the status of aquaporins either directly or via C water deficit) to maintain leaf above a species- a rise in pH brought about by NO3 assimila- and cultivar-specific minimum (Hubbard et al., tion (Gloser et al., 2007; Gorska et al., 2008; see 2001; Ryan and Yoder, 1997). In other words, also Chapter 5.3). This means that increasing the photosynthesis in tall vines, such as those concentration of some nutrients in the xylem climbing in trees, may be at least partly sap increases the sap flow rate (and hence hydraulically limited. Note, however, that water uptake by the roots) and, conversely, that growth is more sensitive to water stress than nutrient deficiency strongly reduces sap flow is photosynthesis, so it may be more correct to even under well-watered conditions. It also say that growth, rather than photosynthesis, is means that vines may be able to actively, et al r hydraulically challenged (Ryan ., 2006; see rapidly, and reversibly modify h by altering also Chapter 5.2). Moreover, the larger vines’ the concentration of ions in the xylem sap, for more extensive root system and their usually instance, by remobilizing/resorbing nutrients greater trunk diameter tend to counter this in the surrounding parenchyma cells. hydraulic limitation trend. Grapevines sustain very rapid water flow up As in electricity, resistances encountered in their trunks that matches the canopy trans- series (e.g., in the roots, trunk, cordon, shoots, piration rate plus the water “bound” by the r ¼ r þ r þ and leaves) are additive (i.e., h h1 h2 growing tissues. The flow rate in field-grown ... þ r hn), whereas for those acting in parallel vines has been estimated to fluctuate daily from (e.g., several shoots on the same cordon) it is low values at night (<0.1 L h 1) to high values the individual conductances that are additive during days with clear sky (1–3 L h 1), but r 1 ¼ r 1 þ r 1 þ ... þ r 1 (i.e., h h1 h2 hn ). Never- it decreases to 10–20% of clear-sky values r theless, whole-plant h is not constant for a during rainy or overcast days (Tarara and 98 3. WATER RELATIONS AND NUTRIENT UPTAKE
r r Ferguson, 2006). Other estimates, however, are which in turn is correlated with s. Thus, h considerably lower at 2 or 3 L day 1 averaged defines how wide the stomatal pores can be over the growing season (Currle et al., 1983). open without desiccating the leaves (Brodribb r Aquaporin activity in the root cell membranes, and Holbrook, 2003; Jones, 1998). A larger h r C and hence radial h, approximately matches will result in a greater decrease in leaf when the diurnal oscillations in flow rate to enable transpiration increases. Water will flow from C appropriate water uptake by and flow across the soil to the vine as long as xylem is lower et al C the roots (Luu and Maurel, 2005; Maurel ., than soil, and it will flow from the xylem to et al C C 2008; Vandeleur ., 2009). The rate of sap the leaf as long as leaf is lower than xylem. C flow also varies with canopy size and water Even in well-watered vines, leaf can fluctuate availability: For instance, maximum values widely during the day (by as much as 1.5 may be approximately 1 or 2 L h 1 in deficit- MPa, depending on cultivar; see Chapter 7.2) irrigated Cabernet Sauvignon (declining to following both opening and closing of stomata <0.5 L h 1 as the soil dries down), but can at dawn and dusk and depending on the evap- reach 3 or 4 L h 1 in well-watered Concord orative demand (i.e., vapor pressure deficit) of et al C (Dragoni ., 2006; Tarara and Ferguson, the air. These changes in leaf (i.e., changes in 2006). A pressure gradient >10 kPa m 1 is C of the cell walls) must be balanced by required to maintain such flow rates in the changes in turgor pressure for the leaf cells to trunk. Rapid transpiration favors hydrostatic maintain volume and solute concentration, (and hence apoplastic) water flow through the which is achieved due to vacuoles and rigid r root and thus reduces h (Steudle, 2000, 2001; but elastic cell walls. Because transpiration- C Steudle and Peterson, 1998). Conversely, when driven water flow is minimal at night, leaf there is little or no transpiration, such as at reaches a maximum (Schultz and Matthews, r night or with dry soil, h increases substantially 1988a). In other words, soil water status deter- C because water flow is driven by osmotic gradi- mines the baseline leaf in the near absence of ents (i.e., across membranes). Of course, cavita- transpiration (Tardieu and Simonneau, 1998). r tion in the xylem also greatly increases h—to As a soil dries, the extraction of water by the infinity in the affected vessel. If cavitation is roots and transport to the shoots become to be avoided, the flow of water from the soil increasingly more difficult for grapevines. To to the leaves must balance the water lost maintain the water potential gradient that through the stomata. By combining the transpi- drives water flow, the C in the shoot must C ration equation with the flow equation, this can decrease. Therefore, the predawn leaf can be C be written as follows: used as a robust indicator of the soil to which ðC C Þðr þ r Þ 1 ¼ðC C Þr 1 the roots are exposed. Transpiration during the leaf air s b soil leaf h C day decreases leaf below the predawn value C As long as transpiration rate and soil-to-leaf so that leaf at a particular time of day is the C flow rate remain constant, leaf will remain result of both soil water status and transpira- C constant. For a given microclimate ( air), which tion driven by evaporative demand (Smart, determines the vapor pressure deficit, and a 1974; Tardieu and Simonneau, 1998). The water C r r given soil water status ( soil), b and s deter- potential equation introduced in Chapter 3.1 mine the transpiration rate, whereas whole- also applies to the soil solution, although, in r C C vine h determines leaf at that transpiration contrast to plant cells, the p of the soil water rate (Tyree and Zimmermann, 2002). In other is generally negligible (approximately 0.01 C C words, the difference between soil and leaf MPa), except in saline soils (see Chapter 7.3). r C is determined by the transpiration rate and h, Consequently, soil is determined mainly by 3.3. WATER AND NUTRIENT UPTAKE AND TRANSPORT 99
C M, which is close to zero in wet soils but can soil–root–shoot pathway. Although nutrients decrease to 3 MPa in dry soils due to the are often concentrated in the biologically active surface tension resulting from expanding air surface soil, water and nutrient availability var- spaces. In drying soil, a point will eventually ies greatly in both space and time. To compli- be reached at which the resistance to water cate matters further, different nutrients are flow is so great that the vine can no longer often not readily available in the same location. DC maintain a sufficient to sustain transpiration For example, NO3 diffuses through the soil þ and becomes drought stressed (see Chapter 7.2). approximately 10 times faster than K and 500 Drought stress can also be induced by low soil times faster than H2PO4 , so percolating water temperature, which increases the resistance to moves and drains (i.e., leaches) nitrate down water uptake and flow inside the roots (i.e., it to the subsoil much more rapidly than other r increases the root’s radial h), ostensibly due to nutrients. As a consequence, superficial roots closure of aquaporins (Maurel et al., 2008). This often take up soil-immobile nutrients (e.g., can be a problem when cold irrigation water is potassium and phosphorus) from the topsoil, applied to rapidly transpiring vines, especially whereas deeper roots tap water and soil-mobile at high atmospheric vapor pressure deficit nutrients (e.g., nitrate) that leach deeper into (see Chapter 3.2). The resulting imbalance the soil profile. These deep roots can also trans- between water uptake by the roots and water port water to the surface roots in a process loss from the leaves can induce cavitation called hydraulic lift or hydraulic redistribution, in the xylem, which may lead to leaf wilting which occurs mostly at night when transpira- and injury (Scheenen et al., 2007). When vines tion is minimal (Bauerle et al., 2008a). Deep are grown in pots, especially in dark pots roots need not even be grapevine roots; some heated by the sun, whose soil temperature can cover crops might also be able to lift water to be several degrees above air temperature, app- the topsoil, where it may be available to the lying cold irrigation water can inhibit growth vines, although this does not apply to covers (Passioura, 2006). dominated by grass species (Celette et al., Water influx is proportional to the surface 2008; Patrick King and Berry, 2005). Hydraulic area of the root system (Steudle, 2001). redistribution keeps the surface roots alive Grapevines often have dense root systems in and, by locally increasing soil moisture due to C < C the topsoil and can extract water effectively water loss from the roots when soil root, from the surface soil layers. Extracting water even enables them to sustain nutrient uptake from lower soil layers is more difficult so that in drying soil. In addition, roots can passively the surface soil dries more quickly than the move water laterally (horizontally) and down- subsoil. Under nonirrigated conditions, roots ward (e.g., when rainfall or irrigation make continue to grow into deeper, wetter soil layers, the surface soil wetter than the subsoil) accord- C et al whereas the roots of irrigated plants proliferate ing to local gradients of soil (Bauerle ., mostly in the topsoil; in both cases, the C of the 2008a; Smart et al., 2005; Stoll et al., 2000). Water advancing root tips remains high as long as redistribution by roots tends to even out local the roots can find water (Hsiao and Xu, 2000). differences in soil moisture (e.g., due to varia- Root water uptake shifts to deeper soil layers tions in soil texture or organic matter), delays as the soil dries and back to the upper layers soil drying, and may assist in maintaining after rainfall or irrigation. symbiotic mycorrhiza. However, water that is Because plant-available nutrient ions are lost from vine roots by diffusion can also be dissolved in the soil solution, nutrient uptake reabsorbed by competitors (e.g., some weeds also depends on water flow through the or cover crops). 100 3. WATER RELATIONS AND NUTRIENT UPTAKE
Epidermis
Casparian band
Endodermis Root hair
Vessel
FIGURE 3.3 Water and nutrient flow from the soil through a root into the xylem via the symplastic (top blue line) and apoplastic (bottom blue line) pathways. Illustration by A. Mills and M. Keller.
As shown in Figure 3.3, water that enters the Steudle, 2001). In other words, cell walls resem- root initially moves through the epidermis and ble dense aqueous gels. Nonetheless, because cortex tissues along both symplastic (or sym- cell walls comprise only approximately 5% of plasmic) and apoplastic (or apoplasmic) routes a tissue’s total volume, they have very limited (Steudle, 2000, 2001; Steudle and Peterson, water storage capacity. In addition, the cell 1998). The symplast is everything inside the walls’ dense network of microfibrils also acts plant that is bound inside a membrane (i.e., like a sieve (pore sizes of 4–14 nm) because it the interconnected cytoplasm of the cells); thus, permits passage only to small molecules such the symplastic pathway is an intracellular path- as nutrient ions, amino acids, and sucrose. way. Within the symplast, water and solutes Water on an apoplastic route flows around the move from cell to cell via small connecting cell membranes; hence, the apoplastic pathway pores called plasmodesmata (Lough and Lucas, is an extracellular pathway. Because there are 2006). Short-distance cell-to-cell movement of no membranes along the apoplast, hydraulic, solutes within the symplast thus occurs by dif- nonselective water flow dominates along this fusion but is also influenced by osmotic gradi- pathway (Steudle, 2000, 2001). Rather than ents across membranes, and long-distance moving exclusively in one or the other com- movement between symplastic domains occurs partment, a third possibility for water flow is by mass flow through the interconnecting straight across cell walls, membranes, and phloem (see Chapter 5.1). The apoplast, on the through cells. In other words, water moves other hand, is everything that is outside the alternately in the apoplast and the symplast. membrane and includes cell walls, intercellular This route is termed the transcellular pathway spaces, and xylem conduits. Cell walls consist and is probably the preferred pathway for of a porous framework of the hydrophilic (i.e., water on its way to the xylem. However, the wettable; from Greek hydros ¼ water, philos ¼ endodermis cells that separate the cortex from loving) polymers cellulose and hemicellulose the stele have thickened, hydrophobic radial (matrix glycans) embedded in a polymer and transverse cell walls called Casparian matrix of pectin, lignin, and proteins, which is strips or Casparian bands, which act like a imbued with water like a sponge so that gasket due to their impregnation with lignin approximately 65–75% of the cell wall is water and suberin (Clarkson, 1993; Steudle, 2000). (Carpita and Gibeaut, 1993; Cosgrove, 1997; This waterproofing blocks diffusion and 3.3. WATER AND NUTRIENT UPTAKE AND TRANSPORT 101 hydraulically separates the cortical and stelar inward rectifying or outward rectifying. apoplasts, forcing water and dissolved nutrient Inward-rectifying channels allow ions to move ions to pass directly through the cell mem- into cells, whereas outward-rectifying channels branes into the cell interior—that is, into the facilitate ion movement out of cells (Tyerman symplast. Thus, on their way to the xylem, et al., 1999). Such one-way movement water and solutes must pass at least two cell allows channels to function like valves, which membranes (i.e., in and out of at least one cell), prevents unnecessary nutrient loss. Because of regardless of whether the initial route is sym- their ability to control the passage of ions, plastic or apoplastic (Tester and Leigh, 2001). membranes also act as electrical isolators, Unlike cell walls, cell membranes are semi- which leads to electrical potential gradients permeable (see Chapter 3.1); that is, they are, across membranes (negative charges typically in principle, impermeable to solutes, including dominate inside the cell) in addition to concen- even small ions such as protons. In reality, they tration gradients. The sum of the gradients of act like selective sieves, permitting entry to concentration and of electric charge of an ion some ions and not to others. The degree of is called the electrochemical potential gradient. selectivity or semipermeability is termed the Some channels, termed ion channels, are reflection coefficient(s), which can vary from specific for cations in general, and some ion 0 to 1. Membranes with s ¼ 0 behave nonselec- channels are specific for particular cations, tively and are equally permeable to water and whereas solute carriers are responsible for the solutes, whereas at s ¼ 1 membranes retain transport across membranes of neutral solutes. solutes completely and have high water but Some neutral solutes (e.g., urea and boric and no solute permeability (Steudle and Peterson, silicic acid) and gases (e.g., CO2 and NH3) 1998). A high s in the endodermis ensures, may also move through aquaporins (Maurel for instance, that nutrients taken up by the et al., 2008). A channel is called specific if it is roots do not leak back out into the soil solution, selective, or discriminates, for a particular mol- which also permits the buildup of root pressure ecule or group of molecules such that other under conditions of slow transpiration. Pores molecules cannot pass through the channel, created by special channel and transport pro- regardless of whether their movement is pas- teins embedded in the membranes regulate sive or active. Ions moving passively simply the passage of water and nutrient ions, and diffuse across channels down the electrochemi- each cell type or tissue seems to be equipped cal potential gradient. As with water flow, the with a unique set of these proteins. Many chan- direction of movement is from high to low nel proteins, especially in the parenchyma tis- potential, typically from high to low concentra- sue around vascular bundles, are specific for tion. Ions moving actively must be transported water (aquaporins; see Chapter 3.1) and control (pumped) across membranes by transport pro- water flow by opening or closing (Maurel et al., teins called transporters or carriers against their 2008; Steudle, 2000). Thus, they control the rate electrochemical potential, ordinarily from low but not the direction of the flow, which simply to high concentration (Grossman and Takaha- and passively follows the DC (Tyerman et al., shi, 2001). Active movement typically draws þ 1999). For instance, aquaporins close at night on a proton (H ) gradient across the membrane but open during the day; this opening and requires an input of energy in the form of r decreases the radial h, which enables the roots ATP, which is hydrolyzed to ADP and phos- to meet the water demand for transpiration. phate (Gilroy and Jones, 2000; Lalonde et al., In contrast with the bidirectional aquapor- 2004). However, rather than burning ATP ins, many ion channels are said to be either directly for nutrient transport, the ATP is used 102 3. WATER RELATIONS AND NUTRIENT UPTAKE
þ to power pumps, termed H -ATPases, which reached after nitrogen fertilizer applications þ export H from the cell interior to the apoplast and are relatively common for potassium; þ þ at a rate of approximately 1 H per “consumed” moreover, the passive uptake of NH4 and þ ATP (Britto and Kronzucker, 2006; Sondergaard Na can sometimes result in ion toxicity (see et al., 2004). The resulting decrease in apoplast Chapter 7.3). Some transporters (e.g., those for þ pH (to pH 4 or 5) generates an electrochemical NO3 and K ) can function in both active and gradient of -50 to 200 mV that, in turn, fuels passive mode, depending on the external con- nutrient uptake via the membrane’s channels centration, and anion channels can also trans- þ and carriers. Thus, H -ATPases act as ener- port organic acids (Britto and Kronzucker, gizers that convert the chemical energy released 2006). Although controlled ion uptake is also by ATP hydrolysis into chemiosmotic energy; important to maintain a neutral electrical the resulting ADP is then recharged by respira- charge balance in the plant, uptake rates of tion. Whereas in the so-called antiport one cations and anions are rarely precisely equal; þ proton is pumped out of the cell for each charge roots restore neutrality by releasing H , HCO3 , of an incoming cation, two protons are needed or OH into the soil immediately surrounding per charge of an incoming anion that then them, termed the rhizosphere (Clarkson and enters together with the backflowing protons in Hanson, 1980). In addition, the high uptake the so-called proton cotransport or symport capacity of the passive channels is usually (Amtmann and Blatt, 2009). accompanied by nutrient efflux from the roots. Active transport enables grapevines to Whereas anions can leak back passively in concentrate nutrient ions inside the roots well exchange for protons, cations must be pumped above their concentration in the surrounding back out actively (Britto and Kronzucker, 2006). soil solution (Keller et al., 1995, 2001b), Therefore, because either uptake or efflux is although some concentration is also possible active, even passive nutrient uptake could be simply due to the electrical potential gradients called active in terms of energy expenditure (Amtmann and Blatt, 2009). Most ions occur for the related active efflux. For most nutrient in the soil water at much lower concentration ions save phosphate, this efflux increases than would be required for plant cellular with rising external concentration and becomes functions, and without a concentration mecha- almost equivalent to ion uptake at high nism nutrient uptake would not be possible. soil nutrient concentration, resulting in 2 Whereas anions (e.g., NO3 ,SO4 ,orCl )must considerable—and apparently futile—nutrient always be taken up actively against their elec- cycling across the membranes (Britto and trochemical potential gradient, some cations Kronzucker, 2006). þ þ þ (e.g., K ,NH4 ,orNa ) are taken up passively Once inside the stele, most water and dis- across relatively generic (i.e., nonselective) ion solved nutrients ultimately return to the apoplast channels when their availability in the soil solu- when the water is released into the xylem tion is high and actively and very selectively conduits across the pit membranes separating across carriers when availability is low (Tester the vessels and tracheids from the surrounding and Leigh, 2001; Ve´ry and Sentenac, 2003). For parenchyma cells (see Chapter 1.3), which most macronutrients, the carriers become themselves have acquired the solutes via their saturated at approximately 1 mM in the soil plasmodesmata from other parenchyma cells or solution, and the channels are switched on from the cell wall apoplast (Sattelmacher, 2001; when the nutrient concentration rises above Sondergaard et al., 2004). Because the cell walls this threshold (Britto and Kronzucker, 2006). of the xylem conduits are waterproof around the Such high nutrient concentrations may be pits, release into the xylem is another bottleneck 3.3. WATER AND NUTRIENT UPTAKE AND TRANSPORT 103
r of high radial h in the vine’s hydraulic system the nonxylem apoplast, but this diffusion is at after the passage through the endodermis. In least one order of magnitude slower than the addition, the presence of aquaporins and selec- movement in the xylem (Kramer et al., 2007). tive channel or transport proteins in the paren- This is because cations interact with, and chyma cell membranes provides a chance to hence are slowed by, the negatively charged car- actively control access to the xylem (De Boer boxyl groups of the cell wall pectins that func- and Volkov, 2003; Tester and Leigh, 2001). Such tion like a cation exchanger and also because controlled release and reversible exchange the pores in the apoplast hinder the passage of between parenchyma cells and xylem conduits solutes and increase their path length (Sattelma- also enables the vine to modify the xylem sap cher, 2001). composition and hence to fine-tune nutrient The long-distance transport of water in the supply with the changing demand by the shoots. xylem, the so-called transpiration stream, Once inside the xylem conduits, the tension occurs by bulk flow, in which large numbers P (negative x) induced by transpiration and the of H2O molecules move together, and is r very low axial h of the tracheary elements driven by a hydrostatic pressure gradient 1 DP facilitate rapid transport of water (up to 2 m h ( x) caused by transpiration, which “sucks” during a sunny day) and its dissolved nutri- water up the vine against gravity and friction. ents to the shoots, where they are delivered The flow rate increases exponentially (actually initially to the apoplastic compartment of to the power of 4) with increasing diameter of transpiring organs. the xylem conduits, and so does the velocity of Even with the concentration mechanism dis- the flow (actually to the power of 2), due to the r cussed previously, xylem sap is very dilute decreasing h with increasing diameter. Thus, (i.e., it has a low p) compared with phloem sap the large vessels formed in spring can trans- (see Table 5.1) and the contents of shoot and port water much more rapidly than the leaf cells. Nevertheless, due to the transpiration narrower ones formed during summer. More- stream, it transports large amounts of nutrient over, older vines with more annual rings and ions by mass flow, although the nutrient con- a larger trunk diameter, and hence more large centration can fluctuate severalfold between vessels, are able to transport much more water day and night due to the diurnal change in to support a larger leaf area than can young transpiration rates. Lower concentrations are vines, and their generally deeper and more associated with higher flow rates, which in turn extensive root system is able to access soil are associated with greater water uptake by the moisture at deeper layers, which is especially roots; despite the lower concentration, how- important later in the season as the soil dries ever, the total nutrient flow to the shoots is down from the top. The roots, trunks, and cor- greater because transpiration-driven water dons also act as important water storage com- influx is strongly coupled with nutrient uptake partments; that is, they behave as hydraulic (Wegner and Zimmermann, 2009). In addition capacitors. These reservoirs, which are also to nutrient ions, the xylem can also transport called hydraulic capacitance, are depleted in- sugars, especially during springtime remobili- the morning, when the stomata open and zation of storage reserves as discussed previ- transpiration rises, and replenished in the ously, as well as other organic molecules such afternoon (Schultz and Matthews, 1988a; as amino acids as nitrogen carriers (see Chapter Steppe and Lemeur, 2004). Depletion and 5.3), organic acids as carriers of metal ions (see replenishment of the internal water reserves Chapter 7.3), and hormones such as abscisic leads to daily cycles of shrinkage and expan- acid and cytokinins. Solutes can also diffuse in sion of the trunk diameter. Older vines can 104 3. WATER RELATIONS AND NUTRIENT UPTAKE store more water due to their higher number nonxylem) resistance predominates. This of xylem (and phloem) parenchyma cells; this means that the distance between the veins and buffering capacity could make them less vul- the evaporative surfaces is directly correlated r nerable to xylem cavitation so that they can with, and controls, leaf h, which in turn limits cope with drought stress better than young photosynthesis (Brodribb et al., 2007). Uptake vines. Moreover, the xylem parenchyma cells of water and solutes into the leaf mesophyll mayalsobeabletoabsorbnutrientionssuch cells from the xylem vessels occurs initially as nitrate and potassium during periods of through pits (from the apoplast back into the abundant supply and release them back into symplast of neighboring parenchyma cells) the xylem in times of starvation. In other and then through plasmodesmata or, across words, these cells may serve as short-term the cell membranes, via aquaporins (for water) storage reserves of both water and nutrients. and energy-dependent carrier proteins (for The xylem constitutes more than 99% of the solutes) that are stimulated by the growth total length of the transpiration stream; this hormone auxin. Therefore, the mesophyll is proportion increases with increasing plant size somewhat hydraulically isolated from the (Sperry et al., 2003). Nevertheless, a few dozen remainder of the transpiration stream. micrometers in the roots and in the leaves exert During the day, grapevines lose most of the r the vast majority of whole-vine h, even though absorbed water in transpiration, but a small the entire transpiration stream can be several fraction ( 1%) is used for cell expansion, cell r meters long, and h increases with increasing metabolism, and phloem transport. At night, transport distance (i.e., with increasing plant when the stomata are essentially closed and height) (Sperry et al., 2002). Thus, one of two transpiration is drastically reduced, those small major resistances to water flow through a amounts may become the dominant component grapevine is not in the trunk or shoots but in of water flow. In fact, water is circulated the roots, where the radial (i.e., nonxylem) between the phloem and xylem, and this circu- rather than axial (i.e., xylem lumen) resistances lation maintains water flow in the xylem in the dominate (Steudle and Peterson, 1998); within absence of any significant transpiration, even the xylem, the pit membranes between adjoin- at night (Ko¨ckenberger et al., 1997; Windt ing vessels and tracheids can account for half et al., 2006). The water taken up to balance this r et al or more of the total h (Choat ., 2008). The “growth water” and phloem counterflow other main resistance is in the leaves, at the ter- amounts to a transpiration-independent water minal section of the transpiration stream, flow that is sufficient to transport nutrients where the water flows through orders of veins absorbed by the roots; hence, nutrient uptake in series and in parallel, leaves the xylem net- and long-distance transport in the xylem are work at the veins across and around the bundle not dependent on transpiration (Tanner and sheath parenchyma encircling the veins, and Beevers, 2001). Of course, the increase in moves into and around the mesophyll cells water flow due to transpiration results in a before it evaporates into the air spaces, chang- corresponding increase in the speed at which ing from the liquid to the vapor phase, and dif- dissolved molecules move up the xylem (Peuke fuses out of the stomata (Comstock, 2002; Sack et al., 2001; Wegner and Zimmermann, 2009). and Holbrook, 2006; Sack et al., 2003; see also Rapid transpiration thus increases nutrient r Chapter 1.3). In a transpiring grapevine, h of uptake by the roots, especially when nutrient the leaves accounts for approximately one- availability in the soil is high (Alleweldt et al., fourth of the whole-plant resistance to water 1984a; Hsiao, 1973; Keller et al., 1995). Over flow, and within the leaf the post-vein (i.e., the course of a day, however, root water uptake 3.3. WATER AND NUTRIENT UPTAKE AND TRANSPORT 105 must exceed transpiration sufficiently to also DC is required to sustain growth (Schultz and supply water for growth because plant growth Matthews, 1993). If water ceases to move out is mainly caused by cell expansion, which of the xylem, shoot and leaf growth stops results almost entirely from an increase in the immediately because cells must take up water cell’s water content (Boyer, 1985; Hsiao and to expand, whereas root growth is less affected Xu, 2000; Schopfer, 2006). Water import into (Boyer and Silk, 2004; Hsiao and Xu, 2000; Wu C cells is driven by cell wall loosening and and Cosgrove, 2000). Lower xylem is also accumulation of solutes inside the cell and thus responsible for the smaller berry size of water- depends on a water potential gradient between stressed vines. Moreover, it explains why pre- DC ¼ C the cell and the supplying xylem ( xylem veraison water deficit has a larger effect on C cell; see Chapter 3.1). Therefore, expanding berry size than post-veraison water deficit: C tissues normally have lower cell than mature Water influx into the berry changes at veraison (nongrowing) tissues. Because growth com- from predominantly via the xylem to pre- petes with transpiration for xylem water, cells dominantly via the phloem (see Chapters 1.3 C must also maintain lower cell during the day and 6.2). Photosynthesis of mature leaves is less than at night if they are to sustain expansion affected by decreasing C than is expansion of during the day (Boyer and Silk, 2004). Indeed, young leaves (Hsiao and Xu, 2000), and solute grape berries, which must remove water from transport in the phloem can continue at the transpiration stream surging up the shoot decreasing C. As a consequence, the import of and past the fruit clusters, grow mostly at solutes may exceed their use so that solutes night, especially before veraison and when the often accumulate in sink organs, such as roots water demand by the transpiring canopy is or berries, of vines that experience mild water high (Greenspan et al., 1994, 1996; Matthews deficit. Therefore, berries grown on mildly and Shackel, 2005). water-stressed vines are not only smaller but Growth rates of all tissues change rapidly also have higher sugar content. Yet the C with fluctuating xylem (i.e., with extracellular decrease in photosynthesis under more severe water status), and growth is extremely sensitive water stress leads to a reduction in berry sugar to water deficit (Boyer, 1985; Boyer and Silk, (see Chapter 7.2). As in the berries, some of the 2004; Hsiao and Xu, 2000). Shoot growth of water for root growth comes from the phloem C grapevines declines with decreasing xylem in addition to direct influx from the soil (Boyer C and stops completely when xylem around mid- and Silk, 2004). Moreover, because xylem cells day reaches approximately 1.0 to 1.1 MPa. are differentiated behind the growing root tip, C Because water stress decreases xylem, a major water uptake for transpiration also occurs cause for growth inhibition under water deficit behind the expanding cells. Due to this hydrau- may simply be the smaller DC, which reduces lic isolation, the root tip usually experiences water uptake by the expanding cells (Nonami slightly higher soil moisture than the mature et al., 1997). Moreover, due to the small, imma- root. All of this makes root growth somewhat C ture xylem conduits near the shoot tip and in less vulnerable to fluctuations in xylem and r young petioles and expanding leaves, the h in ensures that it is favored over shoot growth the “growth zone” is much greater than in the when the soil dries (Boyer and Silk, 2004; Hsiao mature portion of the shoot so that a steeper and Xu, 2000). CHAPTER 4
Photosynthesis and Respiration
OUTLINE
4.1. Light Absorption and Energy 4.4. Respiration 118 Capture 107 4.5. From Cells to Plants 121 4.2. Carbon Uptake and Assimilation 111 4.3. Photorespiration 117
4.1. LIGHT ABSORPTION AND enables us to describe light energy as a stream of ENERGY CAPTURE energy-carrying particles called quanta (singular: quantum) or photons. The window of the spectrum called visible Photosynthesis is the most important way of light, because our eyes are sensitive to it, also obtaining energy for all plants including grape- happens to be the range of wavelengths vines. As the name suggests, photosynthesis (approximately 400–700 nm) that is important (Greek photos ¼ light, synthesis ¼ building a for photosynthesis. This region is only a small whole) is a process by which sunlight is con- portion of the entire spectrum of electromag- verted to chemical energy that is used to pro- netic radiation, and other organisms can see duce (synthesize) organic compounds within light of other wavelengths. Insects and birds, the plant from inorganic compounds acquired for instance, often distinguish between objects from outside the plant. Sunlight is a form of such as flowers or eggs according to their color electromagnetic radiation that has properties of patterns in the ultraviolet (UV) range, and both waves and particles. The wave nature of plants can detect both UV and infrared radia- light can be visualized with a prism separating tion in addition to visible light (see Chapter a ray of visible or white light into a continuous 5.2). All forms of electromagnetic radiation array of different colors according to their travel at the speed of light (c 300,000 km s 1), wavelengths (Figure 4.1). Light’s particle nature but the energy content (E) of photons is higher
The Science of Grapevines 107 Copyright # 2010 Markus Keller. Published by Elsevier Inc. All rights reserved. 108 4. PHOTOSYNTHESIS AND RESPIRATION
the conversion of CO2 (which is discussed in Chapter 4.2) occurs in the fluid chloroplast matrix referred to as stroma. This chapter pro- vides a brief summary of the most important aspects of the photosynthetic processes. Because of their importance for life on Earth, the majority of these processes are virtually identical in all plants and are covered in great FIGURE 4.1 A portion of the electromagnetic spectrum. detail in general texts on plant physiology, Reproduced from Taiz and Zeiger (2006). from which most of the following information is derived (e.g., Atwell et al., 1999; Salisbury the shorter their wavelength (l), or the higher and Ross, 1992; Taiz and Zeiger, 2006). their frequency (u), within the spectrum. This Before the light energy can be used, the light can be written as follows: must first be absorbed and translated into a flow of electrons that are derived from the E ¼ hu ¼ hcl 1 splitting of water. The green chlorophyll is where h is the Planck constant (6.63 10 34 Js). the main light-absorbing pigment that makes The energy contained in the photons is light energy capture possible. It is the world’s exploited by plants during photosynthesis, most abundant pigment. Chlorophyll is green which proceeds in two stages. The first is the because it absorbs predominantly blue (430 nm) photochemical capture of solar energy, trans- and red (680 nm) light and reflects most of ported in the form of photons, and its tempo- the green light with wavelengths intermediate rary storage in high-energy chemical bonds of between blue and red (Figure 4.2). In plants, 0 adenosine 5 -triphosphate (ATP; the universal chlorophyll comes in two slightly different biological energy currency) and nicotinamide versions called chlorophyll a and chlorophyll adenine dinucleotide phosphate (NADPH). b, both of which, along with various other The second stage uses this energy to enzymati- pigments called accessory pigments (e.g., caro- cally convert carbon dioxide (CO2) and water tenoids such as carotenes and xanthophylls), (H2O) to sugar (carbohydrate), which in turn are located in the thylakoid membranes. When is used to produce all other organic compounds a chlorophyll molecule absorbs a photon, it is throughout the plant. Of course, in the absence converted from its low-energy state, called of light there is no energy source for photosyn- ground state, to an excited state by abruptly thesis, and there is no carbohydrate produc- shifting one of its electrons from a lower tion. The two stages of photosynthesis occur energy atomic orbital, or shell, close to the in different regions of specialized organelles atomic nucleus to a more distant, higher termed chloroplasts that contain the photosyn- energy orbital. This instantaneous electron thetic machinery. Although the cells of all “jump” from one energy level to another is green plant organs and tissues have chloro- termed a quantum leap. The excited chloro- plasts, the leaves are their main residence, and phyll is extremely unstable and returns to its they are well designed to intercept the maxi- ground state within nanoseconds (10 9 s) by mum amount of light (see Chapter 1.3). The releasing its available energy as either heat or conversion of light energy into chemical energy fluorescent radiation (in which the photon is happens in the membranes of small, disc- transferred to xanthophylls and reemitted), or shaped, stacked sacs called thylakoids (Greek by transferring either the energy or an electron thylakos ¼ sac) inside the chloroplast, whereas to another pigment molecule. 4.1. LIGHT ABSORPTION AND ENERGY CAPTURE 109
FIGURE 4.2 Structure of the chlorophyll a molecule (left) and absorption spectra for different light-absorbing pigments (right).