Effects of water stress on vessel size and xylem hydraulic conductivity in Vitis vinifera L.

Transcription

1 Journa of Experimenta Botany, Vo. 49, No. 321, pp , Apri 1998 Effects of water stress on vesse size and xyem ydrauic conductivity in Vitis vinifera L. Caudio Lovisoo1,3 and Andrea Scubert2 1 Dipartimento di Coture Arboree de Università di Torino, Via Leonardo da Vinci, 44, I Grugiasco, Itay 2 Centro Migioramento genetico e Bioogia dea Vite, CNR, Torino, Itay Received 13 October 1997; Accepted 13 November 1997 Abstract impications. On te one and te aim of investigations focused on crop water management is to improve water Modifications of vesse size and of soot ydrauic use efficiency (Jones, 1990) wist on te oter and, conductivity induced by different water avaiabiity factors affecting water transport parameters are important eves (y 0.35 MPa, 0.6 MPa and 0.8 MPa, determinants of drougt toerance and reative abitat respectivey) were investigated in container-grown preferences of native and cutivated species (Sperry and grapevine pants. Pant water oss, measured as xyem sap fow and as transpiration, was ower in waterand predictors of carbon and water baance in environ- Tyree, 1990; Cocard et a., 1994; Pockman et a., 1995), stressed pants. Morpometric measurements on xyem sowed tat vesses of water-stressed pants menta modes ( Wiiams et a., 1996). ad ower transectiona areas. Soot ydrauic concontroed by te conductance of te components of te Water movement from te roots to te atmospere is ductivity (k ), soot specific conductivity (k ) and s specific conductivity (k ) were ower in water-stressed water patway. Traditionay, stomata conductance and pants. Wen conductivities were measured on soot root conductivity ave been considered te main contro- portions, differences between treatments were par- ing factors of water fow in te pant (Jones, 1983). ticuary ig at te basa internodes. At te ower However, te efficiency of te water transport system (te stress eve no emboism was detected, and reduced xyem vesses in angiosperms) can aso significanty affect conductivity coud be expained by te reduction of water movement by imposing conductivity constraints vesse diameter, according to te Poiseuie equation. (Tyree and Ewers, 1991) and peraps by te reguation At te iger stress eve k was furter reduced by of deivery to te eaves of root cemica signas (Davies formation of vesse emboisms. Te tension gradient and Zang, 1991; Davies et a., 1994; Jackson, 1997). aong te soot increased ony at te iger stress Xyem conductivity is determined by te structure and eve. Stomata conductance was ineary correated size of te vesses (Scutz and Mattews, 1993; Tyree wit k at ow stress eves, suggesting a concerted and Ewers, 1991) and by teir efficiency, wic may be reguation of water fow, wie at iger stress eves affected by te presence of emboisms ( Tyree and stomata conductance decreased wit no canges of Sperry, 1989). k. Reduced deveopment of xyem vesses in grapev- Wen exposed to water stress, pants sow modificaines subjected to moderate water stress may contrib- tions of water fow rate. Modifications of te conductivity ute to te contro of water fow and to a reduction in components of te transpiration patway (root, soot vunerabiity to xyem emboism. and stomata) contribute to determine suc canges. It Key words: Sap fow rate, transpiration, stomata conducttivity by inducing emboism in te xyem vesses (Scutz as been reported tat water stress affects soot conducance, xyem vesse anatomy, vesse emboism. and Mattews, 1988; Tyree and Sperry, 1989; Tognetti Introduction et a., 1996). However, very itte information is avaiabe on te effects of water stress on soot conductivity via Xyem water transport in pants is te subject of intensive modifications of vesse size. Te experimenta approac researc because of its agronomica and ecoogica used ere was to subject grapevine pants to water stress 3 To wom correspondence soud be addressed. Fax: E-mai: scubert@cvt.to.cnr.it Oxford University Press 1998

2 694 Lovisoo and Scubert of different intensity, and to measure te size of xyem vesses, te presence of emboism and canges in soot ydrauic conductivity. Te grapevine was cosen as te experimenta species because of its reativey wide vesses (Saeo et a., 1985). Materias and metods Water movement trougout te pant and stomata conductance Te amount of water fowing trougout te pant was estimated by te measurement of xyem sap fow and by gravimetric measurements. Measurements were taken at te end of te differentia water management period on four consecutive days (tota number of repicates=4). Te stem eat baance (SHB) metod (Sakuratani, 1981; Baker and van Bave, 1987) adapted to woody pants was used to measure xyem sap fow. Gauges camped to te stem, incuding eaters and sensors, were sef-constructed according to te indications of Steinberg et a. (1989). A CR10 (Campbe Scientific Corporation, Logan, Uta, USA) dataogger was programmed to record measurements every 15 s and to average and store tem every 15 min. Data were transferred to a persona computer for cacuations of sap fow. Seat conductance of te gauge was daiy recacuated using minimum predawn vaues obtained after nigt caibration oops. A stem terma conductance vaue of 0.54 W m 1 C 1 was determined according to Sakuratani (1979). Tota pant transpiration was assessed at oury intervas during te day on te same four pants using a scae. Stomata conductance and transpiration were measured on seven eaves per pant (one every tree nodes aong te soot) on te same pants were xyem sap fow was assessed. Leaf gas excange measurements were taken at and using an open-system ADC-LCA3 infrared gas anayser equipped wit a Parkinson Leaf Camber (Anaytica Deveopment Company, Hoddesdon, UK). At and 13.30, temperature and reative umidity were, respectivey, 31.9 C and 35.7 C, and 46.4% and 39.5%. Two minutes before measurement, eaves were oriented to a 90 ange wit incident igt (1200 mmo m 2 s 1) to eiminate variations of stomata conductance due to igt intensity. Growing conditions and irrigation treatments Two experiments were carried out in two subsequent years. In 1995, 20 4-year-od pants of Vitis vinifera cv. Freisa grafted on Vitis riparia berandieri Kober 5BB were grown in a gassouse. Eac pant grew in a 25 container fied wit a substrate composed of a sandy-oam soi/expanded cay/peat mixture (45251, by vo ), fertiized once a mont wit 30 g of a compex ( ) fertiizer. In January, pants were pruned to a singe-bud spur. Average budbreak took pace on 5 Apri. Te singe soot of eac pant was trained verticay by being tied to a stick. Pants were irrigated wit a drip system deivering a programmed amount of water once a day (09.00 ). In te 85 d foowing budbreak, irrigation was managed to maintain te substrate in te containers at container capacity (Casse and Niesen, 1986). Container capacity was previousy assessed aowing water to drip freey from te oes in te bottom of te container and it was daiy cecked in two reference containers (wit one pant eac) not incuded in te experiment. During te experiments, te surface of te pant containers was covered wit gas-impermeabe pastic fim, to avoid water oss due to evaporation. Eigty-five days after budbreak, two groups of 10 omogen- eous repicate pants, grown to 21 internodes and randomy distributed in te gassouse, were subjected to differentia water treatment for 40 d. Tereafter for 35 d unti te end of te growing season, water was again suppied to container capacity to a pants. During te differentia water treatment, pants in te first group ( IRR treatment) were watered to container capacity (y = 0.01 MPa; soi voumetric water soi content=22%; y at midday= 0.35 MPa). Te pants of te second group were subjected to a water stress ( WS1) treatment (y = 0.07 MPa; soi voumetric water content=15%; y at soi midday= 0.60 MPa), by irrigating wit te same frequency, but for af te time. Just before te differentia water treatment y at midday was in average 0.30 MPa witout significant differences among a pants; 1 week after te end of te differentia water treatment y at midday was on average 0.40 MPa and sowed no significant differences between a of te pants. In 1996 te same pants were used, and te growt conditions of 1995 were appied, wit te exceptions described beow. In tis experiment pants were divided in tree groups (7 repicates), eac subjected to an irrigation treatment. Two treatments were as in 1995 (IRR and WS1 treatments). In te additiona treatment ( WS2), irrigation was wited for 20 d (starting 20 d after te beginning of te differentia water management in te WS1 treatment). At te end of tis period y was 0.20 MPa, soi soi voumetric water content was 11% and y at midday was 0.80 MPa. Pant growt In order to assess area an indirect metod was used. In bot years and for eac treatment te widt of a eaves of a pants was measured, and area was cacuated wit reference to a regression equation obtained on 50 eaves per treatment (Scubert et a., 1995). Soot base diameters were measured wit a and vernier caiper, averaging te arger and te smaer diameter. Soot ydrauic conductivity (k ) k was assessed by two different metods. It was measured on cut soot portions at te end of te season in bot years, and it was estimated in vivo (on growing pants) in At te end of te 1995 (15 September) growing season, soots of te four pants used for sap fow and gas excange determinations were cut at teir base and divided in 0.2 m ong portions, eac incuding two nodes. Conductivity measurements were taken immediatey after cutting. A controed pressure system was used according to Scubert et a. (1995). After 2 min at 0.3 MPa m 1 to eiminate emboisms, and after a furter 5 min at 0.1 MPa m 1 to aow stabiization, for eac soot segment two subsequent fow measurements for 2 min were made at a constant pressure gradient (0.1 MPa m 1). Soot ydrauic conductivity was cacuated from pressure gradient and fow measurements. Soot specific conductivity k s and specific conductivity k (Zimmermann, 1983) were cacuated, respectivey, dividing k by te xyem transectiona area at te midde of te measured internode, and by te area dista to te measured internode. Te teoretica soot conductivity was cacuated using Poiseuie s equation, assuming a circuar sape of te vesses, wit te diameter measured as described beow. In order to measure ydrauic conductivity at different internodes aong te soot, in te 1995 experiment soot portions were cut sorter tan te majority of vesses (Sperry

3 Vesse size and ydrauic conductivity 695 et a., 1987). To verify te resuts on unsevered vesses, at te observed at a stereomicroscope ( 100). Soot, xyem and end of te 1996 growing season k measurements were made as vesse areas were cacuated from te average of two ortogona in te previous year, but on onger soot portions. In addition, measurements of soot or vesse diameter, and of xyem in tis experiment te formation of vesse emboisms were tickness, respectivey. Vesse transectiona area was cacuated assessed. On eac of te four pants per treatment, conductivity from a te vesses observed inside tree xyem wedges, 120 was first assessed on two 1.5 m ong soot segments at constant one to te oter, per section. One section was cut and observed pressure gradient (0.1 MPa m 1). Te 1.5 m portions were ten per eac soot portion and per pant. cut in 0.5 m segments. Conductivity was assessed on tese segments at 0.1 MPa m 1, before and after a ig pressure water fus (0.4 MPa m 1 for 10 min) designed to eiminate Resuts soot emboisms (Sperry et a., 1987). Te 1.5 m soot portion engt is iger tan te average vesse engt, and te 0.5 m In 1995, pant growt at te end of te differentia water engt is iger tan te average engt of 82% of te vesses management period was iger in irrigated pants tan in Vitis abrusca and V. riparia (Sperry et a., 1987). In vivo k was cacuated on te same four pants per in water-stressed ones. In 1996, pant growt was again treatment according to te water fux equation, as te ratio iger in irrigated pants tan in bot oter treatments between water fow (F, kgs 1) trougout te soot and te ( Tabe 1). water tension gradient from te base to te apex of te soot As expected, water stress decreased te amount of (dy/dx, MPa m 1) causing te fow. Water fow (F) was water ost by te pant. Xyem sap fow measured trougmeasured by te SHB metod at te end of te differentia water management period, as described above. Te water out 24 by te SHB metod at te end of te differentia tension gradient aong te soot (dy/dx) was assessed on te water management period was about twice as ig in same pants immediatey after te end of te fow measurements. irrigated tan in water-stressed pants, in bot experi- On eac pant tree eaves at te base and tree at te apex of ments ( Tabe 1). Te reative differences in xyem sap te soot were wrapped in te evening wit a doube ayer, fow were simiar trougout te day (data not sown). inside pastic and outside auminium, bag according to Liu et a. (1978). Te water potentia of te bagged eaves was Tota pant transpiration was ineary correated wit sap measured te foowing day wit a pressure camber fow trougout te soot in bot irrigated and stressed (Soimoisture Equipment Corp., Santa Barbara, CA, USA), pants. Te correation coefficients (R2) for measurements and it was assumed to represent te water potentia of te taken between and were, respectivey, corresponding soot xyem. Sap fow and water potentia and 0.951, and te sap fow/transpiration rate measurements were taken between and was Water tension gradient Stomata conductance (g ) was measured in 1995 and s Water tension gradient aong te soot (dy /dx) was it was in te average 35% iger in irrigated pants tan (base-apex) estimated in 1995 at seven node positions aong te soot, in te water-stressed ones ( Tabe 1). based on te ratio between te average transpiration from Te soot, xyem and vesse transectiona areas were eaves dista to te node (cacuated from te unit area measured in A were ower in water-stressed tan transpiration mutipied by te area), measured at te end in irrigated pants in a te soot portions tested of te differentia water management period, and te k at te same node (Tyree and Ewers, 1991). In 1996, water tension ( Tabe 2). Te ratio between xyem and soot transec- gradient was directy measured by pressure camber determinations as described in te previous section. pants at a internodes tested (66% and 67%, respect- tiona areas was simiar in irrigated and water-stressed ivey). Te summation of te transectiona areas of vesse Morpometric measurements umina (wa excuded) was in te average 11.3% of tota Soot and xyem transectiona areas, and te areas of singe xyem transectiona area in bot water-stressed and irrigvesse umina were measured in 1995 at internodes 5, 11 and 17 on te same soot portions used for conductivity measurements. ated pants. Te vesse diameters more frequenty meas- Soot sections about 0.5 mm tick were cut wit a and-ed ured ranged between 60 and 80 mm for bot irrigated and scape midway between te nodes. Sections were directy water-stressed pants; owever, diameters arger tan Tabe 1. Growt parameters (tota area, soot engt, basa soot diameter), daiy water oss, and stomata conductance of grapevine pants subjected to different irrigation treatments in two foowing years (averages ±standard errors) A measurements were taken at te end of te differentia water treatment (125 d after budbreak). IRR, irrigated to container capacity; WS1 and WS2, water stress treatments. Year of experiment Water management treatment IRR WS1 IRR WS1 WS2 Tota area (m2) 0.489± ± ± ± ±0.03 Soot engt (m) 3.45± ± ± ± ±0.28 Basa soot diameter (mm) 7.56± ± ± ± ±0.98 Daiy water oss (kg m 2) 1.02± ± ± ± ±0.07 Stomata conductance (mmo m 2 s 1) 208.6± ±16.8

4 696 Lovisoo and Scubert Tabe 2. Soot and xyem transectiona areas measured in 1995 at internodes 5, 11 and 17 from te soot base on grapevine pants subjected to two irrigation treatments (averages±standard errors); IRR, irrigated; WS1, water-stressed Treatment Internode IRR Soot cross-sectiona 44.5± ± ±1.6 area (mm2) Xyem cross-sectiona 28.6± ± ±1.6 area (mm2) WS1 Soot cross-sectiona 35.5± ± ±1.0 area (mm2) Xyem cross-sectiona 22.9± ± ±0.7 area (mm2) Fig. 1. Reative frequency of vesses of different diameter cass of soot sections at intenode 5, 11 and 17 in grapevine pants subjected to different irrigation treatments. Fied symbos, irrigated to container capacity (IRR); empty symbos, water-stressed ( WS1). Fig. 2. Water conductivity measured in 1995 at different internodes 80 mm were aways more frequenty measured in irrigated aong te soot of grapevine pants subjected to different irrigation treatments (averages±standard errors). Fied symbos, irrigated to pants ( Fig. 1). As a consequence, average vesse transec- container capacity (IRR); empty symbos, water-stressed ( WS1). (a) tiona area was about 35% ower in water-stressed tan Hydrauic conductivity per unit pressure gradient (k ): measured data in irrigated pants at a tested positions aong te soots. (squares) and cacuated teoretica data (trianges), predicted by Poiseuie s equation. (b) Specific conductivity k (=k per unit xyem In te 1995 experiment ydrauic conductivities were s transectiona area). (c) Leaf specific conductivity (k ), cacuated as te measured at different internodes. Te soot ydrauic rate between k and te area dista to te node. conductivity (k ) decreased from basa to apica segments in bot irrigated and water-stressed pants. In irrigated were conductivity was measured ( Zimmermann, 1983), pants, k was iger tan in water-stressed ones; te was aso ower in water-stressed pants tan in irrigated differences were arger at te basa internodes ( Fig. 2a). contros at te basa internodes, wie smaer differences Te cacuated teoretica k predicted by Poiseuie s aw, were observed at internodes ( Fig. 2c). assuming xyem vesses to be capiaries, sowed te same Te average specific conductivity (cacuated from trend, atoug, as expected, absoute vaues were in te vaues measured in te different soot portions) was average 4 or 5 times iger tan measured ( Tyree and correated wit stomata conductance, but te reationsip Zimmermann, 1971; Kramer, 1983). was different in irrigated and water-stressed pants: con- Te specific conductivity k (=k per unit xyem tran- sistent variations of k corresponded to smaer variations s sectiona area) sowed a simiar pattern: it decreased in g in irrigated pants; in water-stressed pants g canges s s from te base to te apex of te soot in irrigated pants, were not matced by k canges (Fig. 3). wie in stressed pants it was neary constant and it was In 1996, k was measured on soot portions of two ower tan in irrigated pants ( Fig. 2b). different engts. It was in te average 23% iger in Leaf specific conductivity (k ), cacuated as te ratio 0.5 m tan in 1.5 m ong soot portions (Tabe 3), and of k and te surface of eaves dista to te internodes te measurements taken on te 1.5 m soot portions were

5 Vesse size and ydrauic conductivity 697 Fig. 4. Water tension gradient aong te soot (dy/dx) of grapevine pants subjected to different irrigation treatments (averages±standard errors) cacuated in 1995 as te transpiration rate/k ratio at different Fig. 3. Reationsip between stomata conductance (g ) and specific s internodes in irrigated (IRR, fied symbos) and water-stressed (WS1, conductivity (k ), measured in Fied symbos, irigated pants empty symbos) grapevines. (IRR); empty symbos, water-stressed pants ( WS1). Squares are measurements taken at ; circes are measurements taken at Eac symbo is te average±standard errors of measurements wen no differences were found between te IRR and te taken on seven different nodes in a singe pant. WS1 treatments (corresponding to te treatments of 1995) tat were MPa m 1 and MPa m 1, respectivey. significanty correated wit te average of te tree However, te water tension gradient significanty measurements taken on te tree 0.5 soot portions increased in te WS2 treatment (dy /dx on woe (base-apex) obtained from te 1.5 portion (R2=0.71). Te differences pant=0.056 MPa m 1). between measurements taken on te soot portions of different engt can be expained by te reduced presence in sorter soot portions of transverse ce was (Scutz Discussion and Mattews, 1988), and te correation between te Higer pants respond to water stress wit a variety of two measurements sows tat te subsequent measurements, pysioogica and moecuar mecanisms. A common taken on 0.5 m portions, are overestimated but aspect of suc responses is a decrease of te conductivity representative of te true vaues. to water fow aong te soi pant atmospere water In 1996, conductivity measurements were taken before patway (te so caed buk conductivity; Moreset et a., and after eimination of emboisms in water-stressed 1990), wic can reduce water oss to te atmospere. pants. Conductivity was not affected by removing xyem Stomata resistance is a key component of buk resistance emboisms in te IRR and in te WS1 treatments, wie to water fow (Meinzer et a., 1996); owever an increase it increased about 24% after removing emboisms in te of te resistance to water fow of oter segments of te WS2 treatment ( Tabe 3). water patway, suc as te roots and te soot, can be In 1996, te in vivo k was cacuated, based on water part of te pant response to water stress. fux and water tension gradient measured at te end of Soot ydrauic conductivity decreases in response to te differentia water management period. Te in vivo k water stress. Hydrauic conductivity can cange as a was ower in bot water stress treatments tan in te resut of (i) interruption of te water coumn in te irrigated treatment, and it was ower in te WS2 treatment vesses (emboism) or (ii) modifications of te size of te tan in te WS1 treatment (Tabe 3). xyem vesses. Severa reports ave sown tat water In 1995, te water tension gradient in te soot stress induces emboism and oss of function of te vesses increased from te base to te apex. Te water tension (Scutz and Mattews, 1988; Sperry and Tyree, 1990; gradient sowed no significant differences at any node Hargrave et a., 1994). Te decrease of xyem conductivity position between te irrigated and water-stressed treatments due to vesse emboism can directy contribute to reduce (Fig. 4). Tese resuts were confirmed in 1996, water fow across te soot (Scutz and Mattews, 1988) Tabe 3. Soot ydrauic conductivity k (kg s 1 MPa m ) measured in 1996 wit different tecniques on grapevine pants subjected to different irrigation treatments (averages±standard errors); IRR, irrigated to container capacity; WS1 and WS2, water-stressed Treatment IRR WS1 WS2 Cacuated in vivo k 38.3± ± ±2.8 Measured k on 1.5 m cut soot portions 30.4± ± ±0.9 Measured k on 0.5 m cut soot portions, before fusing 33.5± ± ±0.4 Measured k on 0.5 m cut soot portions, after fusing 32.8± ± ±1.2

6 698 Lovisoo and Scubert and at te same time it can induce stomata cosure, of simiar magnitude as canopy conductivity, i.e. te wic in turn avoids furter emboisms and imits transpiration conductivity of te water patway from te soot xyem (Sperry, 1986; Meinzer and Grantz, 1990; Sperry to te substomata cambers ( Running, 1980; Moreset and Pockman, 1993). et a., 1990). It is difficut to compare iquid pase (e.g. Te resuts of te 1995 experiment add a new compon- xyem) and gas pase (e.g. stomata ) water conductivities, ent to tis picture, sowing tat water stress induces a but an indication of teir reative importance can be decrease in te average diameter of grapevine vesses and obtained comparing teir canges as externa conditions a decrease of xyem ydrauic conductivity. A negative vary. Suc reationsips can be diverse: in some cases effect of water stress on vesse size was ypotesized by modifications of stomata resistance ave been sown to Zimmermann and Miburn (1982) and is impied in te prevent a decrease in ydrauic conductivity due to xyem observation tat in periods of drougt, wood xyem rings emboism (Sperry, 1986), wie in oter cases, stomata deveop ess tan wen water is avaiabe. However, to conductance was found to vary in accordance to ydrauic our knowedge, no direct evidence of tis effect as been conductivity ( Meinzer and Grantz, 1990). In te present reported unti now. In tis experiment no emboisms were case te canges in water avaiabiity affected bot detected in te xyem. Water-stressed pants ad ower stomata conductance and ydrauic conductivity. As soot growt and tota transectiona xyem area, wic specific conductivity decreased, stomata conductance was can bot affect ydrauic conductivity, owever, ydrauic initiay ineary correated wit specific conductivity, conductivity was aso ower wen expressed per unit a situation simiar to tat described by Meinzer and xyem cross-sectiona area (k ) and per unit area (k ). Grantz (1990). However, an excessive decrease in s Tese resuts suggest tat te decrease in vesse transectiona ydrauic conductivity was avoided by stomata cosure, area due to a diminised growt as a response to as reported by Sperry (1986). Tese resuts suggest tat water stress was te main factor affecting conductivity. vesse size modifications contribute to reguate water fow Tis concusion is furter supported by te observation trougout te pant at reativey ow water stress eves, tat, as predicted by Poiseuie s aw, atoug reative wie at ower y te reguation of water fow is taken differences in vesse transectiona areas (dependent on te over by stomata. square of te vesse radius) between irrigated and waterstressed Atoug te resuts of te 1995 experiment sowed pants were simiar at a node positions tested, tat moderatey water-stressed vines (y 0.6 MPa) reative differences in k (dependent on te fourt power ave smaer vesses, but no detectabe emboisms, Scutz of te vesse radius) were arger in basa internodes, were and Mattews (1988) reported tat severe water stress vesse transectiona areas are arger. Oter factors, owever, may contribute to conductivity canges: vesse dia- same pant. In tat paper, reduction of y from 0.41 (y down to 1.2 MPa) induced xyem emboism in te meter and engt are positivey correated in Angiosperms to 0.77 MPa decreased ydrauic conductivity about ( Zimmermann and Jeje, 1981), and so reduced vesse 96%. However, tese autors did not measure vesse size. engt may ave co-operated wit reduced vesse transectiona In order to assess te incidence of vesse size modification area in decreasing xyem conductivity in water- and of vesse emboism on ydrauic conductivity separ- stressed pants. atey, a new experiment was conducted in 1996 wit vines Conductivity was measured 35 d after te end of te subjected to water stress treatments of different severity water stress period, and te question may arise weter (y 0.6 and 0.8 MPa, respectivey). In tis experi- conductivity was aso affected in growing pants at te ment a y decrease from 0.35 to 0.8 MPa decreased end of te water stress period. To answer tis question, ydrauic conductivity by 72%, in agreement wit te in te 1996 experiment xyem conductivity was cacuated resuts of Scutz and Mattews (1988). In addition, from fow and pressure gradient measurements. Tis it was found tat at te iger stress eve conductivity approac was foowed as te sap fow/canopy transpir- was about 24% ower before eimination of xyem ation rate was cose to 1, sowing tat te pants used in emboi, impying partia emboism of te vesses. tis experiment ad a very ow or no capacitance at te Correspondingy, in tis treatment water tension aong time of te day wen measurements were taken (between te soot increased above te vaues measured in te and ) (Steinberg et a., 1989; Lascano IRR and WS1 treatments. In WS1 te pants adjusted to et a., 1992). Te resuts were cose to conductivity vaues keep te pressure gradient equa to tat of irrigated measured at te end of te growing season, and confirm pants. Tis required ower g to compensate for te ower s tat conductivity was negativey affected by water stress k in te stressed pants. Wen water stress was more at te time stress was imposed on te pants. severe, tis baance was not maintained, presumaby g s Te decrease of xyem ydrauic conductivity induced was not reduced sufficienty to baance te y gradient, by water stress may contribute to reduce te buk conduc- and emboism occurred. It was concuded from tese tivity. However, oter resistances affect fow across te resuts tat grapevine soots respond to water stress by pant. Previous reports sowed tat soot conductivity is two parae mecanisms: modification of xyem vesse

7 Vesse size and ydrauic conductivity 699 size, wic is induced at moderate stress eves, and vesse References emboism, wic requires more severe water stress. Baker JM, van Bave CHM Measurement of mass fow Tese two conductivity responses to water stress require of water in te stems of erbaceous pants. Pant, Ce and different conditions and may pay different roes. It was Environment 10, observed tat a reduction of xyem size after 40 d of Casse DK, Niesen DR Fied capacity. In: Kute A, ed. exposure to a moderate water stress as a growt decrease Metods of soi anaysis. Madison, Wisconsin, USA: American response. Moreover, te reationsip between stomata Society of Agricuture Inc. Cocard H, Ewers FW, Tyree MT Water reations of a conductance and specific conductivity suggests tat tropica vine-ike bamboo (Ripidocadum racemiforum) reduction of xyem size affects soot water fow at rater root pressures, vunerabiity to cavitation and seasona moderate water stress eves ( WS1 treatment, see above). canges in emboism. Journa of Experimenta Botany 45, On te contrary, emboisms formed at iger stress eves in tis experiment, wic is te case in most angiosperms Davies WJ, Tardieu F, Trejo CL How do cemica signas work in pants tat grow in drying soi? Pant Pysioogy ( Tyree and Ewers, 1991). Modification of vesse diameter 104, and soot emboism aso differ in teir capacity to be Davies WJ, Zang J Root signas and te reguation of eiminated once te stress is over: in te first case again growt and deveopment of pants in drying soi. Annua a sow growt process is needed, wie emboisms can Review of Pant Pysioogy and Moecuar Bioogy 42, recover reativey quicky (Sperry, 1986; Sperry et a., Düring H, Loveys BR Diurna canges in water reations and abscissic acid in fied grown Vitis vinifera cvs. I. Leaf 1987; Yang and Tyree, 1992). Tus in a arge-vesse water potentia components and conductance under species suc as te grapevine, reduction of vesse size may umid temperate and semiarid conditions. Vitis 21, be an adaptation to a persistent situation of moderate Hargrave KR, Kob KJ, Ewers FW, Davis SD Conduit water stress, wie emboism may be induced by a sort diameter and drougt-induced emboism in Savia meifera and more severe water stress. Conditions of moderate Greene (Labiatae). New Pytoogist 126, Jackson M Hormones from roots as signas for te water stress ( 0.6 MPa in tis experiment) in te grapev- soots of stressed pants. Trends in Pant Science 2, ine are common even in umid temperate viticutura Jones HG Pants and microcimate. Cambridge areas (Düring and Loveys, 1982). Tus modifications of University Press. vesse size are ikey to be a common mecanism of Jones HG Pysioogica aspects of te contro of water response to water stress in tis pant. status in orticutura crops. HortScience 25, Kramer P Water reations of pants. Orando, Forida: If reduction of vesse size and xyem emboisms are Academic Press. distinct mecanisms, tey are not independent. Te sus- Lascano RJ, Baumardt RL, Lipe WN Measurement of ceptibiity of xyem vesses to emboism is inked to te water fow in young grapevines using te stem eat baance size and structure of teir pit pores, and, in genera, metod. American Journa of Enoogy and Viticuture 43, smaer vesses, probaby as tey ave ess deveoped Liu WT, Wenkert W, Aen LH, Lemon ER Soi pant pores, are ess susceptibe to emboism (Saeo et a., 1985; water reations in a New York vineyard: resistances to water Sperry and Tyree, 1988; Hargrave et a., 1994; Lo Guo movement. Journa of te American Society of Horticutura et a., 1995). In V. vinifera, Saeo et a. (1985) interpreted Science 103, te ower diameter of apica vesses mainy as a protection Lo Guo MA, Saeo S, Piaceri EC, Rosso R Reations against emboism in te apica part of te soot. In between vunerabiity to xyem emboism and xyem conduit dimensions in young trees of Quercus cerris. Pant, Ce and addition, anoter factor wic can increase soot embo- Environment 18, ism, water tension across te soot, was not affected by Meinzer FC, Fownes JH, Harrington RA Growt indices te WS1 treatment, wic induced canges in k due to and stomata contro of transpiration in Acacia koa stands modifications of vesse size. Tus te modification of panted at different densities. Tree Pysioogy 16, vesse diameter at te basa internodes induced by moderconductance in growing sugarcane: stomata adjustment to Meinzer FC, Grantz DA Stomata and ydrauic ate water stress may not ony ave te effect of reducing water transport capacity. Pant, Ce and Environment 13, water oss from te pant, it may aso ep to avoid emboisms, wic woud impair functioning of te woe Moreset S, Coen Y, Green GC, Fucs M Te branc or tree. It is tus a passive mecanism, wic partitioning of ydrauic conductances witin mature orange may ep te pant reducing water oss and preventing trees. Journa of Experimenta Botany 41, Pockman WT, Sperry JS, O Leary JW Sustained and oss of function of te xyem at te cost of a ower water significant negative water pressure in xyem. Nature 378, avaiabiity in te Running SW Fied estimates of root and xyem resistances in Pinus contorta using root excision. Journa of Experimenta Acknowedgements Botany 31, Sakuratani T Apparent terma conductivity of rice stem We tank E Peterunger, A Peressotti and G Dee Vedove for in reation to transpiration stream. Journa of Agricutura ep in te setting up of te experimenta system for sap fow measurements. Meteoroogy 34, Sakuratani T A eat baance metod for measuring

8 700 Lovisoo and Scubert water fux in te stem of intact pants. Journa of Agricutura Tognetti R, Rasci A, Beres C, Fenyvesi A, Ridder HW Meteoroogy 37, Comparison of sap fow, cavitation and water status of Saeo S, Lo Guo MA, Oiveri F Hydrauic parameters Quercus petraea and Quercus cerris trees wit specia reference measured in 1-year-od twigs of some mediterranean species to computer tomograpy. Pant, Ce and Environment wit diffuse-porous wood: canges in ydrauic conductivity 19, and teir possibe functiona significance. Journa of Tyree MT, Ewers FW Te ydrauic arcitecture of trees Experimenta Botany 36, and oter woody pants. New Pytoogist 119, Scubert A, Restagno M, Noveo V, Peterunger E Effects Tyree MT, Sperry JS Vunerabiity of xyem to cavitation of soot orientation on growt, net potosyntesis, and and emboism. Annua Review of Pant Pysioogy and ydrauic conductivity of Vitis vinifera L. cv. Cortese. Moecuar Bioogy 40, American Journa of Enoogy and Viticuture 46, Tyree MT, Zimmermann MH Te teory and practice of Scutz HR, Mattews MA Resistance to water transport measuring transport coefficients and sap fow in te xyem of in soots of Vitis vinifera L. Pant Pysioogy 88, red mape stems (Acer rubrum). Journa of Experimenta Scutz HR, Mattews MA Xyem deveopment and Botany 22, ydrauic conductance in sun and sade soots of grapevine Wiiams M, Rastetter EB, Fernandes DN, Gouden ML, Wofsy (Vitis vinifera L.): evidence tat ow igt uncoupes water SC, Saver GR, Meio JM, Munger JW, Fan S-M, transport capacity from area. Panta 190, Nadeoffer KJ Modeing te soi pant atmospere Sperry JS Reationsip of xyem emboism to xyem continuum in a Quercus Acer stand at Harvard Forest: te pressure potentia, stomata cosure, and soot morpoogy reguation of stomata conductance by igt, nitrogen and in te pam Rapis excesa. Pant Pysioogy 80, soi/pant ydrauic properties. Pant, Ce and Environment Sperry JS, Hobrook NH, Zimmermann MH, Tyree MT , Spring fiing of xyem vesses in wid grapevine. Pant Yang S, Tyree MT A teoretica mode of ydrauic Pysioogy 83, conductivity recovery from emboism wit comparison to Sperry JS, Pockman WT Limitation of transpiration by experimenta data on Acer saccarum. Pant, Ce and ydrauic conductance and xyem cavitation in Betua occi- Environment 15, dentais. Pant, Ce and Environment 16, Zimmermann MH Xyem structure and te ascent of sap. Sperry JS, Tyree MT Mecanism of water stress-induced Berin: Springer Verag. xyem emboism. Pant Pysioogy 88, Zimmermann MH, Jeje A Vesse-engt distribution in Sperry JS, Tyree MT Water-stress-induced xyem embo- stems of some American woody pants. Canadian Journa of ism in tree species of conifers. Pant, Ce and Environment Botany 59, , Zimmermann MH, Miburn JA Transport and storage of Steinberg S, van Bave CHM, McFarand MJ A gauge to water. In: Pirson A, Zimmermann MH, eds. Encycopedia of measure mass fow rate of sap in stems and trunks of woody pant pysioogy, Vo. 12B. New York: Springer Verag. pants. Journa of te American Society of Horticutura Science 114,