great extent III previous investigations on plant spacing of grapevines.

Transcription

1 Plant Spacing Implicatins fr Grafted Grapevine 1. Sil Characteristics, Rt Grwth, Dry Matter Partitining, Dry Matter Cmpsitin and Sil Utilisatin J.J. Hunter ARC-Fruit, Vine and Wine Research Institute, Nietvrbij Centre fr Vine and Wine, Agricultural Research Cuncil, Private Bag X5026, 7599 Stellenbsch, Suth Africa Submitted fr publicatin: March 1998 Accepted fr publicatin: Octber 1998 Key wrds: Vitis vinifera, Pint nir, grapevine, plant spacing, rt grwth, dry matter partitining, dry matter cmpsitin, sil utilisatin Results presented in part at the Furth American Sciety fr Enlgy and Viticulture Internatinal Sympsium n Cl-Climate Viticulture and Enlgy, Rchester, USA, 1996, the Fifth Office Internatinal de la Vigne et du Vin Internatinal Sympsium n Grapevine Physilgy, Jerusalem, Israel, 1997, and the Twenty-First Suth African Sciety fr Enlgy and Viticulture Annual Cngress, Cape Twn, Suth Africa, Six plant spacings (3 x 3 m, 3 xl,s m, 2 x 2 m, 2 x 1 m, 1 x 1 m, 1 x 0,5 m) were evaluated fr their effect n abvegrund and subterranean grwth, dry matter partitining and dry matter cmpsitin f different plant parts f vertically trellised Vitis vini/era L. cv. Pint nir vines grafted nt rtstck 99 Richter. The vines were pruned t six buds/m2 sil surface area and supplementary irrigated just after pea berry size and veraisn stages. Rt distributin was studied by using the prfile wall methd and by excavating the whle plant. Apparently higher bulk densities ccurred in the tp sil layers f clser spacings. In the case f fine «0,5 mm), extensin (0,5-2 mm) and permanent (2-5 mm) rts, clser spaced vines significantly cmpensated in terms f rt number per prfile wall area. Higher rt densities ccurred with clser spacing. Abvegrund and subterranean grwth f the clserspaced vines were reduced; a distinctin between in-rw spacings f 1,5 m and wider and thse 1 m and narrwer ccurred. Abvegrund and subterranean grwth were psitively related. The angle f rt penetratin and size f the rt system increased with wider spacing. The spread f the rt system was apparently nt affected by either inter-rw r in-rw spacing. The majrity f rts were lcated within the allcated in-rw distance. Cane and rt mass and ttal vine dry mass per m2 sil surface indicated ptimum utilisatin f sil vlume fr medium-spaced vines (2 x 2 m, 2 x 1 m). Evidently higher starch cncentratins ccurred in the crdns, trunks and rts f widerspaced vines. Rts cntained the highest starch cncentratins, fllwed by the trunk, crdn, canes and rtstck trunk. Starch cntents f abvegrund and subterranean plant parts were similar. The rati f abvegrund:subterranean starch cntent was, hwever, lwer fr clsely spaced vines. Althugh the selectin f a particular plant spacing n a given lcality is ne f the mst imprtant lng-term decisins in the cultivatin f grapevines and is critical fr sustained prductivity, this aspect f grapevine cultivatin has been extensively studied during recent years by nly a few investigatrs (Shaulis, 1980; Archer & Strauss, 1990; Bandinelli, Cesari & Di Cllalt, 1993; Reynlds, Wardle & Naylr, 1995; Valenti, Tnni & Cisani, 1996). Furthermre, in spite f the prnunced effect f the rt system n plant perfrmance (Van Zyl, 1988; Hunter & Le Rux, 1992; Hunter et al., 1995), knwledge f the effect f plant spacing n rt develpment and distributin is limited. Except fr earlier classical studies and thers quted by Champagnl (1984), later cmprehensive studies dealt with three-year-ld (Archer & Strauss, 1985), six-year-ld (Bandinelli et al., 1993; Kliewer, Benz & Mran, 1996) and nine-year-ld (Mrlat, Remne & Pinet, 1984) vines. Lng-term effects, dry matter partitining, reserve nutrient status f the different plant parts, and the very imprtant cncept f land utilisatin, have all been neglected t a great extent III previus investigatins n plant spacing f grapevines. Prir t planting, vine spacing is dictated by the sil's physical and chemical prperties, which determine water-hlding capacity as well as efficient rt develpment and distributin fr the supply f water, minerals and grwth regulatrs t abvegrund parts (Richards, 1983 and references therein). The yielding capacity f the sil can be greatly increased by e.g. sil management practices such as ridging, liming and irrigatin (Raath & Saayman, 1995) as well as judicius nitrgen fertilisatin, which all impact directly n grwth and fruit cmpsitin (Cnradie, 1991; Christensen et al., 1994). Hwever, the physical cnfinement f the rt system and the change in sil yielding capacity created by different spacing f vines will undubtedly affect distributin as well as the mrphlgical and bichemical cmpsitin f the rt system, eventually affecting the perfrmance f the whle vine. It is quite cnceivable that the cnstraint induced Acknwledgements: Technical assistance by G.W. Fuche. D..l. Ie Raux. e.g. Vlschenk. A..l. Heyns, W..l. Hendricks. L.M. Paulse, E. Rhde and L.F. Adams as well as financial supprt by the Suth African Vine and Wine Industry thrugh Winerech are greatly appreciated. The Plant Prtectin, and Farm and Maintenance Divisins fnietvrbij are thanked fr quantificatin f silbrne pests, and help with canpy management practices and excavatin f rt systems. respectively. 25

2 26 Plant Spacing Effects n Rt Grwth by this wuld affect the accumulatin f strage tissue reserve cmpunds, ntably starch, which has been identified as the main reserve cmpund (Winkler & Williams, 1945). In this study, sil characteristics, abvegrund and subterranean grwth, as well as dry matter partitining and cmpsitin, were quantified in rder t determine the respnse f a mature, supplementary irrigated Vitis vinifera L. cv. Pint nir/99 Richter vineyard t different plant spacings n a mediumptential sil. MATERIALS AND METHODS Vineyard and treatments: A 14-year-ld Vitis vinifera L. cv. Pint nir (clne BKV)/99 Richter (clne RY 1/30/1) vineyard planted t six different spacings was investigated. The sil was classified as a medium-ptential Glenrsa (Sil Classificatin Wrking Grup, 1991) develped frm Malmesbury Shale. Vines were spaced (inter-rw/in-rw) 3,0 x 3,0 m, 3,0 x 1,5 m, 2,0 x 2,0 m, 2,0 x 1,0 m, 1,0 x 1,0 m, and 1,0 x 0,5 m, representing 1 111, 2 222, 2 500, 5 000, 10000, and vines/ha, respectively. They were trained t a five-strand hedge and spur pruned t six buds/m2 sil surface area. Canpy management practices, namely suckering (remval f all shts nt lcated n spurs at 30 cm sht length) and sht psitining (twice during the seasn) were applied, whereas all shts grwing in excess f 30 cm abve the tp wire were tpped just after pea size stage. Over the last five years f the experiment, supplementary irrigatin f 50 mm was applied just after pea size and veraisn stages, respectively. Vines were previusly grwn under dryland cnditins (Archer & Strauss, 1985). Sil cnditins: Befre planting, the sil was deep delved in tw directins t a depth f 1 m using a wing plugh. In autumn a cver crp (rye) was swed between the rws and killed with a herbicide befre bud break. The sil's physical characteristics were determined accrding t standard ARC-Nietvrbij methds in three layers, i.e cm, cm, and cm. Bulk density was determined in 0-30 cm and cm sil layers. Sil-brne pests (phyllxera, margardes, nematdes) were quantified accrding t methds described by De Klerk (1970, 1978) and Lubser (1985). Abvegrund and subterranean grwth, dry matter partitining: Measurements included cane mass, crdn mass, trunk mass and circumference, rtstck trunk mass and circumference, rt mass and rt distributin. The latter was determined by using tw methds: the prfile wall methd f Bhm (1979) as mdified by Hunter & Le Rux (1992), and whle rt system excavatin. The prfile wall methd cnsisted f the digging f a trench f apprximately 1,4 m deep parallel t the vine rw and 30 cm frm the vine trunk. After expsure f the rts, a 20 x 20 cm grid system was set up against the prfile wall. The width f the grid system extended t the centre between tw adjacent vines. Rts were pltted in six sil layers (0-20 cm, cm, cm, cm, cm, cm depth) and in five rt thickness classes «0,5 mm, 0,5-2 mm, 2-5 mm, 5-10 mm, > 10 mm). They were categrised accrding t Richards (1983) as fine «0,5 mm), extensin (0,5-2 mm), permanent ( 2-5 mm) and framewrk ( 5-10 mm and> 10 mm) rts. Cane, crdn, trunk, rtstck trunk, and rt samples were taken frm these vines fr chemical analyses (vid. "dry matter cmpsitin"). Fr the excavatin f the whle rt system, trenches were dug utside the surface area allcated per vine, whereafter the rts were carefully expsed up t a depth f 1,2 m by using pegs and water jets. The intact plant was then phtgraphed. After remval f the whle vine, the riginal spatial distributin f the intact rt system was recreated and the plants again phtgraphed (with wide-angle lens frm a fixed psitin) t determine the size f the plants as well as hrizntal and vertical distributin f the rts. The excavated plants were als used t determine ttal dry mass f canes, crdn, trunk, rtstck trunk and rts (dry matter partitining) as well as trunk and rtstck trunk circumference. Circumference f the rtstck trunk was measured in ne psitin, whereas that f the trunk was measured in three psitins (tp, middle, bttm); means are presented. Dry matter cmpsitin: Starch, glucse and fructse were extracted frm freeze-dried cane, crdn, trunk, rtstck trunk and rt samples as described by Hunter et ai. (1995). Starch was determined enzymatically (Hunter et ai., 1995), whereas glucse and fructse were analysed by HPLC, using the equipment and cnditins described by Hunter, Visser & De Villiers (1991). Experimental design and statistical analyses: The full experimental layut cnsisted f cmpletely randmised treatments and replicates (five) allwing a minimum f 33 and a maximum f 48 vines per replicate. T eliminate side-effects, brder rws were included. The rt study was dne during the winter f 1995 n three f the abve replicates per treatment. One vine per replicate was used fr each f the prfile wall and whle rt system excavatin studies. Student's t-lsd was used t test fr significant differences. RESULTS AND DISCUSSION Sil cnditins: The numbers f sil-brne pests were insignificant and did nt affect grapevine health (data nt shwn). A general increase in clay and silt and a decrease in sand cntent with an increase in depth were fund (Table 1). The sil cmpactin index (measured as bulk density) exceeded the critical value f 1,5 g/cm3, beynd which rt penetratin is believed t decline (Richards, 1983) (Fig. 1). Apparently higher bulk densities in the tp sil layers (0-30 cm, cm) f clser spacings indicate higher sil cmpactin between the rws, prbably because mvement was cnfined t a small surface area between the rws. As these znes are nrmally highly clnised by rts (Fig. 2) (cf. als Hunter et ai., 1995 and references therein), rt develpment and distributin f clser spacings may have been impeded. Rt penetratin is, hwever, als affected by water ptential f the sil as well as size and rigidity f sil pres (sil prsity) (Richards, 1983). As shwn befre, sil water ptential may have prnunced effects n the perfrmance f grapevines planted t different spacings (Archer & Strauss, 1985, 1989, 1990).

3 Plant Spacing Effects n Rt Grwth 27 TABLE 1 Physical characteristics f a medium-ptential Glenrsa sil. Depth (mm) Silt (%) Sand (%) Clay «0,002 mm) Fine Carse Very fine Fine Medium (%) (0,002-0,02 mm) (0,02-0,05 mm) (0,05-0,1 mm) (0,1-0,25 mm) (0,25-0,5 mm) ,9 11,8 11,0 17,5 22,6 12, ,4 12,7 11,8 17,0 21,0 10, ,2 18,3 11,2 15,2 18,2 9,4 Carse (0,5-2,0 mm) 9,1 8,2 6,2 Values represent the means f five replicatins X 3 m ~ 3 X 1.5 m D 2 X 2 m 0 2 X 1 m IZl 1 X 1 m 0 1 X 0.5 m Sil layer (cm) FIGURE 1 Vine-spacing effect n bulk density in different sil layers. ~ Q) 14 ~ c.. 12 JL) ~ E ::J s:::: ~ Sil layer (cm) FIGURE 2 Number f rts in different sil layers, irrespective f treatment

4 28 Plant Spacing Effects n Rt Grwth Abvegrund and subterranean grwth, dry matter partitining: Given the in-rw distance, it is evident that clserspaced vines in the case f fine, extensin and permanent rts significantly cmpensated in terms f rt number per prfile wall area, resulting in higher rt densities (Table 2). Althugh clse relatinships between cane mass and rt density were previusly fund fr varius rtstcks under different cultivatin practices (Swanepel & Suthey, 1989; Hunter et ai., 1995 and references therein), cane mass (Table 3) was negatively related t rt density/m2 prfile wall (Table 2) in this study. Therefre, althugh rt density is recgnised as an imprtant indicatr f grapevine perfrmance (Freeman, 1983; Richards, 1983), this parameter cannt slely be used t predict abvegrund grwth under the restrictive sil vlumes f different planting densities. Nevertheless, despite the lwer ttal number f rts per prfile wall (Table 2), ccurrence f higher densities f particularly fine «0,5 mm) and extensin «0,5-2 mm) rts in the sil vlumes f clser-spaced vines (Table 2) (cf. als Archer & Strauss, 1985; Band~nelli et al., 1993) culd have cntributed t sustained and even increased perfrmance f these vines n a m2 sil surface basis, as is evident frm Table 3 and Fig. 3. The presence f rapidly grwing fine and extensin rts wuld nt nly increase the absrptive capacity and activity f the rt system and allw mre efficient utilisatin f water and nutrients frm the sil, but als enhance the prductin f grwth regulatrs (particularly cytkinin) invlved in the regulatin f sht and fruit develpment (Freeman, 1983; Richards, 1983). Cnsidering dry mass f all abvegrund and subterranean vegetative parts f the vine, the clser-spaced vines cnsistently displayed lwer values (Table 3). Values btained fr in-rw spacings f 1,5 m and wider and fr thse f 1 m and narrwer TABLE 2 Vine-spacing effect n rt number and rt density. Spacing Ttal number Rt density/m2 prfile wall/rt size (mrn) Ttal rt (m) f density/m2 rts/prfile < 0,5 0, >10 prfile wall wall 3x3 506,0 a 120,1 c 12,2 c 5,6 c 1,3 a 1,4 a 140,6 b 3 xl,s 307,3 bc 133,0 bc 15,6 c 7,1 c 2,6 a 1,7 a 160,1 b 2x2 382,7 b 124,5 c 23,8 bc 8,5 c 1,7 a 1,1 a 159,4 b 2xl 337,3 bc 230,3 a 33,3 ab 14,2 ab 1,7 a 1,7 a 281,1 a 1 x 1 301,7 bc 208,9 ab 30,6 b 9,4 bc 1,1 a 1,4 a 251,4 a 1 x 0,5 236,3 c 265,7 a 44,9 a 15,7 a 1,4 a 0,5 a 328,2 a Values in clumns fllwed by the same letter d nt differ significantly (p s 0.05). TABLE 3 Vine-spacing effect n mrphlgical dimensins and dry matter partitining. Spacing Canes/vine Crdn/vine Trunk/vine Rtstck Ttal rt Ttal rt (m) trunk/vine dry dry mass Length Mass Mass (kg)/m2 Length Circum- Mass Circum- Mass Circum- Mass mass/vine (kg)/m2 (m) (kg) sil surface (cm) ference (kg) ference (kg) ference (kg) (kg) sil surface (cm) (cm) (cm) 3x3 27,3 a 0,45 a 0,05 245,3 a 9,5 a 2,37 a 14,4 a 0,88 a 24,5 a 1,48 a 2,08 a 0,23 3 xl,s 24,1 a 0,41 a 0,09 147,3 c 9,0 a 1,36 b 13,9 a 0,73 a 20,0 ab 1,37 a 1,51 b 0,34 2x2 27,4 a 0,52 a 0,13 180,7 b 9,2 a 1,57 b 14,1 a 0,79 a 20,7 a 1,38 a 1,61 b 0,40 2xl 11,3 b 0,19 b 0,10 90,0 d 9,3 a 0,78 c 10,2 b 0,48 b 14,6 bc 0,76 b 0,88 c 0,44 1 x 1 8,8bc 0,16 bc 0,16 83,0 d 8,7 ab 0,59 cd 9,6 b 0,35 bc 11,9 cd 0,50 bc 0,68 cd 0,68 1 x 0,5 4,0 c 0,05 c 0,31 41,0 e 7,3 b 0,23 d 7,4 c 0,23 c 9,0 d 0,32 c 0,40 d 0,80 Values in clumns fllwed by the same letter d nt differ significantly (p S 0,05). S. Afr. J. Enl. Vitie., Vl. 19, N.2, 1998

5 Plant Spacing Effects n Rt Grwth 29 were distinctly different, indicating a dminating effect f in-rw spacing n grwth. Psitive relatinships between abvegrund and subterranean grwth were fund and were apparently nt changed by plant spacing. Althugh maximum and minimum values were fund fr clsely versus widely spaced vines, cane and rt mass (Table 3) and ttal vine dry mass (Fig. 3) n a m2 sil surface basis shwed that ptimum utilisatin f available sil vlume ccurred fr medium-spaced vines (2 x 2 m, 2 x 1 m), i.e. a stabilisatin in perfrmance was bserved at medium spacing. Regardless f plant spacing, rts were predminantly lcated in the 0-80/100 cm sil zne (data nt shwn). This is in agreement with previus findings under different cnditins f cultivatin (Van Zyl, 1988; Swanepel & Suthey, 1989; Hunter & Le Rux, 1992; Hunter et ai., 1995). The frmatin and vertical penetratin f rts in different sil layers are prbably affected by genetic factrs, depth f sil preparatin, irrigatin practices and sil structure. The angle f rt penetratin as well as size f the rt system prgressively increased the wider the plant spacing (Figs. 4a & 4b). This was als fund by Archer & Strauss (1985) and is evidence f the physical cnstraint caused by the rt systems f adjacent vines, particularly under cnditins f narrw spacing. It is, hwever, nticable that neither inter-rw nr in-rw spacing had any bvius directinal bearing n the hrizntal distributin f rts, i.e. rts were apprximately equally distributed in all directins. In the case f narrw- and medium-spaced vines, rts penetrated slightly beynd the in-rw distance per vine. Hwever, rts f 3 x 3 m and 3 x 1,5 m spaced vines penetrated well shrt f and well beynd the allcated in-rw space, respectively. In spite f this, the great majrity f the rts were still lcated within the available in-rw space, which is further evidence f the restrictive effect f in-rw spacing n grwth. Dry matter cmpsitin: Wider-spaced vines apparently cntained higher starch cncentratins in the crdns, trunks (Table 4) and rts (Table 5), which may indicate that these vines were subjected t a lwer degree f plant stress during the grwth seasn. It has been fund that stred starch can be mbilised frm perennial parts f the plant under stress cnditins (Candlfi Vascncels, Candlfi & Kblet, 1994). Irrespective f size, rts cntained the highest starch cncentratins, fllwed in general by the trunk, crdn, canes and rtstck trunk, the latter tw having almst identical cncentratins. Furthermre, cnsidering ttal starch cntent f each abvegrund part versus that f the ttal rt system, it is evident that the rt system is the main starch strer. In line with previus findings (Hunter et ai., 1995), extensin (0,5-2 mm) and permanent rts (2-5 mm) cntained the highest starch cncentratins. It is interesting t nte that the starch cntent f the vines was apprximately equally distributed between abvegrund (Table 4) and subterranean grwth (Table 5). Hwever, the rati f abvegrund: subterranean starch cntent was lwer than ne fr clsely spaced vines (l x 1 m, 1 x 0,5 m) (Table 5), indicating a preference f clsely spaced vines fr abvegrund sucrse exprt instead f transitry starch frmatin, i.e. the carbhydrate demand f abvegrund grwth was seemingly higher. This is als indicated by the apparently higher hexse (glucse, fructse) cntents in abvegrund grwth f clser-spaced treatments. 3 X 3 m ~ 3 X 1.5 m D 2 X 2 m M 2 X 1 m Q 1 X 1 m IS] 1 X 0.5 m ~ -CD r::: ~ tn fa E i:!' CD 0 -@ ::s tn '0 tn C\I -E tn fa E i:!' 0 Spacing FIGURE 3 Vine-spacing effect n dry mass per vine and dry mass per m2 sil surface.

6 30 Plant Spacing Effects n Rt Grwth FIGURE 4a Vine-spacing effect n the angle f rt penetratin and the size f the rt system. Representative examples are shwn. s. Afr. J. Enl. Vitic., Vl. 19, N.2, 1998

7 Plant Spacing Effects n Rt Grwth 31 FIGURE4b Vine-spacing effect n the angle f rt penetratin and the size f the rt system. Representative examples are shwn. s. Afr. J. Enl. Vitic., Vl. 19, N.2, 1998

8 32 Plant Spacing Effects n Rt Grwth TABLE 4 Vine-spacing effect n starch and hexse cntent f abvegrund dry matter. Plant Carbhydrate Spacing (m) part 3x3 3 x 1,5 2x2 2xl Canes Starch mg/g 51,82 b 58,33 ab 52,69 b 74,77 a g/vine 23,58 ab 24,40 a 27,16 a 13,62 bc Glucse mg/g 1,59 ab 1,06 b 3,29 a 3,31 a g/vine 0,83 b 0,40 b 1,70 a 0,66 b Fructse mg/g 3,86 ab 2,96 b 4,83 a 4,19 ab g/vine 1,78 ab 1,23 bc 2,50 a 0,81 bc Crdn Starch mg/g 74,89 a 78,51 a 63,85 a 59,24 a g/vine 178,86 a 106,11 b 100,27 b 47,11 c Glucse mg/g 4,21 a 1,97 b 4,78 a 5,42 a g/vine 9,78 a 2,73 bc 7,41 a 4,26 b Fructse mg/g 2,79 b 2,87 b 5,82 a 6,21 a g/vine 6,45 b 3,95 bc 9,04 a 4,88 bc Trunk Starch mg/g 86,63 a 78,23 ab 84,80 a 76,33 ab g/vine 77,74 a 56,48 ab 65,65 a 36,68 bc Glucse mg/g 4,07 a 5,01 a 4,22 a 5,98 a g/vine 3,77 a 3,66 a 3,42 a 2,87 ab Fructse mg/g 5,50 ab 6,25 ab 4,81 ab 7,11 a g/vine 5,13 a 4,56 a 3,89 ab 3,43 ab Rtstck Starch mg/g 51,47 a 55,76 a 67,40 a 54,62 a Trunk g/vine 76,14 ab 76,34 ab 97,54 a 42,17 bc Glucse mg/g 0,74 b 2,22 ab 1,71 ab 3,64 ab g/vine 1,15 a 3,03 a 2,10 a 2,76 a Fructse mg/g 2,51 b 2,95 ab 2,90 ab 2,82 ab g/vine 3,74 a 4,04 a 3,48 a 2,15 a 1 x 1 1 x 0,5 59,28 ab 61,53 ab 9,83 c 3,22 c 2,74 ab 1,89 ab 0,45 b 0,09b 4,27 ab 4,29 ab 0,68 c 0,23 c 59,30 a 60,01 a 34,64 c 13,98 c 3,34 ab 4,05 ab 2,08 bc 0,92 c 4,87 a 4,38 ab 2,93 cd 0,98 d 73,60 ab 65,55 b 25,58 c 15,80 c 5,06 a 3,78 a 1,77 ab 0,83 b 6,10 ab 4,16 b 2,10 ab 0,99 b 57,06 a 62,63 a 28,40 c 20,04 c 4,62 a 3,09 ab 2,22 a 0,99 a 6,62 a 4,18 ab 3,19 a 1,34 a Values in rws fllwed by the same letter d nt differ significantly (p ::::; 0,05).

9 Plant Spacing Effects n Rt Grwth 33 TABLE 5 Vine-spacing effect n starch and hexse cntent f subterranean dry matter. Rt size Carbhydrate Spacing (m) (mm) 3x3 3 x 1,5 2x2 2 xl 1 x 1 1 x 0,5 < 0,5 Starch : mg/g 146,08 a 151,18 a 170,37 a 165,22 a 147,18 a 134,45 a g/vine 6,59 a 3,37 b 7,30 a 3,03 bc 2,59 bc 0,40 c Glucse : mg/g 5,23 a 5,55 a 0,98 c 2,99 abc 2,64 bc 2,41c g/vine 0,23 a 0,13 b 0,03 c 0,05 c 0,04 c 0,D1 c Fructse : mg/g 6,66 ab 7,64 a 4,32 b 4,02 b 4,67 b 4,34 b g/vine 0,29 a 0,17 b 0,18 b 0,07 c 0,07 c 0,02 c 0,5-2 Starch : mg/g 192,73 a 210,00 a 194,72 a 172,80 a 152,82 a 168,73 a g/vine 43,69 ab 60,00 a 68,24 a 37,02 ab 21,26 b 18,05 b Glucse : mg/g 2,18 ab 4,43 a 1,30 b 2,04 ab 2,10 ab 2,69 ab g/vine 0,48 b 1,31 a 0,43 b 0,45 b 0,33 b 0,29 b Fructse : mg/g 5,47 a 6,23 a 4,46 a 3,26 a 3,80 a 4,41 a g/vine 1,24 a 1,76 a 1,75 a 0,72 a 0,56 a 0,48 a 2-5 Starch : mg/g 186,55 a 202,02 a 177,76 a 174,85 a 172,26 a 172,26 a g/vine 135,80 a 65,11 bc 79,53 b 67,84 bc 44,93 bc 30,69 c Glucse : mg/g 2,57 a 4,77 a 4,46 a 3,81 a 3,04 a 3,88 a g/vine 1,89 a 1,72 a 2,01 a 1,51 a 0,81 a 0,99 a Fructse : mg/g 3,69 a 6,24 a 5,48 a 4,60 a 5,28 a 3,95 a g/vine 2,73 a 2,15 a 2,42 a 1,79 a 1,39 a 0,72 a 5-10 Starch : mg/g 158,66 a 180,76 a 154,74 a 128,67 a 143,12 a 131,22 a g/vine 97,33 a 94,03 a 63,83 ab 22,50 bc 35,90 bc 6,88 c Glucse : mg/g 2,17 b 5,39 a 3,91 ab 3,41 ab 3,54 ab 3,88 ab g/vine 2,20 ab 2,78 a 1,70 ab 0,61 cd 0,88 bcd 0,21 d Fructse : mg/g 3,31 a 7,76 a 4,85 a 4,57 a 5,69 a 5,36 a g/vine 2,24 ab 3,99 a 2,15 ab 0,81 b 1,36 b 0,29 b >10 Starch : mg/g 121,66 a 109,58 a 94,50 a 100,65 a 110,16 a 112,11 a g/vine 62,43 a 35,23 ab 33,68 ab 9,31 b 3,21 b 8,78 b Glucse : mg/g 4,32 a 4,78 a 4,18 a 3,78 a 3,13 a 4,64 a g/vine 2,11 a 1,52 ab 1,46 ab 0,36 b 0,09 b 0,47 ab Fructse :mg/g 5,60 a 6,03 a 5,14 a 5,53 a 5,01 a 5,11 a g/vine 3,01 a 2,09 a 1,87 a 0,59 a 0,15 a 0,49 a Mean starch cncentratin (mg/g) 161,1 170,7 158,4 148,4 145,1 143,8 Ttal subterranean starch (g/vine) 345,8 a 257,7 a 252,6 a 136,6 b 106,8 b 61,9 b Rati abvegrund:subterranean starch 1,03 1,02 1,15 1,02 0,92 0,86 Values in rws fllwed by the same letter d nt differ significantly (p ~ 0,05).

10 34 Plant Spacing Effects n Rt Crwth CONCLUSIONS The study shwed that bth abvegrund and subterranean grwth are greatly affected by physical cnfinement f grapevines by means f plant spacing. Higher densities f particularly fine and extensin rts in sil vlumes f clser-spaced vines prbably played a large rle in the sustained perfrmance f these vines n an area basis. Althugh dry matter partitining between abvegrund and subterranean grwth was nt affected, distinct differences between in-rw spacings f 1,5 m and wider and thse f 1 m and narrwer shw that grwth is largely dminated by in-rw spacing. Althugh it is recgnised that the distributin f rts may vary under different cultivatin cnditins, this aspect has serius implicatins fr ptimal land utilisatin and shuld be brne in mind whenever recmmendatins regarding the spacing f vines are made. Under the cnditins f this investigatin abvegrund and subterranean grwth as well as land utilisatin shwed that medium-spaced vines (2 x 2 m, 2 x 1 m) cnsistently perfrmed ptimally. This can therefre be used as guideline fr prducers. Althugh all rt types are likely t cntribute t sme extent t the ttal uptake f required substances, it wuld appear that the cmpsitin f the rt system has a critical rle in its efficient functining, supply t abvegrund parts f the vine, and eventual realisatin f the full ptential f the vine in terms f prductivity and lngevity, particularly under cnditins where the perceptin prevails that vines are t clsely spaced. As the cmpsitin f the rt system relates t rt activity, the data can be interpreted as supprting suggestins that sht and rt utput (size x activity) are in equilibrium and hmestatically interrelated. Hwever, the questin arises as t the way in which a rt system with primarily bigger-sized rts versus ne with primarily smaller-sized rts will affect grape cmpsitin and eventually wine quality. The differential cntributin f these tw cmpnents shuld be determined. LITERATURE CITED ARCHER, E. & STRAUSS, H.e., The effect f plant density n rt distributin f three-year-ld grafted 99 Richter grapevines. S. Afr. J. Enl. Vitic. 6, ARCHER, E. & STRAUSS, H.e., The effect f plant spacing n the water status f sil and grapevines. S. Afr. J. Enl. Vitic. 10, ARCHER, E. & STRAUSS, H.C., The effect f plant spacing n sme physilgical aspects f Vitis vinifera L. (cv. Pint nir). S. Afr. 1. Enl. Vitic. 11, BANDINELLI, R., CESARI, D. & DI COLLALTO, G., Risultati di ricerche su viti allevate a differenti densita di piantgine in ambiente cllinare tscan. Vignevini 4, BORM, w., Methds f studying rt systems. Eclgical Studies, Vl. 33, Springer Verlag, Berlin. CANDOLFI-VASCONCELOS, M.e., CANDOLFI, M.P. & KOBLET, w., Retranslcatin f carbn reserves frm the wdy strage tissues int the fruit as a respnse t defliatin stress during the ripening perid in Vitis vinifera L. Planta 192, CHAMPAGNOL, E Elements de physilgie de la vigne et de viticulture generale. Dehan, Mntpellier. CHRISTENSEN, L.P., BIANCHI, M.L., PEACOCK, W.L. & HlRSCHFELT, DJ., Effect f nitrgen fertilizer timing and rate n inrganic nitrgen status, fruit cmpsitin, and yield f grapevines. Am. J. Enl. Vitic. 45, CONRADIE, WJ., Translcatin and strage f nitrgen by grapevines as affected by time f applicatin. In: J.M. RANTZ (ed.). Prc. Int. Symp. n Nitrgen in Grapes and Wine, June 1991, Seattle, USA. pp DE KLERK, e.a., Bilgie van die wingerdfillksera Phyllxera vitifliae (Fitch) Hmptera: Phyllixeridae) in Suid-Afrika. M.Sc. Thesis, University f Stellenbsch, 7600 Stellenbsch, RSA. DE KLERK, e.a., Mrphlgy and taxnmy f the Suth African species f the genus Margardes (Hemiptera: Margardiae), with detailed studies n the bilgy f tw vine-infesting species. Ph.D. Dissertatin, University f Stellenbsch, 7600 Stellenbsch, RSA. FREEMAN, B.M., At the rt f the vine. Austral. Crapegrwer WinemakeI' 232, HUNTER, J.J., RUFFNER, H.P., VOLSCHENK, e.g. & LE ROUX, D.J., Partial defliatin f Vitis vinifera L. cv. Cabernet Sauvignn/99 Richter: Effect n rt grwth, canpy efficiency, grape cmpsitin and wine quality. Am. J. Enl. Vitic. 46, HUNTER, J.J. & LE ROUX, DJ., The effect f partial defliatin n develpment and distributin f rts f Vitis vinifera L. cv. Cabernet Sauvignn grafted nt rtstck 99 Richter. Am. J. Enl. Vilie. 43, HUNTER, J.J., VISSER, J.H. & DE VILLIERS, O.T., Preparatin f grapes and extractin f sugars and rganic acids fr determinatin by high-perfrmance liquid chrmatgraphy. Am. 1. Enl. Vilic. 42, KLIEWER, W.M., BENZ, M. & MORANO, L Rtstck and vine spacing effects n grwth, rt distributin and water utilizatin by Cabernet Sauvignn grapevines in Napa Valley. In: Abstracts Furth Int. Symp. n Cl-Climate Viticulture & Enlgy. LOUBSER, J.T., A mdified sieving-sedimentatin methd fr extracting nematdes frm sil. Hart. Sci. 3, MORLAT, R., REMONe, M. & PINET, P., Influence de la densite de plantatin et du mde d'entretien du sl sur I'enracinement d'un penplement de vigne plante en sl favrable. Agrnmie 4, RAATH, PJ. & SAAYMAN, D., Nitrgen mineralisatin in vineyard sils f the Western Cape as affected by sil management practices. S. Afr. 1. Enl. Vitic. 16,7-13. REYNOLDS, A.G., WARDLE, D.A. & NAYLOR,A.P., Impact f training system and vine spacing n vine perfrmance and berry cmpsitin f Chancellr. Am. 1. Enl. Vitic. 46, RICHARDS, D., The grape rt system. Hart. Rev. 5, SHAULIS, N.J., Respnses f grapevines and grapes t spacing f and within canpies. In: WEBB, A.D. (ed.). Prc. Grape and Wine Centennial Symp., June 1980, University f Califrnia, Davis. pp SOIL CLASSIFICATION WORKING GROUP, Sil classificatin. A taxnmic system fr Suth Africa. Dept. Agric. Develpment, Pretria. SWANEPOEL, J.J. & SOUTHEY, J.M., The influence f rtstck n the rting pattern f the grapevine. S. Aji-..I. Enl. Vitic. 10, VALENTI, L., TONNI, M. & CISANI, E, Effect f training system and vine spacing n vine grwth and prductivity f cv. "Barbera" in S. Clmban Al Lambr (Nrth Italy). First results. In: PONI, S., PETERLUNGER, E., IACONO, F. & INTRIERI, e. (eds). Prc. Wrkshp Strategies t Optimize Wine Grape Quality. Acta Hrt. 427, VAN ZYL, J.L., The grapevine rt and its envirnment. Van Zyl, J.L. (cmp.). Tech. Cmm Dept. Agric. Develp., Pretria, RSA. WINKLER, A.J. & WILLIAMS, W.O., Starch and sugars f Vitis vinifera. Plant Physil. 20,