Temperature studies on kiwifruit vines using relocatable greenhouses

Transcription

1 New Zealand Journal of Experimental Agriculture SSN: (Print) (Online) Journal homepage: Temperature studies on kiwifruit vines using relocatable greenhouses W. P. Snelgar, G. S. Bayley & P. J. Manson To cite this article: W. P. Snelgar, G. S. Bayley & P. J. Manson (1988) Temperature studies on kiwifruit vines using relocatable greenhouses, New Zealand Journal of Experimental Agriculture, 16:4, , DO: 1.18/ To link to this article: Published online: 24 Jan 212. Submit your article to this journal Article views: 19 View related articles Citing articles: 8 View citing articles Full Terms & Conditions of access and use can be found at

2 New Zealand Journal of Experimental Agriculture Vol. 16: /88/ $2.5/ Crown copyright Temperature studies on kiwifruit vines using relocatable greenhouses W. P. SNELGAR G.S.BAYLEY P.J.MANSON Division of Horticulture and Processing Department of Scientific and ndustrial Research Private Bag, Auckland, New Zealand Abstract A set of eight relocatable greenhouses has been built for studies of temperature effects on mature kiwifruit vines (Actinidia deliciosa (A. Chev.) C. F. Liang et A. R. Ferguson) growing in a typical orchard. Each greenhouse covers one half of a vine planted at 6 m within-row spacing on at-bar trellis. Temperatures within the greenhouses are maintained at a constant differential above ambient air temperature. Details of the construction, operation, and performance of the greenhouses are provided. n an initial experiment carried out in spring 1985, the mean air temperature was increased by.7 C or 4.9 C for 41 days. At the higher temperature, bud burst was advanced by 7 days and full bloom by 14 days. n the warmer greenhouses, growth of the apical shoot was increased and the percentage of shoots which produced flowers was reduced. An interaction between the warmed and untreated halves of the vines was observed. Keywords Actinidia deliciosa; kiwifruit; flowering; temperature; greenhouses; flower production NTRODUCTON The need for detailed information on the temperature responses of kiwifruit vines was recognised some time ago (Davison 1977). Because of the large size of kiwifruit vines and the lack of a small-test plant system as used for grapes (Mullins & Rajasekaran 1981), some of the earlier studies on kiwifruit were carried out using unrooted cuttings (Brundell976). Received 2 May 1988; accepted 12 August 1988 While this approach has yielded valuable information, there are limitations to the usefulness of this technique. More recently, other workers have studied the temperature responses of small potted kiwifruit vines in a phytotron (Laing 1985; Morgan et al. 1985; Warrington & Stanley 1986). n these studies, vines were grown under a wide range of temperature and light regimes, and suggested optimal ranges for growth and photosynthesis determined. However, for phytotron work, the vines are necessarily small and cannot be trained and pruned in an orthodox manner. This makes it difficult to obtain the large quantities of fruit needed for detailed studies of fruit quality and storage life. Furthermore, the results obtained in phytotrons may not be directly applicable to vines in a commercial orchard. n order to overcome such difficulties, some researchers have studied plant responses by manipulating temperatures in the field. This can be accomplished using fixed glasshouses (Squire 1978; Monteith et al. 1983), or semi-mobile units (Millar etal.1984). However, these systems have only been used on annual crops, which are more easily managed in a glasshouse, or on small trees, which can be moved into the glasshouse for a short-term experiment. A further step towards true field studies is the use of portable enclosures which can be taken into standard orchards and fitted to mature trees (Lilleland 1936; Tukey 1952; Tukey 1958). For practical reasons, most of the earlier units were limited to covering only a small part of a tree, and often onl y night temperatures could be modified. An exception is the portable limb enclosure described by Mellenthin & Bonney (1972). These units are larger, easily moved, allow normal diurnal temperature fluctuations, and provide for the continuous monitoring of temperatures. The present work describes a set of eight relocatable greenhouses which can be used to modify the temperature around mature kiwifruit vines trained on at-bar trellis, in a typical orchard block. This system will be used to study the responses of kiwifruit to short periods of raised temperatures applied at various stages of growth.

3 33 New Zealand Journal of Experimental Agriculture, 1988, Vol. 16 MATERALS AND METHODS Experimental greenhouses The purpose-built greenhouses have a curved roof and are 3.2 m high at the apex, 2.8 m long and 3.7 m wide (Fig. la). The frame is constructed from 2 mm galvanised iron waterpipe joined by steel clamps. For ease of assembly and disassembly, alloy 'quick fit' clamps (K wik-grip are used on all joints that are not permanent Each frame is composed of four polythene-covered panels (.2 mm thick' Agphane 11 ') which are assembled around the wires of the vine support structure (T -bar trellis) to form the ends of the greenhouse. The two ends are then connected together by lengths of pipe which also form a support system for the roof cover. Finally, the roof and sides of the greenhouse are covered with a 3 X 9 m sheet of polythene. This is held in place by two-piece aluminium locking strips (polylock extrusion) at the sides and by ropes inside hems at the ends of each unit. A roof (and side) cover can be installed or removed by two workers in less than 12 min. The eight greenhouses can be erected or removed by two workers in about 2 days. There is a 1.3 X.8 m door in the north end of each greenhouse to allow entry for servicing and plant measurements. Temperature control and ventilation The temperature control and air circulation equipment for each greenhouse is mounted in a.6 X.6 X 1.2 m plywood box which is placed inside the greenhouse with one end projecting through the southern end wall. The equipment (see Fig. b) consists of: - One 38 W, 46 mm diam. axial flow fan (Rita EF18141) rated at 1.8 m 3 /s. This runs continuously. A 1 X.6 m diam. perforated polythene tube fixed to the exit of the fan box ensures an even distribution of air throughout the greenhouse. - A.6 X.6 m vent and aluminium louvre system directly in front of the fan. The louvre is connected via a rod to a second louvre (.8 X.8 m double walled polycarbonate) which is immediately above the first, at the apex of the greenhouse. The setting of this louvre system is adjusted automatically via a single damper actuator (Honeywell sm 24-srl). - One custom built 1.8kW finned heating element. - A purpose-built electronic control unit fitted with two temperature sensors (AD59). One sensor measures the outside air temperature in a unaspirated screen (white PVC tube), and the other monitors the greenhouse air temperature in an aspirated screen (Young 1978). The controller maintains the greenhouse temperature at a preselected (-1 C), constant differential above ambient During any experiment, two differentials can be preset and the controller can be programmed to switch from one to the other automatically. Thus it is possible to set night and day differentials independently. n practice, temperature modification within the greenhouse is limited by the capacity of the heater unit to approximately 4-5 e above ambient during the night and 1 e during the day. Control logic and operation f the greenhouse temperature differential is above the set value, both louvres are opened and the fan draws in fresh air through the lower vent and forces it through the greenhouse, thereby forcing warm air out of the top vent. When operating at full speed the fan is theoretically capable of replacing all of the air in the greenhouse every 16 s. The louvre system is under modulated control via a 'read and hold' circuit which adjusts the louvre setting every 9 s. The controller circuit also allows the operator to set a limit on vent closure in order to ensure that the vents are not completely closed during the day. This limit can be switched in or out at.25 h intervals with a 24 h time switch. f the greenhouse is below the set temperature the louvres are progressively closed. When completely closed, the fan recirculates air within the greenhouse. When necessary the heater, which is under proportional control, is turned on. n order to minimise wear on the mechanical and electronic components both the louvre and heater control circuits have a ±.5 C deadband.. Experimental design and methods An initial experiment was carried out on the Department of Scientific and ndustrial Research (DSR) Reseach OrchardatKumeu, Auckland,New Zealand (36 44'S, ) in spring The vines were 9 years old and were cv. Hayward grafted onto seedling rootstocks and planted at 6 m withinrow spacing. All vines in the block were irrigated by minisprinklers (one per vine). The greenhouses were installed on eight vines all in the same row and temperature regimes were assigned randomly within

4 Snelgar et al.-relocatable greenhouses 331 (a) (b) F. A l B '\ \ E \ "- DC "- \ "- /- Fig. 1 (a) Side view of a greenhouse positioned over a dormant vine. Roofcover is partially removed. (b) Scale diagram of greenhouse (side view). Arrows indicate direction of air flow. A, door; B, perforated polythene tube; C, heating element; D, fan; E. bottom louvre (aluminium); F. upper louvre (transparent polycarbonate).

5 332 New Zealand Journal of Experimental Agriculture, 1988, Vol. 16 pairs of consecutive vines. The two temperature regimes used were: Ambient = air temperatures were maintained as close as possible to the outside air temperature Warm = air temperatures were maintained 5 C above outside air temperature both day and night. The responses of the half-vines outside each greenhouse were also monitored. These are denoted ambient outside and warm outside treatments. The greenhouses were installed on 27 August, approximately 2 weeks before bud burst, and removed on 7 October; when bud burst was complete. Since the vines were 6 m long, each greenhouse covered one half of a vine. n order to minimise any shading effects the greenhouses were always positioned over the southern half of each vine. The standard minisprinklers used for irrigation were replaced by two smaller sprinklers, one inside the greenhouse, and one outside. Environmental monitoring Temperatures and photsynthetically active radiation (PAR) were monitored using an electronic datalogger. Air temperatures (1.6 m above ground level, aspirated screen) in the centre of each greenhouse, and near a control vine (unaspirated screen, see Henshall & Snelgar 1988), were measured using thermistors. These were scanned every 3 s and the hourly means were recorded on cassette. Soil temperatures 15 cm below the surface were monitored in two greenhouses, and under an untreated vine. The PAR levels incident on the orchard, and inside two representative greenhouses, were measured every 6 s (LCOR 19s quantum sensors) and recorded hourly. Vine and fruit measurements Five canes were tagged on each half of each vine. At 3-5 day intervals during spring, bud burst (as defined by Brundell 1975a), the number of flowers at full bloom (see Brundell975b), and the length of the apical shoot were determined for each cane. The total number of flower buds in each leaf axil of each shoot were counted during the last week of October, about 2-4 weeks before full bloom. Analyses of the effect of temperature on shoot fruitfulness were made separately for the apical node (shoot), then on consecutive groups of three nodes down to nodes Canes which lost shoots because of wind damage were excluded from these analyses. A minimum of 14 canes in each treatment remained. Flowers inside warm greenhouses were hand pollinated since they reached full bloom before male vines in the block had flowered. The growth of five fruit per vine (from the halves of the vines which had been inside the greenhouses) was monitored at 1-4 week intervals by measuring fruit diameter. After harvest, seeds from these fruit were extracted by forcing the ripened fruit through a sieve. Seed counts were made by weighing the air-dried seeds then weighing a 1 seed subsample. Fruit quality and storage Two weeks and one week before harvest, five fruit from each of the treated vines were picked and the flesh firmness and soluble solids concentrations of each fruit were measured. Flesh firmness was measured with an Effigi type penetrometer (7.9 mm head) on the flat and rounded faces of each fruit after the skin had been removed. Soluble solids concentrations were measured at each end of the fruit using an Atago hand-held refractometer. At harvest, 15 fruit from each half vine were packed into standard kiwifruit trays and stored at O C. Flesh firmness and soluble solids concentrations were measured on the day of harvest, and after 4 and 8 weeks cool storage. The significance of results was tested by anal ysis of variance using the Minitab and Genstat statistical packages. RESULTS Greenhouse performance The transmittance of the greenhouses was estimated by regressing PAR levels inside the greenhouses against incident PAR. During 1 days of measurement, the mean transmittance was 85% in both warm and ambient greenhouses. Fig. 2 gives typical plots of the temperature differentials recorded in two greenhouses at a preset differential of approximately 5 C temperatures were generally within ± 1 C of the set difference (Fig. 3). n the greenhouse operated at a differential ofo C, the temperature was occasionally2-3 C higher than the outside temperature during the day. These fluctuations effectively increased the daily maxima but had only a minor effect on the overall mean temperature (Table 1). Previous trials with a prototype greenhouse demonstrated that the vertical temperature gradient within a greenhouse is small. Mean temperatures (in aspirated screens) during a -week run were 22.6 C

6 Snelgar et al.-relocatable greenhouses 333 Fig. 2 Typical variations of air temperature within two greenhouses compared with outside air temperature. 25, ,---_, ,_--_, 2, U 15 :J E - o -- - Air temperature Warm differential Ambient differential, ' J-\ Date September 1985 Table 1 Summary of temperatures inside greenhouses (warm or ambient) and outside. Vines were covered from 17 August to 7 October, Temperature Mean regime temperature Max. Min. Outside Ambient Warm From outside daily maxima and minima. Mean daily differential Mean soil temperature Max. Min. (15 cm, 6 Sep---2 Oct) m above ground level, 22.4 at 1.6 m, and 22.3 at 2.8 m. Differences between hourly readings at the standard sampling height (1.6 m) and the lower and upper heights were also small (94 % of the readings within ± O.7 C). The mean soil temperature at 15 cm was increased by.4 C in the ambient greenhouse and 3.1 C in the warm greenhouse. Vine measurements The warm temperature treatment advanced bud burst by 7 days and full bloom by 14 days in comparison with the ambient treatment (Fig. 4). Bud burst and flowering on the halves of the vines not inside the greenhouses coincided with that of the vines in the ambient greenhouses (data not presented). The rate of growth of the apical shoot was greater on warmtreated vines than on ambient-treated vines. Bud burst on canes in warm greenhouses was only 25% complete when the apical shoot was 4 cm long, while in the ambient greenhouses bud burst was 63% complete at this stage of shoot development. Shoot growth on the canes outside the greenhouse was similar to that inside the ambient greenhouses (data not presented). The temperature treatments applied did not affect the percentage of buds which burst, or the number of flowers per flowering shoot (Table 2). However the warm treatment significantly reduced the percentage of shoots that produced flowers, and thus the number of flowers per winter bud. The effects of the temperature treatments applied on flower production varied with position of the shoot

7 '--' >- u c ::J '"... ll '--' >- u c ::J '" l- ll r-- 1 o f-- r- r- -- r-- -rilh New Zealand Journal of Experimenral Agriculture, 1988, Vol. 16 (a) Ambient greenhouse (b) Warm greenhouse J - ;--- ;--- r- r- - - h o Temperature differential (oc) Fig. 3 Histograms of hourly differences between outside air temperature and greenhouse temperature for the greenhouses depicted in Fig. 2. Data covers entire experimental period. The ambient differential was set ato C, and the warm differential at 5 C. Table 2 Bud burst and flower bud production in kiwifruit vines maintained at.7 C (Ambient) or 4.9 C (Warm) above the prevailing air temperature for 41 days during spring. Only half of each vine (inside) was treated, the remaining half (outside) being left as a control. Canes were 1.7 m long with 21 buds per cane. Temperature % Bud % Flowering % Fruitful Flowers per Flowers per regime burst shoots bud burst flowering shoot winter bud Ambient, outside Ambient, inside Warm, outside Warm, inside LSD (P <.5) NS 14 7 NS.34 on the cane. n each treatment, almost all of the apical shoots produced flowers (Fig. 5a). However on canes in the warm greenhouses, buds proximal to the apex generally produced a high percentage of vegetative shoots. A similar but less marked tendency was recorded on canes from ambient greenhouses. On shoots positioned on the basal half of warm- or ambient-treated canes, the number of flowers per shoot tended to be low (Fig. 5b). When shoots which did not flower are excluded from the analysis, canes from the warm greenhouses still tended to produce fewer flowers per shoot (Fig. 5c). n contrast, the apical shoots on canes outside warm greenhouses produced significantly more

8 Snelgar et al.-relocatable greenhouses Fig. 4 Relationship between bud burst, extension of the apical shoot (broken lines), and flowering for 5 kiwifruit vines maintained O.7 C E ( (Ambient) or 4.9 C (Warm) above.:;.. the prevailing air temperature for.<: 4 41 days during spring. Arrows, c indicate the period during which greenhouses enclosed vines. 3.<: n '- 2 -n '- :J. "'(J :J ( 1 Sep )1 i i i '" i? Shoot length Oct Date 1985 Nov Flowering Dec _ ::J E ::J () 4 '-../ E 3. ::J '> c ()!... CL (J)!... >: ll. flowers than the apical shoots of canes from any other treatments. Fruit growth ripening and storage Early in the growing season, the mean diameter of fruit on vines which had been warmed was significantly greater than that of fruit on ambienttreated vines (Fig. 6). This difference decreased as the fruit enlarged and by harvest the diameters were equal. However measurements made on the day of harvest showed that fruit from the warm greenhouses were longer (65 v.6mm,p<o.o)andheavier(l6 v. 99 g,p<.5) than fruit from ambient greenhouses. Fruit from warm greenhouses had more seeds than those from ambient greenhouses (16,68 respectively' P<O.OO). Before harvest and at harvest, fruit from warm greenhouses had higher soluble solids concentrations than fruit from ambient greenhouses (Table 3). However no differences were found after fruit had been in cool-store for 8 weeks. DSCUSSON The greenhouses The eight greenhouse units functioned reliably and controlled the temperature within acceptable limits throughout the experimental period. Light transmission of the greenhouses was high, and was not influenced by the different temperature regimes imposed. The results of Snelgar & Hopkirk (1988) imply that reducing PAR levels by 15 % is unlikely to measurably affect the flowering and cropping behaviour of kiwifruit. Temperature response of kiwifruit n most instances, vine response was related directly to the prevailing temperature, and interactions between the enclosed and the unenclosed halves of the vines were not observed. However one variate, the number of flowers on the apical shoot, was significantly increased on the northern half of a vine when the southern half of the vine was warmed. This may have resulted from the southern half breaking bud, and developing foliage, before the northern half. Kiwifruit have been shown to translocate carbohydrates over distances of several metres (Snelgar et al. 1986), and research on other fruit trees suggests that an enhanced supply of carbohydrate may increase flowering (Goldschmidt et al. 1985). nteractions between different parts of plants have not been noted by other authors using similar portable enclosures on apricots (Lilleland 1936), apples (Tukey 1956),pears (Mellenthineta.1972), and cherries (Tukey 1952). t is not clear whether this difference in response is a result of plant differences, or differences in experimental techniques, since sometimes only a small part of the whole plant was treated. Furthermore, data on the untreated portions of plants have not always been presented. n future work on kiwifruit, our relocatable greenhouses will be used in orchard blocks planted at 3 m spacing. This will allow us to carry out

9 336 New Zealand Journal of Experimental Agriculture, 1988, Vol Ambient out ---- Ambient n,, '" 8 -- Worm out 'P,,... ', ---- Worm in ),,,.c 'l-----q () '" 6 ", n > :;:;... 4 > >,,, 2, c '"... 8 c '" 4 G: 1-//,,,,, -fr---:/p-'" 1_-- " _ (b) Fig. 5 Relationship between position on the cane and (a) percentage of vegetative shoots, (b) number of flower buds per shoot, (c) number of flower buds per flowering shoot, for kiwifruitvines maintained.7 C (Ambient) or 4.9 C (Warm) above the prevailing air temperature for4! days during spring. Only half of each vine was treated (inside), the remaining half (outside) being left as a control. Error bars depict the LSD at P =.5. For clarity LSD's are given only when the ANOV A indicated that a difference was statistically significant c '" >1 c... ;;:: c..... '" 4 2 G: Shoot position on cane Apex 2 (c) 25 Base temperature studies on mature vines that are fully enclosed in the greenhouse. ncreasing the mean air temperature by 4.9 C advanced the time of full bloom by 14 days. Although this early flowering resulted in fruit being larger early in the growing season the growth rate of these fruit was slower than that of fruit at ambient temperature, and by harvest there was no difference in the mean diameters of the fruit. This is surprising, since the fruit from ambient greenhouses were poorly pollinated and this factor alone could be expected to cause a 16 g difference in mean fruit weights (see Pyke & Alspach 1986). The differences in fruit shape between the two temperature treatments (fruit from the ambient greenhouses were less elongate) may also have been a consequence of poor pollination, rather than a direct effect of temperature on fruit shape.

10 Snelgar et al.-relocatable greenhouses 337 Fig. 6 ncrease in diameter of kiwifruit on vines which had been 6 maintained.7 C (Ambient) or 4.9 C (Wann) above the prevailing air temperature for 41 days during,,--... spring. Error bars depict the LSD E 5 atp=.5. E '--"! E 4 =Ambie " Warm +- ::l 3!... u... 2 Dec Jan Feb Mar Apr May Month 1985/86 Table 3 Soluble solids concentrations and firmness of fruit from kiwifruit vines maintained at.7 C (Ambient) or 4.9 C (Wann) above the prevailing air temperature for 41 days during spring. Only half of each vine (inside) was treated, the remaining half (outside) being left as a control. Soluble solids concentration (%) Ambient, outside Ambient, inside 6.3 Wann, outside Wann, inside 6.8 LSD (P <.5).3 Flesh firmness (kg/f) Ambient, outside Ambient, inside 9.8 Warm, outside Wann, inside 8.8 LSD (P <.5).6 Date of measurement 29 Apr 5 May 12 May 17 Jun 15 Jul (-2 weeks) (-1 week) (Harvest) (+4 weeks) (+8 weeks) NS NS NS NS By April 29, the fruit from warmed vines had soluble solids concentrations.5% higher than fruit from vines in, ambient greenhouses. Since soluble solids levels were increasing at a rate of.4 %/week this implies that warming the vines during spring advanced the date of commercial maturity (6.2% soluble solids) by about 1 week. Such an increase would be of considerable practical importance in areas where the risk of preharvest frost is high. The earlier commercial maturity appears to result from an earlier maturation of the fruit, rather than a higher concentration of total carbohydrate, since fruit from both treatments reached the same level of soluble solids as a result of ripening during storage. ncreasing the ambient temperature by 4.9 C (compared with O.7 C) did not alter the percentage of apical shoots which produced flowers, or the number of flowers per apical shoot. However the

11 338 New Zealand Journal of Experimental Agriculture, 1988, Vol. 16 higher temperature decreased the number of flowering shoots proximal to the apical shoot by 36%. This reduction was correlated with earlier and more rapid growth of the apical shoot at a time when many of the proximal buds had not reached bud burst These findings appear to conflict with those of Warrington & Stanley ( 1986) who reported that the proportion of fruitful shoots was highest under their warmest regime (12 C before bud burst and 16 C afterwards). t is possible that, in their study, the critical period of apical shoot development occurred before the vines were moved to the warmer regime. The low percentage of flowering shoots on the phytotron-grown vines may also have influenced their results. The fmdings of the present work corroborate the suggestion of Grant & Ryugo (1982), that the flowering potential of proximal buds is eliminated at an early stage during the growth of the apical shoot. Although mean spring air temperatures of 17 C are unlikely to occur in the areas of New Zealand in which kiwifruit are grown, effects similar to those observed in this work could occur in warmer countries and may contribute to the poor flowering usually attributed to a lack of winter chilling. Furthermore, the relationship between growth of the apical shoot and suppression of flower production may be influenced by the duration of the critical stage of apical influence. Hence warm periods which force the apical bud to burst early, followed by cooler periods which slow shoot growth but also delay the breaking of other buds, may also tend to suppress the formation of flowers. ACKNOWLEDGMENTS We thank Dr. K.Young andmr R M"faggart for assistance with designing the controller units, and the staff of the Mount Albert Research Centre workshop for building these units. Dr. R.M. Davison is thanked for initiating this project. ThefmancialsupportoftheNewZealandKiwifruit Authority is gratefully acknowledged. REFERENCES Brundell, D. J.1975a: Flower development of the Chinese gooseberry (Actinidia chinensis Planch.). Development of the flowering shoot. New Zealand jourtul of botany 13: b: Flower development of the Chinese gooseberry (Actinidia chinensis Planch.) ll. Development of the flower bud. New Zealand jourtul of botany 13: : The effect of chilling on the termination of rest and flower bud development of the Chinese gooseberry. Scientia horticulturae 4: Davison, R. M. 1977: Some factors affecting flowering and cropping in kiwifruit. Pp 23-27, in: Proceedings of the Kiwifruit Seminar, Tauranga, New Zealand Ministry of Agriculture and Fisheries, Tauranga. Grant, J. A.; Ryugo, K. 1982: nfluence of developing shoots on flowering potential of donnant buds of Actinidia chinensis. Hortscience 17: Goldschmidt, E. E.; Aschkenazi, N.; Herzano, Y.; Schaffer, A. A.; Monselise, S. P. 1985: A role for carbohydrate levels in the control of flowering in citrus. Scientia horticulturae 26: Henshall, W.R.; Snelgar, W.P. inpress: Asmallunaspirated screen for air temperature measurement. New Zealand JourTUl of crop and horticultural science 17: Laing, W. A. 1985: Temperature and light response curves for photosynthesis in kiwifruit (Actinidia chinensis) cv Hayward. New Zealand jourtul of agricultural research 28: Lilleland,J.1936:Growthstudyoftheapricotfruit.llThe effect of night temperature. Proceedings of the American Society for Horticultural Science 33: Mellenthin, W. M.; Bonney, D. 1972: A portable limb enclosure for temperature modification of tree fruits. Hortscience 7: Mellenthin, W. M.; Wang, C. Y.; Wang, S. Y. 1972: nfluence of temperature on pollen tube growth and initial fruit development in d' Anjou pear. Hortscience 7: Millar, J. M.; Wickenden, M.F.; Jackson, J. E. 1984: Effects of pre-blossom temperatures on Cox fruit set potential. EastMalling Research Station Report for 1984 p. 11 Monteith, 1. L.; Marshall, B.; Saffell, R. A.; Clarke, D.; Gallagher, 1. N.; Gregory, P. 1.; Ong, C. K.; Squire, G. R.; Terry, A. 1983: Erivironmental control of a glasshouse suite for crop physiology. JourTUl of experimental botany 34: Morgan, D. C.; Warrington,. 1.; Halligan, E. A. 1985: Effect of temperature and photosynthetic photon flux density on vegetative growth of kiwifruit (Actinidia chinensis). New Zealand jourtul of agricultural research 28: Mullins, M. G.; Rajasekaran, K. 1981: Fruit cuttings: Revised method for producing test plants of grapevine cultivars. American jourtul of enology and viticulture 32: Pyke, N. B.; Alspach, P. A. 1986: nter-relationships of fruit weight, seed number and seed weight in kiwifruit. New Zealand agricultural science 2:

12 Snelgar et al.-relocatable greenhouses Snelgar, W. P.; Thorp, T. G.; Patterson, K. J. 1986: Optimal leaf:fruit ratios for fruit growth in kiwifruit. Acta horticulturae 175: Snelgar, W. P.; Hopkirk:, G. 1988: Effect of overhead shading on yield and fruit quality of kiwifruit (Actinidia deliciosa). Journal of Horticultural Science 63: Squire, G. R. 1978: A simple temperature-controlled glasshouse for field experimentation in the tropics. Experimental agriculture 14: Tukey, L. D.1952: Effect ofnighttemperature on growth of the fruit of the sour cherry. Botanical gazette 14: : Some effects of night temperatures on the growth of M9ritosh apples, 1. Proceedings of the American Society for Horticultural Science 68: : Effects of controlled temperatures following bloom on berry development of the Concord grape (Vilis labrusca). Proceedings of the American Society for Horticultural Science 71: Warrington,. J.; Stanley, C. J.1986: The influence of preand post-budbreak temperatures on flowering in kiwifruit. Acta horticulturae 175: Young, K. 1978: A simple psychrometer for continuous field use. Journal of agricultural engineering research 23: