Growth, Gas Exchange, Water Uptake, and Drought Response of Seedling- and Cutting-propagated Peach and Citrus Rootstocks

Growth, gas exchange, root hydraulic conductivity, and drought response of seedling and rooted cuttings of Lovell and Nemaguard peach [Prunus persica (L.) Batsch], and Carrizo (Poncirus trifoliata x Citrus sinensis) and sour orange (C. aurantium L.) citrus rootstocks were compared to determine the influence of propagation method on these characteristics. Rooted peach cuttings had a higher proportion of root biomass in fibrous roots (≤1 mm in diameter) and lower root : shoot ratios than seedlings, although this did not occur in citrus. Net CO2 assimilation (A) was higher for peach seedlings than for cuttings, but similar for ‘Redhaven’ (RH) scions on either seedlingor cuttingpropagated rootstocks, suggesting that leaf-associated factors were responsible for differences. As in peach, A was higher for Carrizo seedlings than for cuttings, but A was not affected by propagation method in sour orange. Peach seedlings maintained higher A than cuttings as water potentials declined during short-term soil drying, although in citrus this occurred only for Carrizo. RH scions on either root type exhibited similar declines in A as soil dried, indicating the lack of a rootstock effect. Root hydraulic conductivity (Lp) was similar between seedlings and cuttings of all cultivars when expressed on a length basis. Leaf conductance and osmotic adjustment were similar for RH scions on seedlingor cutting-propogated rootstocks during 45 days of drought stress, indicating the lack of a rootstock effect on long-term stress response. Development of new clonal rootstocks or own-rooted scions for peach and other Prunus species is a worldwide concern due to the vast number of soil-related factors limiting orchard productivity, and the need to reduce variation among trees (Rom, 1982, 1983). Currently, there is only minor use of clonal rootstocks for peach, although own-rooted scions have achieved limited commercial success (Couvillon, 1985). Studies examining the behavior of peach scions on several rootstocks vs. own-rooted scions are just beginning to appear in the literature. Survival on short-life sites (Reighard et al., 1990), variation among cultivars in yield efficiency (Hammerschlag and Scorza, 1991), and drought stress response (Couvillon et al., 1989) differed between own-rooted and grafted peach trees. However, flower bud hardiness (Durner, 1990) and winter injury (Ferree and Schmid, 1988) were similar for trees on their own roots and those grafted to common peach rootstocks. Reduced tree size of own-rooted scions vs. grafted trees seems to be consistent across location in peach (Couvillon et al., 1989; Frecon, 1986; Marini, 1985), and this effect has been reported for ‘Bartlett’ pear (Pyrus communis L.) as well (Westwood et al., 1976). Mechanisms that reduce drought stress in peach trees are worthy of investigation, because supplies of irrigation water are becoming limited, and fruit size and yield are directly related to water availability during the final phase of fruit growth (Chalmers and Wilson, 1978). The objectives were to determine if rooted cutting and seedling peach trees differ with respect to root growth, carbon partitioning, canopy gas exchange, root water uptake, and response to drying soil, all factors that may explain differential growth and drought response. Citrus was compared to peach to determine if effects of propagation method were species specific or similar in widely different taxa. Citrus ceived for publication 15 Jan. 1992. Accepted for publication 27 Apr. 1992. contribution of the Georgia Agricultural Experiment Stations, College Stan, Athens. I thank Michael Duemmel for technical assistance. The cost of blishing this paper was defrayed in part by the payment of page charges. der postal regulations, this paper therefore must be hereby marked adverement solely to indicate this fact. 4 rootstocks were ideally suited to this study because seedlings are nucellar and therefore genetically identical to rooted cuttings, whereas in peach, cuttings and seedlings differ genetitally. This comparison permitted determination of the propagation effect exclusive of genetic variability. Materials and Methods Plant material. Terminal semihardwood cuttings and seed of Nemaguard and Lovell peach were collected from the same trees growing on the Horticultural Research Farm in Athens, Ga., in Sept. 1989. Cuttings were rooted as described by Couvillon and Erez (1980) and were 10 to 20 cm long with dry weights of 1 to 2 g after rooting. Seeds were stratified at 4C in moist vermiculite for 3 months and planted in speedling trays in commercial soilless medium (50% peat, 15% vermiculite, 35% perlite) in early Jan. 1990. Seedlings were transplanted to 0.9-liter pots of pure sand on 29 Jan. 1990. Rooted cuttings were placed at ≈ 7C until 25 Jan. 1990 and then transplanted to 0.9-liter pots with sand medium as well. Trees were grown in a greenhouse with ≈70% light transmission, and day/night cycles of 20-30/ 15-25C. Ten cuttings and 10 seedlings of each cultivar were selected for uniform growth and used for all measurements. RH peach was budded onto seedlingand cutting-propagated rootstocks of Nemaguard and Lovell on 9 June 1990 to test the influence of rootstock propagation on gas exchange, water relations, and drought response of a common scion. Chip buds were inserted ≈10 cm above the soil line on 9 June and wrapped with parafilm for 2 weeks. Buds were forced on 23 June, and scions had grown 5 to 20 cm before being placed at 4C on 9 July for 13 weeks. Trees were removed from the cooler in October and grown under greenhouse conditions as described until measurements were made in Jan.-Feb. 1991. Only Lovell rootstocks produced a sufficient number of trees suitable for laboAbbreviations: A, net CO2 assimilation; g, stomatal conductance; Lp, root hydraulic conductivity; NUE, nitrogen-use efficiency; RH, ‘Redhaven’; WUE, water-use efficiency; Ψ, water potential. J. Amer. Soc. Hort. Sci. 117(5):834-840. 1992. ratory analyses similar to nongrafted trees; the RH/Nemaguard trees were used only during the longer-term drought response studies described later. Six RH/Lovell seedlings and four RH/ Lovell cuttings were used for measurements of the same variables as the nongrafted trees. The fewer replications of grafted than of nongrafted trees resulted from lack of bud success, nonuniform growth, and failure to induce a lateral shoot on the rootstock shank, which is necessary for drought response and Lp measurements described later (Rieger and Motisi, 1990). Carrizo citrange and sour orange were used for comparison to peach trees, because both are highly nucellar and commonly used citrus rootstocks. Seed were obtained from a nursery in Florida, and terminal cuttings, ≈20 cm in length (six to 10 nodes), were taken from 1.5-year-old seedlings in Sept. 1989. Leaves were removed from the basal portion of the cuttings, and stems were wounded and dipped in a solution of 2500 μg indole butyric acid/g water. Cuttings were placed in vermiculite flats in a mist bed for 2 months, acclimated to greenhouse conditions for 2 months, then transplanted to 0.9-liter pots of sand on 15 Jan. 1990. Seed of both species were sown in speedling trays of a soilless medium on 19 Dec. 1989, and seedlings were transplanted to 0.9-liter pots of sand on 15 Jan. 1990. For Carrizo, seven replications were used, and six replications were used for sour orange. Again, less replication of citrus than peach was due to lack of uniform growth and lateral shoot induction. All trees were fertilized with half-strength nutrient solution (Jones, 1985) for the first 4 weeks, then full-strength solution twice weekly, alternating with tap water, until measurements were made June-Aug. 1990. Gas exchange and root hydraulic conductivity (described below) were measured by replication within cultivar, i.e., seedlings and cuttings were measured alternatively until all replicates of a particular cultivar were exhausted. Because 2 to 3 weeks elapsed between measurements among cultivars, statistical comparisons were limited to those between propagation method within cultivar. The design was a factorial combination of five taxa × two propagation methods. Data were tested by analysis of variance using PROC GLM (SAS, Cary, N.C.). Gus exchange measurements. Measurements of A, transpiration (E), and stomatal conductance to water vapor (g) were taken on enclosed plant canopies in a semiclosed gas-exchange chamber (Rieger and Motisi, 1990). A vapor pressure gradient (VPG) of ≈2.0 kPa was maintained for all measurements; leaves were maintained at 26 to 29C, and three 400-W metal halide lamps were arranged in a 120° arc to provide saturating irradiance to all areas of the canopy. Irradiance varied from ≈2500 μmol·m ·s at the top of the chamber to ≈1500 μmol·m ·s at the bottom. These conditions produced maximal A and E for all trees in the study. LP. This was estimated on intact trees as described by Rieger and Motisi (1990). Three E rates were induced on each plant through changing VPG and light intensity. After steady-state E was observed for at least 30 min, the water potential gradient (∆Ψ) across the roots was estimated by measuring xylem pressure potential (pressure chamber method) of leaves enclosed in parafilm and aluminum foil at the base of the stems outside the chamber. The rooting medium was saturated (Ψ≈0) throughout the measurement period and had a hydraulic conductivity several orders of magnitude higher than that of the roots. Water potential was regressed on water uptake on a fibrous root length or weight basis to estimate Lp (slope ) and the offset (y-intercept). Response to soil drying. Following estimation of L,, the soil J. Amer. Soc. Hort. Sci. 117(5):834-840. 1992. was allowed to dry for 4 to 7 days, during which time canopy gas exchange and stem water potential were measured to develop a relationship between A and Ψ for each type of plant. Saturating irradiance, a temperature of 26 to 29C, and a VPG ≈2.0 kPa were maintained for each measurement. Linear and nonlinear regressions of A on Ψ were calculated from all points for a given cultivar/propagation method subgroup. Growth measurements. After Lp and gas exchange measurements were made, trees were divided into leaves, stems, large roots (> 1 mm for peach, >2 mm for citrus), and fibrous roots. Length of fibrous roots was estimated by the line-intercept method of Tennant (1975) using a 3 × 3-cm grid. Total root length was estimated first by counting all intercepts, then only suberized roots were counted on a second scan to estimate the fraction of the total root length that was suberized. Leaf area was measured using a LI-3000 area meter (LI-COR, Lincoln, Neb.). Plant materials were dried at 80C for 1 to 2 weeks to obtain dry weights. Leaf nitrogen content was estimated by Kjeldahl digestion and autoanalyzer (Technicon) on dried leaf tissue. Nitrogen use efficiency (NUE) was calculated as A/N, where N is nitrogen content in millimoles per square meter, giving units of mole C/mole N per second. Long-term drought stress. Drought stress was imposed on a separate group of RH scions on either seedlingor cutting-propagated Nemaguard and Lovell rootstocks obtained from the same population as those used for measurements described above. Three trees of each cultivar/propagation method subgroup were watered daily to container capacity, while the other three were given 100 to 150 ml of water every other day to provide a fluctuating water deficit (total = 24 trees). This volume of water allowed trees in the drought stress treatment to wilt between irrigations but did not cause leaf shedding. Trees were maintained under these conditions for 45 days during Jan.-Mar. 1991. Greenhouse conditions were as described above. Leaf conductance was measured using a steady-state porometer (LICOR) at 9:00, 12:00, and 15:00 HR (Eastern Standard Time) on several days during this period to assess the impact of drought stress on stomatal opening. In addition, leaf osmotic potential at full turgor (πο) was measured with a thermocouple psychrometer (Decagon Devices, Pullman, Wash.) at the end of the 45-day stress period to determine if osmotic adjustment occurred. Three leaves were collected from each tree in the morning and placed at 4C with cut ends of petioles immersed in distilled, deionized water. I assumed from previous experience that the water potential of the leaves reached ≈0 after overnight equilibration. Leaves were then held frozen at 80C until πο was determined at a later date. Cell sap osmotic potentials were therefore overestimated, because membrane disruption causes dilution of cell sap with relatively pure apoplastic water. No attempt was made to correct for this dilution, as I assumed that the dilution was similar for all leaves. Leaf conductance and osmotic potential data from long-term stress studies were analyzed separately using analysis of variance for a 2 × 2 × 2 factorial combination of cultivar, propagation method, and irrigation regime, with three single-tree replications (Proc GLM, SAS). The design was completely randomized. Results and Discussion Root system growth and dry weight partitioning. Seedling root systems had a higher dry weight than cuttings for Nemaguard, Lovell, and RH/Lovell, but the opposite effect occurred for the citrus rootstocks (Table 1). In peach, seedlings tended Table 1. Growth characteristics of seedlings and rooted cuttings of peach and citrus and RH peach scions budded onto Lovell seedlingor cutting-propagated rootstocks. NS,*,**,***Paired comparison between cuttings and seedlings within a cultivar nonsignificant at P > 0.05 or significant at P = 0.05, 0.01, and 0.001, respectively. For Lovell and Nemaguard, n = 10; Carrizo, n = 7; sour orange, n = 6; RH on seedling or cutting rootstocks, n =6 and n = 4, respectively. to produce relatively large taproots that comprised most of the root system mass, whereas cuttings produced very few largediameter roots. Consequently, peach seedling root systems had a lower proportion of root biomass in fibrous (≤1 mm) roots than in larger, structural roots, although fibrous root weights were similar for seedlings and cuttings. Similarly, RH/Lovell cuttings had less total root weight and a greater proportion of root biomass in fibrous roots than those on seedling rootstocks. Thus, the larger size of the root system and greater proportion of biomass in large roots of seedlings was conserved from initial growth through budding, cold storage, and regrowth. These morphological features are consistent with those observed previously on larger, nursery-grown peach trees (Couvillon et al., 1989), indicating that pot culture did not qualitatively affect root characteristics. The larger seedling root systems resulted in higher root : shoot ratios for seedlings than cuttings for Nemaguard, Lovell, and RH/Lovell (Table 1). Carrizo and sour orange, in contrast, had higher root : shoot ratios for cuttings than seedlings. A high root : shoot ratio may be beneficial in drought tolerance if all the roots contribute to water uptake equally, indicating better stress tolerance for peach seedlings than cuttings, which is opposite of that previously reported (Couvillon et al., 1989). However, a large portion of seedling root system weights was incorporated into large, woody roots with little surface area, which probably made little contribution to water absorption. A related study on Eucalyptus grandis W. Hill ex Maiden showed that seedlings had a higher root : shoot ratio than cuttings, as found here with peach, and maintained higher daytime Ψ than cuttings (Blake and Filho, 1988). However, the Eucalyptus seedlings also had lower leaf conductance than cuttings, which alone could have been responsible for maintenance of higher Ψ. Probably of greater importance is the fibrous root length : leaf 836 area ratio in regulating plant water status. This ratio was higher for seedlings than cuttings of Nemaguard and Lovell, but was similar for RH/Lovell trees, suggesting that this initial difference is lost after trees are grafted and cut back. Thus, any advantaged seedlings may have over cuttings in the nursery may have little bearing on the subsequent performance of grafted trees. Only for sour orange was the root length : leaf area ratio higher for cuttings than seedlings. Such inconsistencies in results among taxa in this study and between the current and previous studies make conclusions regarding relative drought tolerance of cuttings vs. seedlings and root/shoot balance tenuous. Root length did not differ between nongrafted seedling and cutting peaches but was less for RH/Lovell on cutting than seedling root systems (Table 1). The suberized root percentage was unaffected by propagation method in peach but seemed to increase with plant age, because grafted trees had a much higher suberized root percentage than nongrafted and were also several months older than nongrafted trees. The degree of suberization of roots is known to increase with age in fruit trees and may approach 100% in winter (Atkinson, 1980). In citrus, roots were longer for cuttings than seedlings, but the degree of suberization was different only for Carrizo, where cuttings had three times the number of suberized roots for seedlings. In peach, total plant weight differed between seedlings and cuttings only for grafted trees, where seedling rootstocks produced larger trees than cutting rootstocks (Table 1). This difference was due entirely to the larger root system for trees on seedling rootstocks, because stem and leaf weights of RH scions did not differ with rootstock propagation method (data not shown). Lack of differences in shoot weights between RH scions on cutting and seedling rootstocks may have been due to high variability induced by recent budding and removal of rootstock canopies. Thus, reasons for the smaller canopy size of ownJ. Amer. Soc. Hort. Sci. 117(5):834-840. 1992. rooted vs. grafted peach trees remain unclear but may be related to initial differences in root mass. In citrus, cuttings had higher shoot and root weights than seedlings. This difference may have been due to retention of leaves during and following rooting, resulting in a larger leaf area for a longer duration and consequently greater cumulative carbon gain for cuttings than seedlings over the same growth period. Unlike those of peach, citrus root systems partitioned ≈40% to 45% of the biomass into fibrous roots, regardless of propagation method and despite large differences in root mass. Conflicting results for peach and citrus indicate that growth response to propagation method may be species dependent. Gas exchange. Seedlings had A rates ≈50% higher than cuttings for all nongrafted trees, except sour orange (Table 2). For peach, this difference appeared to be caused by leaf-associated factors, because RH scions with either rootstock type had similar gas exchange characteristics when well-watered and as soil dried progressively (Fig. 1c). This relationship suggests that the juvenile leaf type for peach has a greater capacity for A than the mature leaf type (present on rooted cuttings and grafted trees) regardless of cultivar. The basis for the higher A in Carrizo seedlings than cuttings is unclear, because leaves on both types were similar with respect to leaf morphology, nitrogen content, and NUE (Table 2). Alternatively, higher A for peach seedling leaves may have been related to lower specific leaf weight, lower area per blade (not shown), and/or higher NUE than for cuttings. Higher NUE in peach seedlings occurred despite their lower nitrogen content than in cuttings of both cultivars. This result suggests that juvenile leaves have an inherently greater ability to use nitrogen with respect to A than mature leaves. Cregg et al. (1989) also showed that leaves on seedlings have higher A and different morphology than those on mature trees for several forest species. Table 2. Leaf and gas exchange characteristics of seedlings and root