Abstract
In nursery production, small root balls are transplanted into larger containers and grown for sale or further transplanting into still larger containers. When a root ball is smaller than a container, the amount of plant-available water (PAW) is initially limited to that of the original root ball. With growth, roots colonize new substrate and thereby increase the volume of water available to a shoot. Because of hydraulic gradients in container substrates, PAW is not linearly proportional to the volume of substrate occupied by roots. To practice precision irrigation in nursery production, it is important to know the extent of PAW and how it changes with growth. A method is detailed that calculates in situ PAW in containers based on changes in actual evapotranspiration while irrigation is withheld. The method is applied under field conditions and requires only daily mass loss measurements and corresponding reference evapotranspiration. An example of how PAW changes during production from rooted cuttings to marketable size plants is provided.
Commercially acceptable growth rates of woody ornamental nursery stock can be achieved using managed allowed deficits, i.e., deficit irrigations, of 20% to 40% plant-available water (PAW) before initiating irrigation (Beeson, 2006; Welsh and Zajicek, 1993). Deficit irrigation can reduce irrigation frequency (Beeson, 2006), thereby reducing nursery water runoff and energy consumption. Using deficit irrigation for container production of nursery stock is especially valuable for water conservation when rain occurs, because rainfall can account for part or all of a plant's daily evapotranspiration. Rainfall is more effective in penetrating plant canopies than overhead sprinkler irrigation (Beeson and Yeager, 2003), being nearly 100% efficient and more uniform than overhead irrigation. Therefore, less rainfall is required to return a container to capacity than overhead irrigation. However, to apply deficit irrigation strategies, or determine how much rainfall is sufficient and when additional supplemental irrigation is required, knowledge of PAW is necessary. This is especially critical during the early stages of plant production, when roots have not completely colonized a container's volume. At this stage, the lower substrate may be at or near 100% container capacity, whereas PAW may be critically low within the actual root ball.
PAW can be found by first recording the mass of a plant/container with a saturated substrate from which excess water has been allowed to drain. The plant is then allowed to transpire until wilted. PAW is calculated as the difference in mass between saturated and wilted conditions (Beardsell et al., 1979b). For a container with a full root ball, total PAW (PAWTotal) can be consistent across species if the container size and substrate are identical (Beeson, 2006).
Although a similar approach can be applied to plants growing in a container in which roots have not yet filled the container volume, there are drawbacks. Stressing a plant to physical wilt can cause leaf drop in some species and may stunt shoot growth after stress is relieved. Because roots are growing into uncolonized substrate, volume of PAW continues to increase, delaying shoot wilt and therefore skewing the estimate of PAW. In situations in which root colonization of a substrate is not complete, PAW can be termed effective PAW (PAWEffec). This skewing is especially prominent when evaporative demand is low during the drying period. For evergreen species that are dormant or that have rigid leaves and stems, determining the point of plant wilt can be difficult. Applying supplemental irrigation that exceeds the PAWEffec is a waste of water and can lead to excessive leaching of nutrients. Overestimating PAWEffec would result in underirrigation and perhaps commercially unacceptable growth rates. Precision irrigation of container plants requires an accurate estimate of PAWEffec.
An alternative for plants that have not colonized a container's volume is to determine root ball volume relative to total container volume. Volume of a developing root system can be easily determined by water displacement. One could then calculate a percentage of container volume and infer PAWEffec if PAWTotal is known. However, water distribution in container substrates varies by substrate composition (Beardsell et al., 1979a; Bilderback et al., 1982; Richards et al., 1986), container size (Bunt, 1976), and container configuration (Bilderback and Fonteno, 1987; Davis et al., 1964; Spooner, 1974; Williams, 1978). In general, hydraulic gradients are established by opposing forces of capillary rise from water interacting with pores within a substrate and gravity pulling water to the bottom of a container (Bilderback and Fonteno, 1987). Although gradients can be determined, they vary with substrate composition and root ball height (Bilderback and Fonteno, 1987). Thus, simple determination of percent root ball volume of a developing root system in a container cannot accurately predict PAWEffec, even if PAWTotal is known.
With increasing demands being placed on landscape nurseries to reduce ground or watershed water withdraws for irrigation, and to reduce nutrients in runoff water, nurseries should begin practicing precision irrigation, applying only as much water as needed, when needed. To do so, knowledge of PAW accessible at a given stage of growth is needed to accurately apply deficit irrigation strategies. Previously, there has been no research in this area for container-grown nursery crops. Objectives of the research presented here were to: 1) verify a nondestructive method for determining PAW in containers in situ, 2) verify the method was valid across plant species, and 3) to examine relationships between PAW and canopy variables among species. To demonstrate the use of this method, a partial data set from an experiment to quantify water use of Viburnum odoratissimum was used to calculate PAWEffec and PAWTotal during production from rooted cuttings to market-sized plants.
Materials and Methods
In mid-Oct. 1997, four plants of five species of woody ornamentals were selected and placed in an open-sided, single-layer clear polyethylene-covered shelter to exclude rain fall. Species were selected for their range of plant size, representation in nursery production, and commercial reputation for drought tolerance or intolerance in containers. Species selected, as ranked from least to most drought-tolerant, were Viburnum odoratissimum, Ligustrum japonica, Photinia × fraseri, Elaeagnus pungens, and Rhaphiolepis indica. Each plant had been grown to marketable size in an 11.4-L black polyethylene container (Lerio Corp., Kissimmee, FL) at the Central Florida Research and Education Center in Sanford, FL during the previous 2 years. Substrate consisted of 64% composted pine bark, 27% Florida sedge peat, and 9% coarse sand by volume amended with 0.88 kg·m−3 micronutrients (Peter's Fritted Trace Elements; Scotts Company, Marysville, OH) and 2.9 kg·m−3 dolomite limestone blended by a commercial vendor (Sunrise Landscape, Apopka, FL). Each plant was placed in a suspension lysimeter (Beeson, 2006) and suspended ≈5 cm above the ground. Lysimeters were spaced such that there was ≈1.5 m between plants. Briefly, a lysimeter consisted of a 22.7-kg load cell (SSM-50; Interface, Scottsdale, AZ) suspended from a tripod 2.2 m in height. A plant was placed in a basket suspended from a load cell by small galvanized chains. Load cells were connected to a system of multiplexers (AM-32 and AM-416; Campbell Scientific, Logan, UT) and a data logger (CR-10; Campbell Scientific) that measured and recorded individual mass changes every half hour. Each plant was irrigated at 0100 hr using a single-spray stake (26.6 L·h−1; Roberts Irrigation Products, San Marcos, CA). Plants were irrigated to supersaturate the substrate each night for three nights before the experiment began.
After irrigation on 21 Oct., nightly irrigation was suspended and an initial maximum container mass was recorded at 0500 hr for each plant. This allowed time for excess water to drain. Each plant was then allowed to transpire to the point of leaf wilt, at which point it was harvested the next morning within 1.5 h of sunrise. Leaf wilt was defined as leaves on three or more branches drooping by 1000 to 1200 hr. Shoots on all plants were mature and all buds were dormant throughout data collection. Just before shoot harvest, lysimeter mass was recorded (LMwilt). The shoot was then removed at the substrate surface and wilted mass of the shoot (Shootwilt) determined immediately using an electronic balance (PB5001; Mettler, Mettler-Toledo, Switzerland). Plant shoots were then dried at 65 °C until a constant dry weight was obtained (Shootdwt). Once a shoot was harvested, a spray stake was returned to a container and nightly irrigation resumed. After the last plants were harvested (4 Nov.), irrigation duration was increased to resaturate all root balls within containers. By 14 Nov., all containers had exhibited a near constant mass for three or more consecutive days at 0630 hr, 2.5 h after the last irrigation. Container mass at this time was recorded as the saturated root ball mass (RBsaturated).
Daily actual evapotranspiration (ETA) for each plant was determined by subtracting mass recorded for each plant at midnight by that recorded the next midnight. Daily ETA for each plant was summed to give a cumulative ETA for each plant. The difference between maximum mass the day of harvest (at midnight) and mass just before harvest was included in a cumulative ETA total.
During the drydown period, daily recordings of net solar radiation (LI-190 pyranometer; LI-COR, Lincoln, NE), maximum and minimum temperature (HMP45C; Campbell Scientific), and daily cumulative wind distance (model 014 anemometer; Met-One Instruments, Medford, OR) were recorded using a data logger (CR10; Campbell Scientific). From these, daily potential evapotranspiration (ETP) was calculated using a regional modification of the Penman-Monteith equation (Jones et al., 1984).
Calculations.
Subtraction of LMwilt from maximum hydrated lysimeter mass (LMsaturated) calculated PAWTotal within the saturated container-plant system (Eq. 1).
Subtraction of resaturated root ball mass (RBsaturated) from LMsaturated calculated the shoot mass at presumably 100% relative water content (Shootmax; Eq. 2).
Shoot relative water container at wilt (RWCwilt) was calculated as follows (Eq. 3):
where Shootwilt is shoot mass at wilt, Shootdwt is the shoot dry weight, and Shootmax was calculated from Eq. 2.
To evaluate the method to determine PAW in situ, daily ETA was normalized by dividing each day's ETA by the corresponding ETP (Eq. 4). This calculated a normalized ETA (ETAmm −1) with the units of mL·mm−1 ETP.
Normalized ETA on the first day of the dry down was then used as a divisor to calculate the percent reduction in normalized ETA (%Max ETA; Eq. 5) for each day of the drydown until plant wilt. For each day, the corresponding cumulative ETA (ΣETA) was also calculated. The ΣETA at the time of shoot harvest was considered to be the total PAW (ΣETATotal). For each plant, a plot of %Max ETA as a function of ΣETA was developed (Fig. 1). Any additional ETA between midnight and shoot harvest on a day of harvest was not included.
Representative plots of cumulative ETA versus %MaxETA for five species of woody ornamental plants. Species were (A) Rhaphiolepis indica, (B) Elaeagnus pungens, (C) Ligustrum japonica, (D), Photinia × fraseri, and (E) Viburnum odoratissimum. The solid line was calculated from a linear regression based on five points and 18% minimum %MaxETA.
Citation: HortScience horts 42, 7; 10.21273/HORTSCI.42.7.1700
Representative plots of cumulative ETA versus %MaxETA for five species of woody ornamental plants. Species were (A) Rhaphiolepis indica, (B) Elaeagnus pungens, (C) Ligustrum japonica, (D), Photinia × fraseri, and (E) Viburnum odoratissimum. The solid line was calculated from a linear regression based on five points and 18% minimum %MaxETA.
Citation: HortScience horts 42, 7; 10.21273/HORTSCI.42.7.1700
Representative plots of cumulative ETA versus %MaxETA for five species of woody ornamental plants. Species were (A) Rhaphiolepis indica, (B) Elaeagnus pungens, (C) Ligustrum japonica, (D), Photinia × fraseri, and (E) Viburnum odoratissimum. The solid line was calculated from a linear regression based on five points and 18% minimum %MaxETA.
Citation: HortScience horts 42, 7; 10.21273/HORTSCI.42.7.1700
where i is the days after saturation of the root ball and the first day is 100% of maximum ETA.
Demonstration.
In Mar. 2005, rooted cuttings of V. odoratissimum were transplanted into 11-L containers using the same substrate formulation but supplied by a different commercial company (Florida Potting Soil, Apopka, FL). At times throughout the 13-month production period, plants were selected and placed in similar lysimeters under an open-side, single polyethylene-covered rainout shelter at the Mid-Florida Research and Education Center (Apopka, FL). After an initial saturating irrigation at midnight, 0100 hr and 0200 hr, water was withheld and plant mass recorded every half hour for 5 d or until visible midmorning shoot wilt. Thereafter, containers were hand-irrigated and canopy widest width, width perpendicular to widest width, and average shoot height were recorded. These were then multiplied to calculate a canopy volume [growth index (GI)]. Root ball volume was measured by first carefully removing the container substrate to a point of visible root tips. The remaining undisturbed root ball was then enclosed in a polyethylene bag and root ball volume determined by volume displacement in a 19-L bucket. The bucket was set at a slight angle so that water beaded always at the lowest point on the rim. The volume of water required to return the water surface to the original, maximum bead was determined. Effective PAW was determined as described subsequently for each replication and means were calculated.
For the statistical analysis, each plant was an experimental unit. Because of malfunctions, one replication each of E. pungens and L. japonica was discarded, resulting in all nonregression analyses being conducted using the general linear model (GLM) in SAS (version 9.0; SAS Institute, Cary, NC). GLM was used to compare PAWTotal as a split plot with species as the main plot and two determinations of PAW as a subplot. Similar comparison of LMwilt and sum of RBwilt + Shootwilt were also analyzed as a split plot design with species as the main plot and methods of determination as subplots. Regression equations were calculated using SAS (PROC REG) and used to analyze relationships between %Max ETA and daily ΣETA. Statistics were calculated for both linear and quadratic curves using daily ΣETA as an independent value. Values of total ΣETA were inserted into the resulting equations to estimate %Max ETA that occurred the day of plant wilt (%Max ETAwilt). Regression was also calculated using %Max ETA as an independent variable for the first four and five data points for each replication. GLM was then used to test the strength of four and five data point regressions to estimate total ΣETA measured at plant wilt using both calculated %Max ETA at the day of wilt and with %Max ETA fixed at 18%. The 18% was the mean minimum %MaxETA across all species in which total ΣETA measured at plant wilt was achieved. Analysis was conducted for each species independently and compared paired combinations of measured ΣETA against the three estimates of ΣETA. Analysis of other variables of Shootmax, Shootwilt, Shootdwt, RWCwilt, and %Max ETA mm −1 wilt were compared among species as a one-way analysis of variance using GLM.
Results and Discussion
Estimates of PAW were similar (P = 0.19) when calculated as the difference between LMsaturated and LMwilt (PAWTotal, Eq. 1) and total ΣETA determined by summing daily ETA (data not shown). The greatest difference between these two methods of calculation across all plants was 200 g or 6.01% of total ΣETA. Mean absolute difference among the 18 plants was 50 g or 1.45% of a system's PAWTotal. There was a slight tendency for total ΣETA to overestimate PAWTotal (11 of 19 plants).
Measurement of Shootwilt with an electronic balance was consistent with its calculation from LMwilt at harvest. There were no differences (P > 0.9) between LMwilt mass at harvest and the sum of Shootwilt and root ball mass at wilt (RBwilt). The mean absolute deviation was 15 g. There were differences (P < 0.01) in LMwilt among species. However, because basket and chain mass was not recorded for each lysimeter, comparisons of absolute mass could not be made.
Shoot dry weight was significantly different among species (Table 1). Because plants were market size for 11-L containers, woody stem mass contributed the majority of Shootdwt, especially for the larger canopy species. The species with the largest Shootdwt, L. japonica, P × fraseri, and V. odoratissimum, were all at least 0.75 m in height and had comparable shoot dry weight. These three species are considered to be upright spreading shrubs (Gaskalla, 1998). Rhaphiolepis indicia was the shortest species at ≈0.3 m with E. pungens height midway between the extremes.
Mean values for components of plant mass and relative water content during determination of plant-available water.z
Shootwilt mass varied among species, but relative rankings were similar to that for Shootdwt (Table 1). V. odoratissimum, L. japonica, and P × fraseri had much greater mass, nearly double that of E. pungens and R. indicia at canopy wilt. Similarly, these same three had nearly double the estimated shoot saturated mass (Table 1). Relative ranking by mass at wilt was different from for Shootdwt for the three largest species. Although not significant (P > 0.05), V. odoratissimum had the least Shootdwt but (P < 0.05) greater mass at wilt than P × fraseri (Table 1). Despite differences (P < 0.05) among species in both Shootdwt and Shootwilt, ratios of Shootdwt:Shootwilt was 0.44 ± 0.01 for all species except V. odoratissimum with a ratio of 0.34 (Table 1). This ratio was consistent across most species, although there were substantial differences in plant size and observational responses to water stress. Thus, when water content of a shoot of most plants evaluated declined to ≈2.25 times Shootdwt, a plant was visibly wilted. For V. odoratissimum, generally considered least tolerant of water stress in containers, visible wilt occurred with at a higher level of tissue water content. Mean RWCwilt of V. odoratissimum of 93.8% was higher than that of other species (Table 1). With the exception of L. japonica and E. pungens, RWCwilt was unique and different (P < 0.05) among all species, although Shootdwt:Shootwilt ratios were consistent. Water in the xylem likely contributed substantially to plant mass at wilt and is likely the reason for different RWCwilt among species. This difference in RWCwilt accounts for the higher Shootwilt for V. odoratissimum than P × fraseri relative to their Shootdwt. Importance and significance of the Shootdwt:Shootwilt ratio is unknown and may deserve further investigation.
RWCwilt was not predictable from other shoot measurements such as Shootdwt, Shootwilt, or ratios of these two. Of these species, V. odoratissimum is considered to have low drought tolerance in containers, whereas R. indica is considered to be very tolerant. The relative order of RWCwilt appears reasonable. Previously, Beeson (1992) suggested water potential thresholds for stomatal closure for L. japonica and E. pungens were similar with E. pungens being somewhat more drought-tolerant. However, the lower RWCwilt for P. × fraseri is opposite of what would have been expected based on previous reported stomata thresholds. These differences in RWCwilt suggest a possible quick way to screen landscape shrub species for drought tolerance.
Normalized values of ETA the first day of the drydown were always highest independent of plant species or replication. As days without irrigation progressed, ETA normalized by ETP declined, resulting in continuous declines in %MaxETA for at least the first 4 to 5 d of the drydown (Fig. 1). After ≈5 d, declines in %MaxETA slowed and became variable for some species (Fig. 1A–B). This variability became more conspicuous as duration of the drydown increased as exemplified by R. indica (Fig. 1A). However, for V. odoratissimum, canopy wilt was achieved within a 4- to 5-d timeframe (Fig. 1E). For all plants of the other species, 7 to 14 d was required until visible wilt was observed. This long duration resulted from both all plant tissue being mature and relatively low ETP that occurred during the drydown period. ETP ranged from 1 to 3.7 mm daily with a median value of 3.3 mm.
Regression analysis was applied to all data sets to calculate %MaxETA as a function of ΣETA to determine the best fit relationship between the two variables. Because there were no differences between ΣETATotal and PAWTotal, daily ΣETA is representative of cumulative daily water loss from a plant–substrate system over time. Some plots, particularly those of P. × fraseri, suggested a quadratic relationship between ΣETA and %MaxETA (Fig. 1D), thus, regression analysis was calculated for both linear and quadratic curves for each replication. Correlation coefficients (r2) were similar between linear and quadratic equations for most replications (data not shown). Values for r2 ranged from 0.849 to 0.976. Quadratic equations increased the r2 value by 5% to 11% for three of the four P. × fraseri (average r2 = 0.882 linear to 0.961 quadratic) and by 6% for one R. indica replication. Quadratic terms were not significant (P > 0.5) for V. odoratissimum.
Values of ΣETATotal were inserted into both linear and quadratic equations derived to compare %MaxETA estimated by these lines with %MaxETA derived from the last day of the drydown (%MaxETAwilt) for each replication. Comparing %MaxETAwilt with calculated values found little difference between values generated by linear or quadratic equation (data not shown). Linear equations tended to be more accurate in predicting %MaxETAwilt than quadratic equations (data not shown). Thus, for simplicity and accuracy, only linear equations were carried forward.
Having established a linear relationship between %MaxETA and ΣETA, the next step in establishing a method to estimate in situ PAWEffec was to define the relationship between PAWEffec and some level of %MaxETA. For all species except V. odoratissimum, a linear decline in %MaxETA with increasing ΣETA was followed by a pronounced slowing of that decline, either linearly or with oscillating variability (Fig. 1). The initial linear decline is proposed to be a direct result of gradual stomata closure resulting from decreases in available water within access of a root system. Gradual stomatal closure would reduce transpiration while drying of the substrate surface would reduce evaporation (Kramer and Boyer, 1995). The abrupt deviation in the linear decline is proposed to be the result of stomata closure. Thereafter, water lost would be mainly through cuticular transpiration and substrate evaporation. Because there is little to no photosynthesis with stomatal closure of most temperate woody plants, it is proposed that PAW at this point is essentially the PAWEffec. Although a plant may or may not exhibit wilt at this point, its net carbon balance would be near zero or less. In production situations in which the objective is growth, complete stomata closure is unacceptable. Thus, the minimum %MaxETA was sought where PAWEffec was achieved. Plots of %MaxETA versus ΣETA were visually examined for all plants to determine PAWEffec. These were then inserted into the linear equations derived and %MaxETA calculated. Mean minimum %MaxETA across all species was 18% (Fig. 1). Individual plant values ranged from 15.0% to 21.8%.
The final step was to determine the minimum number of days of a drydown required for accurate calculation of PAWEffec. Using 18% as %MaxETA and linear regression lines calculated from the first 4 and 5 d of a drydown, ΣETA was calculated for both regression equations for each plant. These values of estimated ΣETA were compared with actual measured ΣETA and ΣETA calculated with each plant's actual %MaxETA at PAWEffec. Overall, estimates from regressions based on 5 d of drydown were more accurate than those based on 4 d (data not shown). However, the gain in accuracy was generally small, from near 0% to 5%. On both a species basis and among species, there were no differences (P > 0.05) between ΣETA estimated from the first 5 d of the drydown using 18% as %MaxETA or estimated from the actual %MaxETA. Similarly, both these estimates of ΣETA were not different (P > 0.05) from measured values. Errors using 18% as %MaxETA were between near 0% and 6% with an average error of 2.5% compared with measured values.
Using a 4-d drydown, overall, there were no differences (P > 0.05) between estimated ΣETA using 18 as %MaxETA or actual %MaxETA and measured values. Estimates of ΣETA based on 18 as %MaxETA and a four-point curve had a mean error from the measured ΣETA of 3.9% and ranged from near 0% to 8.2%. Thus, for species evaluated, a 5-d drydown was sufficient for an accurate estimate of PAWEffec and mostly avoided pushing plants into a wilted state. However, with reference evapotranspiration (ET0) greater than 4.5 mm and plants two-thirds grown or better, where declines in PAW occur more quickly than in this experiment, a 4-d drydown would be acceptable and preferred.
To demonstrate changes in PAWEffec during a production cycle, data were extracted from an experiment using V. odoratissimum grown from rooted cuttings (30052202 Jiffy Strips; Jiffy Products of America, Norwalk, OH) in an 11-L container. A month after transplanting in the 11-L container, root exploration into the substrate was approximately one-fourth the total container volume. However, mean PAWEffec of 0.51 L was ≈17% of PAWTotal of a root ball at full container volume (Fig. 2). Roots generally had not reached the side walls at this point, but were growing in a wide cone shape downward. This illustrates the increasing gradient of the PAW of this substrate from top to bottom of the substrate column. Canopy volume was 1526 cm3, 0.4% of that of harvested plants (Fig. 2). Three months later, PAWEffec had almost doubled to ≈32% of PAWTotal with the root ball volume extending to approximately half the available container volume. Canopy volume had doubled to 3899 cm3, but still only ≈1% of the harvested canopy. Although canopies were visually developing at a good rate, root growth was much more rapid, expanding into virgin PAW. At 109 d after transplanting, root ball volume was roughly 25% larger than shoot volume in this substrate blend. Based on observations since 1990, root ball to canopy volume is more dependent on PAW than physical volume (unpublished data). Irrigation managers should take advantage of this greater increase in PAW than canopy volume, conserving irrigation for later in the production when daily ETA can reach one-third of PAWTotal (data not shown). Seven months after transplanting (day 232), PAWEffec had tripled to 3.1 L, although physical root ball volume had only increased ≈20% over the same period. At this point, root systems had not completely engulfed the substrate with the containers, but a few roots were at the bottom of the root ball. Canopy volume, however, increased 5.8-fold to 0.026 m3 over the same period. With the drydown at harvest 13 months after transplanting, PAWTotal was the same (P > 0.05) as that calculated at 7 months, although root ball volume had increased ≈30% to completely encompass the substrate volume. Correspondingly, canopy volume had increased nearly 12-fold to 0.34 m3.
Measured values of effective plant-available water (PAWEffec), root ball volume, and shoot volume (growth index) of Viburnum odoratissimum grown from rooted cuttings to market-sized plants. Plants were of market size and harvested at day 397. Means represent four to seven plant replications, depending on time after transplanting. Error bars represent the sd.
Citation: HortScience horts 42, 7; 10.21273/HORTSCI.42.7.1700
Measured values of effective plant-available water (PAWEffec), root ball volume, and shoot volume (growth index) of Viburnum odoratissimum grown from rooted cuttings to market-sized plants. Plants were of market size and harvested at day 397. Means represent four to seven plant replications, depending on time after transplanting. Error bars represent the sd.
Citation: HortScience horts 42, 7; 10.21273/HORTSCI.42.7.1700
Measured values of effective plant-available water (PAWEffec), root ball volume, and shoot volume (growth index) of Viburnum odoratissimum grown from rooted cuttings to market-sized plants. Plants were of market size and harvested at day 397. Means represent four to seven plant replications, depending on time after transplanting. Error bars represent the sd.
Citation: HortScience horts 42, 7; 10.21273/HORTSCI.42.7.1700
PAWTotal for V. odoratissimum measured at harvest was not as great as that measured in 1997 nor was it consistent with PAWTotal of the other species in similar containers (Table 2). There are several possible reasons for this. In 2006, plants at harvest were overhead-irrigated normally and the root balls were likely not fully saturated as they were with excessive microirrigation in 1997 or as they were in previous measurements in 2005. This is the most probable explanation because PAW at harvest was the same as that at 7 months, although root ball volume at 7 months was not at the maximum as it was at the harvest. Additionally, harvested plants in 2006 were smaller that those used in 1997 as evident by the smaller (P < 0.05) dry weight (272.4 g versus 499.1 g). This would have lowered the amount of PAW derived from woody stem tissue.
Total plant-available water (PAWTotal) based on cumulative ETA (∑ETA) for the five species from the 1997 experiment and for the V. odoratissimum taken from the 2005 to 2006 experiment.z
Through applying this procedure many times since its inception, there are two conditions that are critical to its successful use. One, the first day of a drydown must be one that has the best chance of having an average or higher reference ET0 for the region and season. Midday rain and prolonged cloudiness the first day, followed by a mostly sunny day resulting in higher ET0, can shift the day of maximum ETA to the second day, damaging the correlation of %MAXETA with ΣETA. This is more critical the less time a plant has been in a container with uncolonized substrate. Should low ET0 conditions occur after the first day, consequences are less severe. The other critical item, as was seen in the 2006 harvest results, is to thoroughly saturate a root ball initially to be able to determine an accurate volume of PAW.
Results suggest that determining RWCwilt could be used to rapidly screen plant species for drought tolerance. Lower RWCwilt suggests an indicative tolerance of a species to water stress, which could be verified by pressure-volume curve analysis (Tyree and Hammel, 1972). Evaluation of this hypothesis using established in-ground plants appears warranted. From a practical perspective, quantification of PAWEffec results in more accurate characterization of deficit irrigation on plants growth in containers. It can also be used to quantify root growth in containers nondestructively and repeatedly. In nursery production, PAWEffec could be determined every 4 to 6 weeks over a 5-d period with mass recorded before dawn each day. Measurement the fifth morning would provide the four minimum points to generate the linear regression curve with PAWEffec calculated using 18%MaxETA. With experience, irrigation managers could determine PAWEffec relative to container water capacity. Applying principles of deficit irrigation, irrigation would be frequent and light after transplanting, progressing rapidly to less frequent and greater volume as roots explore the substrate. This would result in less nutrient leaching and greater capture of rainfall. Irrigation in excess of PAW, especially for young crops, wastes water and energy and promotes excessive leaching of nutrients from a substrate. The main limitation with this method is the need for accurate calculation of daily ET0. Although some states have this available through the Internet, in states like Florida where cloud cover can be sporadic, a good onsite weather station would be preferred. These results from market-sized plants suggest PAWTotal is more an effect of container geometry and substrate composition and physical properties than plant species. Differences in root growth and growth direction would influence the rate at which PAWEffec increases with time.
Literature Cited
Beardsell, D.V. , Nichols, D.G. & Jones, D.L. 1979a Physical properties of nursery potting-mixtures Sci. Hort. 11 1 8
Beardsell, D.V. , Nichols, D.G. & Jones, D.L. 1979b Water relations of nursery potting-media Sci. Hort. 11 9 17
Beeson R.C. Jr 1992 Restricting overhead irrigation to dawn limits growth in container-grown woody ornamentals HortScience. 27 996 999
Beeson R.C. Jr 2006 Relationship of plant growth and actual evapotranspiration to irrigation frequency based on managed allowable deficits for container nursery stock J. Amer. Soc. Hort. Sci. 131 140 148
Beeson R.C. Jr & Yeager, T.H. 2003 Plant canopy affects sprinkler application efficiency of container-grown ornamentals HortScience 38 1373 1377
Bilderback, T.E. & Fonteno, W.C. 1987 Effects of container geometry and media physical properties on air and water volumes in containers J. Environ. Hort. 5 180 182
Bilderback, T.E. , Fonteno, W.C. & Johnson, D.R. 1982 Physical properties of media composed of peanut hulls, pine bark, and peatmoss and their effects on azalea growth J. Amer. Soc. Hort. Sci. 107 522 525
Bunt, A.C. 1976 Modern potting compost The Pennsylvania State Univ. Press University Park, PA
Davis, W.D. , Besemer, S.T. & Close, D.H. 1964 Landscaping in planters without natural drainage U.C. Agr. Ext. Serv. AXT 124 1 13
Jones, J.W. , Allen, L.H. , Shih, S.F. , Rogers, J.S. , Hammond, L.C. , Smajstrala, A.G. & Martsolf, J.D. 1984 Estimated and measured evapotranspiration for Florida climate, crops and soils Ag. Exp. Station, Instit. Food and Agric Sci. Univ. of Florida. Bulletin 840
Gaskalla, R. Director Division of Plant Industry 1998 Grades and standards for nursery plants, 2nd ed. Florida Dept. of Agr. and Consumer Services Tallahassee, FL. p. 237
Kramer, P.J. & Boyer, J.S. 1995 Water relations of plants and soils Academic Press, Inc San Diego. CA
Richards, D. , Lane, M. & Beardsell, D.V. 1986 The influence of particle-size distribution in pinebark:sand:brown coal potting mixes on water supply, aeration and plant growth Sci. Hort. 29 1 14
Spooner, L.A. 1974 Two classroom exercises demonstrating the pattern of container soil water distribution HortScience 9 152 153
Tyree, M.T. & Hammel, H.T. 1972 The measurement of turgor pressure and the water relations of plants by the pressure-bomb technique J. Expt. Bot. 23 267 282
Welsh, D.F. & Zajicek, J.M. 1993 A model for irrigation scheduling in container-grown nursery crops utilizing management allowed deficit (MAD) J. Environ. Hort. 11 115 118
Williams, D.J. 1978 Handling plants in containers J. Arbor. 4 184 186
