Plant materials

Experiments were carried out in situ in field-grown plum trees (Prunus x domestica L.) of the cultivar ‘Hauszwetsche Wolff’ at the Horticultural Research Station of the Leibniz University in Ruthe, Germany (lat. 52°14’N, long. 9°49’E). Trees were grafted on ‘St Julien A’ rootstocks. Standard cultural practices for irrigation, fertilization and pruning were applied by our experienced staff. Measurements were made between 55 and 147 days after full bloom (DAFB). Fruit selected for measurement were of uniform developmental stage, as indexed by size and color, and were without visible defects.

Mass, color and osmotic potential

Fruit mass, color and the osmotic potential of the expressed juice were determined individually on 15–20 fruit throughout the season. Fruit mass was determined gravimetrically, fruit color using a spectrophotometer (CM-2600d; Konica Minolta, Tokyo, Japan) and juice osmotic potential using a water vapor pressure osmometer (VAPRO 5600; Wescor, Logan, UT). Before measuring color, the epicuticular waxes were removed by wiping with soft tissue paper.

Fruit volume

The fruit volume (V; mm³) was estimated based on the average of two orthogonal measurements of equatorial fruit diameter (d; mm) and one of fruit height (h; mm) using a digital caliper. Fruit shape was assumed a spheroid with the following formula: It was further assumed, that the measured (LVDT) change in d was accompanied by a proportional change in h, such that the ratio d to h remained constant during any 6-d measurement period (i.e. that fruit shape was conserved over this period). It was also assumed, that the density of the fruit is 1, such that the change of volume was given in μl or ml (1 mm³ = 1 μl).

Xylem, phloem and transpiration flows

Flows were determined as described previously [10]. The plum was positioned in a Lucite frame such that one side was in contact with the frame and the opposite side with an LVDT (SM277.15.2.STX24, resolution 5 μm; Schreiber Meßtechnik, Oberhaching, Germany). In this arrangement, the LVDT quantified changes in fruit equatorial diameter. Temperature and relative humidity (RH) were monitored within the canopy (CS215; Campbell Scientific, Shepshed, UK). Based on these data, the vapor pressure deficit (VPD) was calculated [16]. Data were logged at 5-min intervals (CR3000 data-logger, Campbell Scientific).

First, a developmental time course was established using 27 LVDTs mounted as above. Data were recorded for 72 h. After this time, each LVDT and its fruit was assigned to one or other of three groups of 6 to 9 fruit each (27 fruit in all). Fruit of the first group (C) served as un-treated controls. For a control fruit, the volume change equals the sum of xylem (X) in- and outflows, the phloem flow (P) and the transpiration (T). Note that T takes a negative sign (i.e. T flow is out of the fruit), also T can sometimes include a small positive component of osmotic water uptake through the skin into the fruit. The pedicels of fruit of the second group (G) were steam-girdled using a steam generator (SC952; Alfred Kärcher, Winnenden, Germany). Care was taken not to allow steam to damage the fruit itself. With steam-girdling, all cells in the steamed region of the pedicel are killed and phloem transport (a living process) is interrupted, whereas xylem transport (a non-living process) is unaffected [7,17]. Fruit from the third group (D) were detached by cutting the pedicel thereby interrupting X and P. The cut end of the pedicel was sealed using a fast-curing epoxy glue (UHU plus schnellfest; UHU, Bühl/Baden, Germany). The fruit remained in the same position in the canopy, such that the transpiration of the detached fruit (D) was considered representative of that of the still-attached fruits (C and G). It is worth noting that the skins of stage III stone fruit contain only very few and mostly inactive stomata [18], so their transpiration is unlikely to be affected significantly by treatments applied to their pedicels over the time periods of these manipulations. From the above discussion, it follows that the value of X is the difference in volume change between a steam-girdled fruit (G) and a detached fruit (D) (Eq 1), and also that the value of P is the difference in volume change between an un-treated control fruit (C) and a steam-girdled fruit (G) (Eq 2), whereas the detached fruit (D) represents the transpiration T (Eq 3).

(1)

(2)

(3)

Transpiration of the detached fruit decreased their diameters, so transpiration flows usually had a negative sign. Occasionally, transpiration was found to have a positive sign. This indicates surface water uptake via the skin must have occurred during dew/rainfall. The diameter changes of the fruit of the three groups (C, G, D) were recorded for a further 72 h period and the corresponding changes in fruit volume were calculated using treatment means and the above equation. An estimate of the standard error (SE) of the differences G–D (SE_X) was then obtained using the following equation (Eq 4). In this equation, the SE_G and SE_D represent the SEs for the means of G and D [19]: (4)

The SE_P was calculated accordingly using the SE_C and SE_G. Occasionally, a replicate was lost due to slippage in the Lucite frame caused by rain or wind or by insect damage. An exception was in late stage III of fruit development, where only four fruit remained in the D group. Transpiration flows were further analyzed by calculating the flux density by dividing the transpiration flow through the skin, by the fruit surface area. To normalize for differences in the transpiration driving force, flux density was then divided by the VPD (estimated from the temperature and RH values measured in the canopy). The normalized flux density is referred to as the skin permeance.

Data analysis

The 72 h period after treatment by steam girdling or by detachment was divided into three 24 h intervals and the X, P and T flows calculated for anyone of the three intervals. Calculations were based on the means of the four to eight fruits in the groups C, G and T. These data were used to calculate the correlations and time courses depicted in Figs 4–6. For Fig 7, the means of the three 24 h intervals were averaged. For data on single replicates the reader is referred to S1 Fig.

Data in Figs 1, 4, 5, 6A and 7 are presented as means ± SE. When no error bar is visible, the SE is smaller than the plotting symbol.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1.

Fruit mass, expressed fruit juice osmotic potential (A) and fruit skin color (B) of developing European plum. Fruit skin color is indexed as a change in Hue angle. The vertical dashed line indicates the transition between stage II and stage III. Time (x-axis) is in days after full bloom (DAFB). Data represent means ± SE.

https://doi.org/10.1371/journal.pone.0252085.g001

Data were analyzed by correlation and regression analyses (R 3.6.1; R Foundation for Statistical Computing, Vienna, Austria). Significances of the coefficients of determination (r²) at the 0.05 and 0.01 probability levels are indicated by * and **, respectively.

Xylem, phloem and transpiration flows in developing fruit

The X flow in European plum decreased during stage III. This conclusion is based on the following observations: 1) The relative contributions of X flow to total flow decreased markedly during development. 2) There was a close relationship between atmospheric VPD (the driving force for transpiration) and X flow during stage II (xylem functional), but not during stage III (xylem largely dysfunctional). 3) The diurnal oscillations of X flow decreased as development proceeded. Oscillations in X flow were essentially absent during late stage III. Our observations are consistent with a loss of functionality of the fruit xylem during the development of European plum. This is in line with similar results for other fleshy fruitcrop species, e.g. apple [20,21], grape [22–29], mango [30], kiwifruit [31,32], sweet cherry [10,33,34] and tomato [35,36]. Only in peach does the xylem seem to remain functional into maturity [37]. Two factors have been identified as contributing to decreases in xylem functionality: 1) a physical disruption of the xylem due to growth strain [21,33,38,39] and/or 2) a blockage of the xylem in the pedicel [29].

Phloem flows were low during stage II, but higher during stage III. This change in flow is accounted for by a high demand for carbohydrates during stage III, when fruit mass increases rapidly [40,41]. During stage II, the pit develops with little change in fruit mass [42,43]. The demand for carbohydrates is low [40,41].

The decrease in fruit transpiration during development was somewhat unexpected. Based on the obvious increase in fruit volume and fruit surface area during stage III, along with a decrease in cuticle mass per unit area and an increase in microcracking [44], we would have expected an increase in fruit transpiration, particularly when normalized by dividing by the increasing fruit surface area (to give a transpiration flux density) and then by dividing by the vagaries of atmospheric VPD (to give a skin permeance). This decrease in skin permeance is not unique for European plum, but was also reported in Japanese plums [11]. These findings are contrary to that for sweet cherry fruit, where skin permeance increases towards maturity, passes through a peak and then decreases [10]. Furthermore, transpiration in plum is closely related to the atmospheric VPD during stage II, but it is almost independent of VPD during stage III. The reason for this is that fruit apoplast water vapor pressure is not well coupled to atmospheric VPD. A decrease in fruit transpiration during development can be accounted for by a decrease in permeance of the fruit cuticle. Similar changes have been reported for developing apricots [45], grape berries [46,47], kiwifruit [48] and nectarines [49]. A decrease in fruit skin permeance must over-compensate any increases in cuticular microcracking that may have occurred. Earlier studies established that wax mass increased markedly during development in European plum [44] and that it is the wax within the cuticle that is responsible for its barrier function [50], whereas accumulation of epicuticular wax does not generally offer a significant barrier to the transpiration of fruit [51] or of leaves [52]. This conclusion also holds for European plum, where mechanical removal of the epicuticular wax bloom had essentially no effect on transpiration rate [53].

Xylem backflow and its potential role in neck shrivel

During stage II and occasionally during late stage III, we observed negative X flow i.e. a ‘backflow’ from fruit to tree. This backflow usually occurred in the morning hours, just after sunrise. The prerequisites for a X backflow are: 1) at least some xylem functionality must remain and 2) the driving force (a pressure gradient in the xylem) must be in the ‘reverse’ direction (i.e. so as to pull water from the fruit to the tree). Our results reveal that during stage III about 25% of the total water inflow to a plum is in the xylem. This indicates some persistent xylem functionality. The second factor, a ‘reverse’ pressure gradient, most likely results from high foliar transpiration following sunrise that lowers tree water status to the point at which the apoplastic pressure in the tree falls below the value established in the fruit apoplast during the previous hours of darkness. To help explain this, note first, that leaf transpiration (many open stomata) greatly exceeds fruit transpiration (few, non-functional stomata for example see [18]). Second, that the hydraulic capacitance of a leaf is low (small volume and a relatively rigid structure), whereas the hydraulic capacitance of a ‘soft’ fruit is much higher (a very much larger volume and a relatively flexible structure). Third, fruit are often surrounded and occluded by leaves. This will generate a higher-humidity and a lower-light environment around the fruit, so decreasing the driving force for fruit transpiration relative to leaf transpiration. Under these circumstances a backflow from fruit to tree is likely to occur. This scenario also accounts for the time lag observed between the T flow and the increase in VPD. Similar observations were reported for Japanese plum [11]. Later in the day, the rate of transpiration of the fruit increases. The increased rate of transpiration and–possibly—the fruit osmotic potential act in the opposite direction, pulling apoplastic water into the fruit [34]. The xylem backflow thus reverts to an inflow from the tree into the fruit.

These findings demonstrate that xylem backflow can indeed occur in the pedicel of a European plum during the early hours after sunrise. However, this backflow is essentially limited to stage II of fruit development and occurs only occasionally during stage III, the stage when neck shrivel appears in the field. Furthermore, the net movement of water into the fruit (i.e. X+P+T) over any 24-hour period, was always positive. If xylem backflow were to be a factor in neck shrivel in European plum, the xylem backflow in that 24-hour period would have to exceed the X+P inflows. This never was the case. We suggest this makes xylem backflow a highly unlikely contributor to neck shrivel in European plum. Research should focus on other potential causal factors.