Studies on Water Transport through the Sweet Cherry Fruit Surface: XII. Variation in Cuticle Properties among Cultivars (2024)

Plant material

Sweet cherry fruit of the cultivars Alma, Bianca, Bing, Burlat, Early Rivers, F12/1, Flamingo Srim, Hedelfinger, Hudson, Katalin, Kordia, Margit, Nadino, Namare, Namosa, Oktavia, Rainier, Rebekka, Regina, Ria, Rube, Sam, Silvia, Stella, Summit, Sunburst, Techlovan, Van, Vernon, Viola, and Zeppelin were hand-picked from experimental orchards of the federal fruit cultivar office in Marquardt, Germany (lat. 52°28′ N, long. 12°58′ E). The cultivar Adriana was sampled from an experimental orchard at the Horticultural Research Center LVG, Erfurt, Germany (lat. 50°58′ N, long. 11°01′ E). All cultivars were grafted on P. avium ‘Alkavo’ rootstocks. Fruit was sampled randomly from three trees per cultivar and mean fruit mass was determined on a minimum of 15 fruit per cultivar and season.

CM mass, wax mass, and strain were quantified in ‘Adriana’, ‘Bianca’, ‘Burlat’, ‘Early Rivers’, ‘F12/1’, ‘Flamingo Srim’, ‘Hedelfinger’, ‘Hudson’, ‘Katalin’, ‘Kordia’, ‘Margit’, ‘Nadino’, ‘Namare’, ‘Namosa’, ‘Oktavia’, ‘Rebekka’, ‘Regina’, ‘Sam’, ‘Sunburst’, and ‘Van’ in 1999, 2000, 2001, 2003, and 2004 (only CM mass) at the mature Stage III.

In 2003, CM and wax deposition were quantified in Stage II and mature Stage III ‘Adriana’, ‘Alma’, ‘Bianca’, ‘Bing’, ‘Burlat’, ‘Early Rivers’, ‘F12/1’, ‘Flamingo Srim’, ‘Hedelfinger’, ‘Hudson’, ‘Katalin’, ‘Kordia’, ‘Margit’, ‘Nadino’, ‘Namare’, ‘Namosa’, ‘Oktavia’, ‘Rainier’, ‘Rebekka’, ‘Regina’, ‘Ria’, ‘Rube’, ‘Sam’, ‘Silvia’, ‘Stella’, ‘Summit’, ‘Sunburst’, ‘Techlovan’, ‘Van’, ‘Vernon’, ‘Viola’, and ‘Zeppelin.’ Mean fruit mass averaged 1.9 ± 0.1 g/fruit (range, 0.6 to 3.9 g) during Stage II and 8.3 ± 0.4 g/fruit (range, 1.4 to 12.0 g) during Stage III.

The permeabilities for transpiration and water uptake were determined in ‘Adriana’, ‘Bianca’, ‘Burlat’, ‘Early Rivers’, ‘F12/1’, ‘Flamingo Srim’, ‘Hedelfinger’, ‘Hudson’, ‘Katalin’, ‘Kordia’, ‘Margit’, ‘Nadino’, ‘Namare’, ‘Namosa’, ‘Oktavia’, ‘Rebekka’, ‘Regina’, ‘Sam’, ‘Sunburst’, and ‘Van’ in 1999, 2000, 2001, 2003, and 2004 (P_t, all years; P_f, 2003 and 2004 only) at the mature Stage III. The densities of stomata (all years) and microcracks (2000, 2001, 2003, 2004) were quantified. In 1999, only a simplified two-score rating scheme (with vs. without microcracks) was used.

Procedures

Strain of the cuticular membrane.

The biaxial elastic strain of the ES was quantified as described previously (Knoche and Peschel, 2006). Fully hydrated CM discs were spread on a microscope slide in a drop of water and photographed at ×4 (MZ6 microscope; Leica Mikrosysteme, Bensheim, Germany) using a video camera (HV-C20 E/K-S4; Hitachi Denshi Europa, Rodgau, Germany). The area of the CM disc was quantified by image analysis (analySIS 3.0 software package; Soft Imaging System, Münster, Germany) (Knoche et al., 2004). The biaxial elastic strain [

(percent)] was calculated according to Eq. [1], where

represents the surface area of the CM covered on the fruit before excision and

the area of the relaxed CM following isolation:

The A was calculated as the surface area of a cap of a sphere having a diameter equivalent to the cork borer and assuming a spherical shape of the fruit as a first approximation. The minimum number of replications was five.

Terminology, data analysis, and presentation

The permeabilities [P (m·s⁻¹)] in transpiration (P_t) and in osmotic water uptake (P_f) were calculated from the rate of water transport [F (kg·s⁻¹)] in transpiration (F_t) and in water uptake (F_f), the density of water [

(kg·m⁻³)], the area of the ES exposed in the diffusion cell (A = 38.5 × 10⁻⁶ m²), and the respective driving forces [

(dimensionless)] for transpiration and water uptake according to Eq. [2] (Beyer and Knoche, 2002; Beyer et al., 2005).

In transpiration,

equaled the gradient in water vapor activity between the inside of the diffusion cell (

) and the ambient atmosphere outside of the diffusion cell [

(for details see Beyer et al., 2005)]. In osmotic water uptake, the

was calculated using Eq. [3] where

represented the gradient in water potential (

) between the water donor solution on the surface of the ES and the PEG receiver solution inside the diffusion cell and

See Also

Eat Well & Share: Basil adds a lovely punch of flavor to many dishes

[137.3 MPa at 25 °C (Nobel, 1999)] the molar volume of water

divided by the product of the universal gas constant (R) and the absolute temperature [T (for details see Beyer et al., 2005)].

Rates of transpiration (F_t) and of osmotic water uptake (F_f) were determined on an individual ES basis by fitting a linear regression line through a plot of cumulative transpiration or water uptake vs. time. The slope of this regression line represented the respective F_t and F_f. Only observations on diffusion cells having an r² > 0.50 were used (transpiration mean r² = 0.992, n = 2009; uptake mean r² = 0.927, n = 564). The observations remaining (n = 14 to 25 per cultivar and year) represented 99.0% and 96.6% of the total population of ES for transpiration and water uptake, respectively.

In this equation, represented the variance component for the variance among genotypes, the variance component for the variance among years, the variance component for the variance associated with the interaction genotype × year, and the variance component for the residual variance that resulted from fruit-to-fruit variation and any variability associated with our experimental techniques. This procedure uses the variance associated with years as an estimate for the environmental variance component. To obtain normal distributions, densities of microcracks, and the permeabilities for transpiration and for osmotic water uptake were log-transformed. Linear regression and multiple stepwise regression analyses were carried out using Proc Reg and Proc Stepwise (SAS Version 9.1.3). Data are presented as means and ses. Where error bars are not shown, they were smaller than data symbols.

Mass of cuticular membrane and wax and strain

Among the scion cultivars, CM mass per area and hence thickness ranged from 0.85 g·m⁻² in ‘Rainier’ to 1.61 g·m⁻² in ‘Kordia’ (Table 1). Only the very small fruit from the ‘F12/1’ rootstock had a higher CM mass per unit surface area [2.17 g·m⁻² (Table 1)]. On a whole fruit basis, ‘Burlat’ had the lowest CM mass per fruit [1.68 mg/fruit (data not shown)] and ‘Kordia’ the highest (3.54 mg/fruit). Across cultivars, CM mass per fruit was positively related to fruit mass [r² = 0.34; P < 0.0001 (data not shown)], whereas CM mass per unit area was unaffected (without ‘F12/1’ fruit) or negatively related to fruit mass [with ‘F12/1’ fruit, r² = 0.08, P = 0.0004 (Fig. 1A)].

On average wax mass accounted for one-fourth of CM mass and ranged from 0.21 g·m⁻² in ‘Burlat’ to 0.42 g·m⁻² in ‘Zeppelin’ (Table 1). Fruit from the rootstock ‘F12/1’ had an even higher wax mass per unit area (0.47 g·m⁻²). Across cultivars, wax mass per unit area was positively and linearly related to CM mass per unit area [r² = 0.64, P < 0.0001 (Fig. 1B)].

On excision of the ES and isolation of the CM, the surface area of the CM decreased indicating the presence of significant elastic strain in the CMs of all cultivars. Across cultivars, strain averaged 76.7% and ranged from 56.6% in ‘Namosa’ to 97.0% in ‘Kordia’ (Table 1). Only the CM of the small fruit of the ‘F12/1’ rootstock was less strained [15.8% (Table 1)]. Across cultivars strain was a linear function of fruit mass [r² = 0.33, P < 0.0001 (Fig. 1C)]. Generally, thinner CMs were more strained [r² = 0.34, P < 0.0001 (Fig. 1D)]. Multiple linear regression analysis revealed that 66% of the strain of the CM at maturity (r² = 0.66, P < 0.0001) was accounted for by fruit surface area (partial r² = 0.51, P < 0.0001) and CM mass per unit surface area (partial r² = 0.15, P < 0.0016). There was no improvement of the model (r² = 0.65, P < 0.0001) when strain was related to the increase in surface area between Stages II and III (partial r² = 0.54, P < 0.0001) or to the amount of cuticle deposited between Stages II and III (partial r² = 0.11, P = 0.0059).

Partitioning total variance into variance components caused by the factors genotype, year, the interaction genotype × year, and a remaining residual error revealed that for fruit mass, CM mass, wax mass, and strain of the CM the largest variance component was always associated with the genotype (Table 2). For these traits, estimates of the broad-sense heritability were 40.6% or greater (Table 2).

Across all cultivars investigated, the CM and wax mass during Stage II were positively and linearly related to CM and wax mass at the mature Stage III, respectively, both on an area and a whole fruit basis (Fig. 2; Table 3). For CM and wax mass on a whole fruit basis, the slope terms were close to 1 indicating that little deposition of CM and wax occurred between Stage II and the mature Stage III.

Density of stomata, microcracks, and permeabilities for water transport

The density of stomata on the cheek of the scion cultivars ranged from 0.12 stomata/mm² in ‘Adriana’ to 2.13 stomata/mm² in ‘Namosa’ that was only exceeded by the rootstock ‘F12/1’ [3.59 stomata/mm² (Table 4)]. The heritability of stomatal density was high [67.5% (Table 2)].

Across cultivars and years, mean densities of short (2.31 microcracks/mm²) and long microcracks (0.79 microcracks/mm²) were of similar orders of magnitude as those of stomata (1.10 stomata/mm²), but ranges were markedly larger (from 0.18 to 7.09 and 0.02 to 4.23 microcracks/mm² for short and long microcracks, respectively). Also, the heritabilities of microcrack density were markedly lower than that for stomatal density (Table 2). Furthermore, the large variance component associated with the residual variability of microcrack densities indicated significant fruit-to-fruit variation within cultivars and years (Table 2).

Permeability for transpiration was lowest in ‘Flamingo Srim’ and highest in ‘Nadino’ and that for osmotic water uptake lowest in ‘Adriana’ and highest in ‘Hedelfinger’ (Table 4). Estimating variance components revealed that the variance components for the residual error exceeded those caused by the cultivars. Also, heritability estimates were low, more so for P_t than for P_f (Table 2). Multiple linear regression analysis revealed that the density of long microcracks accounted for 16% and 45% of the variability in P_t and P_f, respectively (Table 5; Fig. 3A–B). CM mass per unit area (partial r² = 0.069, P = 0.0227) further increased the variability in P_t accounted for to 23.3% (P = 0.0003). For P_f, introducing stomatal density (partial r² = 0.080, P = 0.0456) and the strain of the CM (partial r² = 0.045, P = 0.1174) increased the r² to 57.2% of the variability accounted for [P < 0.0001 (Table 5)]. P_t and P_f were significantly and positively related [r² = 0.771, P < 0.0001 (Fig. 3C)]. Both were positively related to stomatal density (Table 6).

Mass of cuticular membrane and wax, strain, and stomatal density

CM and wax mass differed significantly and consistently among cultivars. Across cultivars, CM mass and wax mass were positively related, which is consistent with common biosynthetic pathways for synthesis of fatty acid based precursors of cutin monomers and wax constituents (Peschel et al., 2007; Samuels et al., 2008). The overall deposition pattern of cutin and wax during fruit development was also similar across cultivars. Essentially all CM and wax deposition had occurred by Stage II, when CM and wax mass accounted for 81.0% ± 1.4% and 78.0% ± 1.9% of their respective masses at maturity. Thus, the increase in surface area between Stage II and the mature Stage III distributed an essentially constant mass of CM over an enlarging surface causing significant elastic and plastic strain (Knoche et al., 2004). The consequences of this mismatch of CM deposition and surface expansion were threefold. First, CM mass per unit area was significantly and negatively related to fruit mass [r² = 0.08, P = 0.0004 including fruit of the ‘F12/1’ rootstock; r² = 0.01, P = 0.3480 excluding ‘F12/1’ fruit (Fig. 1A)]. Second, elastic strain of the CM increased as fruit mass increased [r² = 0.33, P < 0.0001 and r² = 0.18, P < 0.0001 with and without ‘F12/1’ fruit (Fig. 1C)]. Third, strain of the CM was negatively related to CM mass per unit area [r² = 0.34, P < 0.0001 and r² = 0.22, P < 0.0001 with and without ‘F12/1’ fruit (Fig. 1D)].

Density of microcracks, permeabilities in transpiration, and water uptake

The traits density of microcracks and permeability for transpiration and that for water uptake shared the following common characteristics. For all three traits, the variance component caused by the residual variability was larger than that caused by the variance among cultivars. As a consequence, the heritability of these traits was low (range H = 6.2% to 30.3%). This is not surprising considering the mechanistic basis of microcracking of the CM and the role of surface defects including microcracks in water transport. Microcracks result from elastic strain of the CM (Peschel and Knoche, 2005) and from extended exposure of the strained CM to liquid water or high concentrations of water vapor (Knoche and Peschel, 2006). Fruit surface wetness duration is primarily affected by environmental conditions, which are highly variable even among fruit on the same tree. As a result, the residual variability of the density of microcracks is expected to be high. Furthermore, large microcracks were an important determinant of water permeability in transpiration and water uptake.

The observation that P_f exceeded P_t by a factor of 24.5 ± 2.5 is consistent with earlier observations (Beyer et al., 2005). The higher P_f is accounted for by viscous flow along polar pathways that in sweet cherry are associated with stomata and microcracks in the CM (Beyer et al., 2005; Weichert and Knoche, 2006). These pathways contributed to osmotic water uptake but to a lesser extent to diffusion, which is the primary mechanism of transport in transpiration (Beyer et al., 2005). This argument is also supported by the significant relationship between the ratio of P_f/P_t and the densities of stomata and of large microcracks. Furthermore, the dependence of P_f on stomatal density that has a high heritability (H = 67.5%) is likely to account for the higher heritability of P_f (H = 30.3%) as compared with P_t (H = 9.7%).

Relationships to rain-induced cracking

From a practical point of view it would be interesting to know to what extent the CM properties quantified herein account for differences in cracking susceptibility among cultivars. Such analysis is mathematically easy to perform using our data. However, several restrictions and limitations need to be recognized when interpreting the output of such calculations. First, we did not quantify the cracking susceptibility of those batches of fruit used in our experiments. Therefore, we can only relate the means of traits across different years to published data on cracking susceptibility and any information associated with the year-to-year variability of the traits investigated and, possibly, the cracking susceptibility is lost. Second, cracking susceptibility has been quantified using a variety of approaches ranging from cracking indices (CIs) quantified in standardized laboratory assays to rating procedures in the field (Christensen, 1996). Cracking indices are advantageous above rating scores based on field observations, because randomly occurring rain events under orchard conditions are replaced by the standardized immersion assay. To our knowledge, the most comprehensive source of CI is the compilation published by Christensen (1996). Correlation analysis revealed that these CIs are significantly correlated with rating scores of field cracking susceptibility performed by the federal fruit cultivar office, Wurzen, Germany (W. Pfannenstiel, unpublished data), but the coefficient of correlation was “only” r = 0.62 (P = 0.0043) for n = 19 cultivars (M. Knoche, unpublished data). Third, the CM properties that we focused on in our study relate to water transport across the fruit surface, which is an important but not the only factor in fruit cracking. A second important aspect is the mechanical constitution of the fruit and, in particular, that of the load-bearing structure, the exocarp. There is essentially no published information on this aspect and any factors associated with the mechanical properties of the exocarp are ignored in such multiple regression or correlation analyses. Finally, empirical correlations do not prove functional relationships and hence are of limited value for mechanistic studies. Nevertheless, when performing multiple regression analyses using published CI (Christensen, 1996) and mass of CM, density of stomata, of microcracks, and the data on P_t as independent variables, the highest r² (r² = 0.48, P < 0.0001, n = 64) were obtained using stomatal density (partial r² = 0.218, P < 0.0001) followed by densities of long (partial r² = 0.141, P = 0.0005) and short microcracks (partial r² = 0.056, P = 0.0192) and the CM mass per unit area (partial r² = 0.065, P = 0.0085) as independent variables. Similarly, performing the same analysis, but using P_f instead of P_t, identified densities of stomata (partial r² = 0.124, P = 0.034), of long microcracks (partial r² = 0.096, P = 0.047), and the CM mass per unit area (partial r² = 0.077, P = 0.0545) as independent variables that accounted for 52.5% (r² = 0.525, P < 0.0003, n = 29) of the variation of the CI. These analyses demonstrate that maintaining the barrier function of the CM by 1) preventing the formation of microcracks; and 2) maintaining a low stomatal density and a thick CM is helpful in developing less cracking-susceptible cultivars.

Table 6.

Parameter estimates and se of linear regression equations for the relationship between permeability (P, m·s⁻¹) in transpiration (P_t) or in osmotic water uptake (P_f) and stomata number on excised exocarp segments (sto) of 15 (P_f) to 19 (P_t) sweet cherry cultivars.^z