Cross-pollination affects fruit colour, acidity, firmness and shelf life of self-compatible strawberry

Cross-pollination affects the fruit characteristics of many crops but the effects of cross-pollination on fruit quality of strawberry (Fragaria × ananassa Duch.) are poorly known. This study determined how cross-pollination affects fruit quality of the strawberry cultivar, Redlands Joy, under controlled environment conditions. Plants were allocated to one of four treatments, with all flowers on each plant receiving either: (1) unassisted self-pollination (Autogamy); (2) hand-pollination with Redlands Joy pollen (Self); (3) hand-pollination with cross-pollen from a small-fruited cultivar (Sugarbaby); or (4) hand-pollination with cross-pollen from a large-fruited cultivar (Rubygem). Cross-pollination did not significantly affect plant yield or fruit mass, size, shape, firmness or shelf life. However, cross-pollination affected fruit colour and taste attributes. Cross-pollinated fruit were 3%–5% darker than self-pollinated fruit. They also had 26%–34% lower acidity and 43%–58% higher Brix:acid ratio. Cross-pollination by Sugarbaby increased fruit P, K, Ca, Fe and Mn, but decreased B, Cu and Zn, concentrations. Cross-pollination by Rubygem increased fruit Mn, but decreased K and Na, concentrations and reduced shelf life. Fruit mass, length, diameter and firmness within all treatments increased with increasing numbers of fertilized seeds per fruit. Hand self-pollinated fruit had a higher percentage of fertilized seeds than fruit arising from autogamy and they were also darker, redder, firmer, and had a longer shelf life, higher protein concentration, and lower Al and Na concentrations. The results indicate that strawberry fruit quality can be affected by both the source of pollen and the number of stigmas pollinated.


Introduction
Pollination is vital to sustain food production for the growing human population because more than 75% of food crops require some form of pollination [1]. Furthermore, cross- cultivars arose from a strawberry breeding and selection program in Australia [40][41][42] that services an industry with approximately 2800 hectares of strawberry plants under cultivation. They have been widely grown commercially and are now available for planting by home gardeners. Fruit of these cultivars are red, medium-glossy, conical to round, of medium firmness, and flavoursome [43][44][45]. Fruit of Redlands Joy are considered attractive, easily harvested, resistant to rain damage, low-acid, medium-sweetness, and consistently well-liked by consumers [43]. Each runner was transplanted into a 4.5 L pot containing coco-peat (EC < 1 mS/cm, pH 5.5-7.0) and perlite (4:1, v:v) with 2.5 g of Osmocote fertilizer (N:P:K = 19.6:16.0:5.0% w:w, plus trace elements) (Scotts International, Heerlen, The Netherlands). The potted plants were then placed in an enclosed glasshouse at the University of the Sunshine Coast, Sippy Downs, Australia (26˚43'S 153˚03'E), with plant spacing of 20 cm from pot to pot [46]. Each pot was top-dressed monthly with 15 g of Osmocote fertilizer. A supplementary 5 mL of 1% (v/v) aqueous PowerFeed 1 foliar fertiliser (Seasol International, Bayswater, Australia) was applied weekly to each plant from transplanting until the commencement of hand-pollination treatments. Irrigation was applied manually to the plant canopy, avoiding direct contact with flowers during the pollination period. Temperatures in the glasshouse (Fig 1a) were recorded using a Tinytag Plus 2 data logger (Gemini Data Loggers, West Sussex, UK). Irradiance was monitored hourly on five cloudless days during the experimental period (Fig 1b) using a quantum sensor (LICOR LI-250A, Lincoln, NE). Pots were rotated frequently to minimise differences in long-term light interception among different sides of each plant. Runners and senesced leaves were removed regularly.

Experimental design and pollination method
Forty plants of cultivar Redlands Joy with uniform height and crown diameter were arranged in a completely randomized design with four treatments: (1) non-emasculated flowers received Each flower on plants in Treatment 1 (Autogamy) was labelled with a miniature paper tag that recorded the date on which the anthers first released pollen (Fig 2a). The flower was then covered with a fine-mesh bag, which was permeable for water vapour but prevented wind and cross-pollination, from first anther dehiscence until its petals started to fall. Each flower on plants in Treatments 2-4 ('Self', 'Sugarbaby' or 'Rubygem' pollen) had its anthers removed before the flower opened and was then covered with a fine-mesh bag until the stigma was receptive. Mature anthers of each cultivar were harvested daily and dried at glasshouse temperature (Fig 1a) from 1200 H to 1400 H in the lid of a 100-mL plastic jar. Released pollen was then transferred to the plastic jars, which were closed and stored overnight in a refrigerator at 4˚C. Flowers in Treatments 2-4 were hand-pollinated using a paintbrush, with separate brushes used for each pollen-donor cultivar. Each flower in Treatments 2-4 was labelled with a miniature paper tag that recorded the date of first hand-pollination and was covered again with a fine-mesh bag to completely exclude other pollen. Pollination was repeated daily until the petals started to fall.

Yield, fruit characteristics and shelf life
Each fruit was harvested at commercial maturity; i.e. when the skin was becoming fully red (Fig 2b) and was numbered consecutively. Time to maturity for each fruit was calculated as the number of days from first anther dehiscence (Treatment 1: 'Autogamy') or first hand-pollination (Treatments 2-4: 'Self', 'Sugarbaby' or 'Rubygem') to harvest. The fresh mass, length and diameter of each fruit, excluding the pedicel and sepals, were recorded. Length:diameter ratio was then calculated as a measure of shape for each fruit. Yield was calculated as the total fresh mass of all fruit harvested from a plant. Colour was assessed for each fruit using a CR-10 colorimeter (Konica Minolta, Chiyoda, Japan) in the L � a � b � colour space. L � , a � and b � values indicate brightness (brighter is more positive and darker more negative), redness (redder is more positive and greener more negative), and yellowness (yellower is more positive and bluer more negative), respectively. Firmness of each fruit, with the exception of the twentieth to twenty-fourth fruit harvested from each plant (i.e. fruits [20][21][22][23][24], was determined using a Bareiss Hpe-II fruit penetrometer (Bareiss Prüfgerätebau, Oberdischingen, Germany) with a 3-mm diameter probe. This was the smallest available probe and was recommended by the manufacturer for testing strawberry fruit. Maximum force was recorded when the penetrometer was pushed 6 mm vertically downwards into the fruit apex at a speed of 100 mm min -1 [23,47,48]. These fruits were then frozen at -20˚C prior to subsequent quality analyses. The twentieth to twenty-fourth fruit harvested from each plant were stored at 4˚C to determine shelf life. Shelf life was calculated as the number of days until the skin became commercially unattractive due to dark spot or mould development.
Ten frozen fruit from each plant (fruits 1, 6, 11, 16, 19, 26, 31, 36, 41 and 46) were used to measure seed number, TSS concentration and acidity. The total number of seeds (i.e. achenes) and the number of fertilized seeds were counted on each fruit. Unfertilized seeds were visibly distinguishable from fertilized seeds due to their smaller size (Fig 2c-2f). The percentage of seeds that were fertilized was calculated for each fruit. These fruits were then individually defrosted for 60 s (1000 W, 2450 MHz) in a Panasonic NN-CD987W microwave oven (Panasonic, Osaka, Japan) and the released juices were filtered through filter paper. TSS (˚Brix) and acid concentration of the filtered extract were determined using a PAL-BX|ACID4 sugar and acidity meter (Atago, Tokyo, Japan) [49][50][51]. Brix:acid ratio was then calculated for each fruit.

Data analysis
Data were analysed using IBM SPSS Statistics v. 26 (IBM, Armonk, NY). Yield data was analysed by 1-way analysis of variance (ANOVA). Fruit quality and shelf life data were analysed using generalized linear models (GLMs). Pollination treatment and plant (nested within pollination treatment) were regarded as main effects, and the date of harvest of each fruit was incorporated as a covariate. Due to highly significant effects of date of harvest, a series of GLMs was used to compare two pollination treatments in each model. Holm-Bonferroni corrections were applied to adjust the significance values for multiple models. Linear regressions were also performed to evaluate effects of the number of fertilized seeds, the percentage of fertilized seeds, the number of unfertilized seeds, or the percentage of unfertilized seeds per fruit as the independent variable and fruit quality parameters as the dependent variable. Treatment differences or regressions were regarded as significant at P < 0.05. Means are presented with standard errors.

Pollination effects on fruit yield, fruit quality and seed fertilization
Cross-pollination, when compared with self-pollination, did not significantly affect yield of Redlands Joy strawberry plants (Table 1). Cross-pollination also had no significant effect on individual fruit mass, length, diameter, length:diameter ratio, or time to maturity (Table 1). These parameters also did not differ significantly between self-pollinated fruit and fruit arising from autogamy (Table 1).

Relationships between fertilized seeds and fruit quality
The strongest linear relationships between seed parameters and fruit quality attributes were obtained using the number of fertilized seeds, rather than the number of unfertilized seeds or the percentages of fertilized or unfertilized seeds, as the independent variable. Therefore, only the relationships between number of fertilized seeds and fruit quality attributes are presented (Table 3). Fruit fresh mass (r 2 values of 0.353-0.615) (Fig 7), length (r 2 values of 0.064-0.491) and diameter (r 2 values of 0.113-0.596) were positively and consistently related to the number of fertilized seeds in all treatments (Table 3). Fruit firmness was also related positively to the

PLOS ONE
Effects of cross-pollination on strawberry fruit quality number of fertilized seeds, although the relationships were not as strong (r 2 values of 0.043-0.199) ( Table 3).

Discussion
This study has demonstrated that pollination can affect many of the characteristics of strawberry fruit. In particular, cross-pollination affected fruit colour and taste attributes. Cross-pollinated fruit could be more appealing to traders and consumers because they were darker, sweeter and had higher concentrations of some mineral nutrients. Autogamously-pollinated fruit had a lower percentage of fertilized seeds and were brighter, less red, less firm and had shorter shelf life than fruit arising from hand self-pollination or from hand pollination by at least one of the cross-pollen sources. The results highlight that both pollen source and pollen quantity can influence strawberry fruit quality. However, pollen source and pollen quantity tended to affect different aspects of fruit quality.
Cross-pollination of strawberry flowers did not significantly affect plant yield, fruit mass, fruit size, or the time taken to reach fruit maturity. These results demonstrate that the study cultivar, Redlands Joy, is self-compatible. However, the study results were among the first to show that the pollen source can affect the appearance and flavour attributes of strawberry fruit.

PLOS ONE
Effects of cross-pollination on strawberry fruit quality Fruit that arose from hand cross-pollination were darker and less acidic than fruit that arose from hand self-pollination or autogamy, but they had similar TSS concentrations and so were sweeter-tasting. Cross-pollination also increases the sweetness of mandarin (Citrus reticulata Blanco) and sweet cherry (Prunus avium L.) fruit [4,56] and cross-pollination by cultivars that have sweeter-tasting fruit increases the sweetness of cherimoya (Annona cherimola Mill.) fruit  Table 2. Mineral nutrient concentrations (mg/100g fresh mass) in Redlands Joy strawberry fruit arising from unassisted self-pollination (Autogamy) or hand-pollination with self-pollen (Self) or cross-pollen from another cultivar (Sugarbaby or Rubygem).

Nutrient
Pollination treatment  [57]. Consumers prefer sweet strawberry fruit [25], and so growers might increase the consumer appeal of their crop by planting multiple varieties in close proximity to each other and managing bee populations to maximise cross-pollen transfer. Honey bees tend to forage along rows rather than across rows on strawberry farms and so cross-pollination could be improved further by planting more than one cultivar within the same row [58]. Cultivar Redlands Joy was found to be self-compatible but it remains possible that other cultivars such as Sugarbaby and Rubygem have some degree of self-incompatibility. If that were the case, then enhanced cross-pollination might benefit their yield and fruit size in addition to affecting their appearance and flavour. Limited pollen deposition also resulted in lower fruit quality. Autogamy significantly decreased the percentage of fertilized seeds per fruit compared with the hand-pollination treatments. This indicates that autogamously-pollinated flowers experienced a pollination deficiency, with insufficient pollen being deposited on the stigmas to fertilize many of the ovules. Pollen deposition was not measured in the current study, but pollination experiments in open fields of strawberry cultivar Albion suggest that hand-pollination provides greater pollen deposition than bee-pollination, and that bee-pollination provides greater pollen deposition than autogamy [58]. Autogamous fruit of Redlands Joy were brighter, less red, less firm and had shorter shelf life than fruit arising from hand self-pollination or cross-pollination by Sugarbaby. Exclusion of insect pollinators from strawberry flowers has resulted previously in fruit with fewer fertilized seeds, higher proportions of fruit deformation, and shorter shelf life [20,21]. Together, these results suggest that autogamous pollination is associated with limited pollen deposition and that this can result in fruit with poor appearance and reduced shelf life, even in self-compatible cultivars like Redlands Joy. Red colour development in strawberry fruit is associated with the accumulation of anthocyanin pigments, which are beneficial for human health [59][60][61][62][63]. Increasing fruit firmness and shelf life play an important role in reducing postharvest losses of strawberry [64]. The current results are the first to show that pollination can change the mineral nutrient concentrations of strawberry fruit. Cross-pollination of Redlands Joy flowers by Sugarbaby increased fruit Ca, Fe, K, Mn and P concentrations while cross-pollination by Rubygem increased fruit Mn concentration. Cross-pollination by Sugarbaby decreased fruit B, Cu and Zn concentrations while cross-pollination by Rubygem decreased fruit K and Zn concentrations. In addition, hand-pollination with either self-or cross-pollen increased N concentrations, and decreased Al and Na concentrations, when compared with autogamous pollination. Nitrogen concentrations are correlated with crude protein concentrations in fruit, with protein concentration being approximately 6 × N concentration [65]. The 26%-32% higher N concentrations in well-pollinated fruit could help consumers to reach the recommended daily protein intake of 20-25 g [66]. However, strawberry fruit do not provide high amounts of dietary protein, with mean protein concentrations in the current study being less than about 1.4%; i.e. less than about 0.14 g per fruit. Interestingly, hand-pollinated strawberry fruit had 27%-56% lower Na concentrations than autogamously pollinated fruit, demonstrating that consumption of Table 3. Coefficients of determination for linear regressions between the number of fertilized seeds and the physical and chemical attributes of Redlands Joy strawberry fruit arising from unassisted self-pollination (Autogamy) or hand-pollination with self-pollen (Self) or cross-pollen from another cultivar (Sugarbaby or Rubygem). well-pollinated fruit could help to reduce salt intake. These results reinforce that cross-pollination affects many fruit characteristics, but also highlight that both the type and the amount of pollen deposited on the stigmas can influence fruit quality. Fruit mass, length, diameter and firmness were related consistently within all treatments to the number of fertilized seeds per fruit, although the relationship between firmness and number of fertilized seeds was weak. These aspects of fruit quality were not influenced significantly by the source of pollen. Limited pollen deposition can reduce the set of fertilized seeds, and this resulted in smaller and softer fruit. Strawberry fruit growth is regulated by the concentration of indole-3-acetic acid, which is produced by fertilized seeds on developing fruit [23]. Filled seeds may also provide structural support to the skin of the fruit [17]. The number of fertilized seeds can, therefore, affect the phenotype of the fruit. Low seed numbers have been related previously to reduced fruit mass or size, and higher frequencies of fruit deformation, in other strawberry cultivars [11,[19][20][21]23].

Dependent variable Treatment
Some components of fruit quality that were affected by pollen source, such as acidity, Brix: acid ratio, and manganese and zinc concentrations, were generally not related to the number of fertilized seeds per fruit. Putative paternity effects on fruit characteristics in multi-seeded fruit are sometimes difficult to distinguish from the potentially-confounding effects of the number of fertilized ovules and developing seeds [67]. However, the current results suggest that some cross-pollination effects on the flavour attributes and mineral nutrient concentrations of strawberry fruit were not attributable to the number of fertilized seeds but may have been regulated largely by the paternity of the seeds. These effects on fruit flavour attributes and mineral nutrient concentrations could, therefore, be true xenia effects [24].

Conclusion
Cross-pollination affected the appearance, flavour attributes and nutrient concentrations of fruit in a self-compatible strawberry variety. Increasing the percentage of fertilized seeds with either hand self-or cross-pollination, compared with autogamy, also increased fruit darkness, redness, protein concentrations, and, for two of the three pollen sources, firmness and shelf life. Improvements in the appearance, taste and dietary composition of fruit can affect marketability because traders and consumers are interested increasingly in healthy, high-quality food. Fresh mass and size of self-compatible strawberry fruit appeared to be determined largely by the number of fertilized seeds whereas other quality attributes such as flavour and some mineral nutrient concentrations appeared to be regulated by seed paternity, independently of the number of fertilized seeds. Strawberry growers need to consider interplanting different cultivars and managing insect pollinators such as bees on farms to ensure the best possible fruit quality.