2023 Volume 92 Issue 4 Pages 384-392
The salak fruit (snake fruit) contains one to three seeds covered with an aril. The size of the fruit primarily depends on the number of seeds. Fruits with more seeds grow larger and attain higher commercial value in fresh markets owing to their superior appearance. In eastern Thailand, during the hot period in the early rainy season, the fruit set of salak decreases, which is empirically believed to be caused by high night temperatures. In this study, we pollinated and incubated salak spadices at various temperatures (18–36°C) to determine the optimal post-pollination temperature range needed to produce valuable fruits. Chronological pollen-tube elongation in the pistil and development of the early embryo and endosperm were then observed anatomically, followed by fruit-set estimation. At ≤ 21°C, pollen-tube elongation was limited. At 24–27°C, although elongation was slow, pollen tubes attained embryo sacs in > 60% of florets within 36 h after pollination. Pollen tubes elongated fastest at 30–33°C and attained embryo sacs within 12 h after pollination. At 36°C, the difference in the elongation speed tended to be higher among the pistils. The percentages of ovules with developing embryos were the highest at 27°C at 48 h after pollination; zygote and early embryos were observed in 10.8% and 55.9% of the ovules, respectively, and the primary endosperm nucleus, dispersing endosperm nucleoplasm, and free endosperm nuclei were observed in 8.6%, 38.7%, and 19.4%, respectively. The second highest percentage was observed at 24°C. The percentages tended to decrease at 30°C and decreased significantly at ≥ 33°C. At ≥ 30°C, shriveled embryo sacs were observed. The estimated fruit-set percentage based on embryo development as the consequent fruit set was the highest (≈80%) at 27°C, while the second highest percentage (≈75%) was at 24°C. At ≥ 30°C, the estimated percentage decreased to less than half that at 27°C. Fruits containing three seeds were expected to grow in the range of 17.6–28.0% at ≤ 27°C. One- or two-seeded fruits were expected to grow at ≥ 30°C. Limited fruit set was expected at 36°C. Our results indicate that salak prefers relatively cool temperatures of approximately 25°C for the fertilization and set of valuable fruits.
Harvesting and marketing strategies significantly influence the profitability of salak (Salacca wallichiana C. Mart.) production; thus, production technologies aimed at postharvest handling were examined. Notably, artificial pollination (Matsuda et al., 2020a, 2021a; Piyarom et al., 2000) is a marketing-related technology in salak cultivation. Salak fruits bear in clusters; each fruit contains one to three seeds and the edible portions are the sweet, sour, and juicy arils that cover each seed. The greater the number of seeds in the fruit, the larger the fruit. Owing to their superior appearance and large edible portion, three-seeded fruits tend to be used for fresh consumption, and have a high market value. Small fruits with fewer seeds are often used for processed products, such as syrup, resulting in low unit prices and profitability. Based on postharvest marketing strategies, growers prefer to produce large fruits with three seeds.
In eastern Thailand, a major production area for salak, also called rakam locally, the plant flowers year-round. Notably, an increasing number of growers are performing artificial pollination to achieve stable production. However, artificial pollination at the beginning of the rainy season (May), reduces fruit set. Therefore, ‘Sumalee’ salak has an offseason from January to February because the relevant pollination period is around May, and the night temperature in May is the highest of the year. This suggests that high night temperatures during the pollination period may cause low fruit set at the beginning of the rainy season.
Even so, the effects of high temperatures on salak pistils remain unclear, although the germination percentage of salak pollen on agar media has been reported to decrease at temperatures above 36°C (Matsuda et al., 2020c). Environmental stress, such as high temperatures, can exhaust pistils. Since pollen tubes elongate heterotrophically in the pistil of angiosperms (Cruzan, 1986; Herrero and Arbeloa, 1989), pistil receptivity was reduced, and pollen-tube elongation after pollination was inhibited by high temperatures in fruit trees such as avocado (Sedgley and Annells, 1981), mango (Sukhvibul et al., 2000), cherimoya (Matsuda and Higuchi, 2012), and prune (DeCeault and Polito, 2010). During the hot rainy season, when the fruit set of salak decreases, daily maximum temperatures often reach approximately 35°C, and nighttime temperatures often remain at approximately 30°C. Under these conditions, pollen germination and subsequent pollen-tube elongation, fertilization, and embryo development may be inhibited.
Controlling the air temperature surrounding detached flowers indoors is easier than controlling the temperatures surrounding flowers on trees in a field. The thermal response of fruit tree reproductive organs was observed by incubating detached pollinated flowers at a constant temperature regime for several days: when pollinated peach flowers were incubated in the range of 10–30°C for 5–10 days, pollen tubes in the pistil elongated rapidly and pistil receptivity decreased rapidly at high temperatures (Hedhly et al., 2005); when pollinated sweet cherry flowers were incubated in the range of 5–30°C for 3 days, pollen-tube elongation in the pistil was inhibited at 5°C and 30°C (Beppu and Kataoka, 1999); and when pollinated mango flowers were incubated for 2 days in the range of 10–30°C, pollen-tube elongation was inhibited above 30°C and below 10°C (Sukhvibul et al., 2000). One day after pollination, fertilized salak embryo sacs were observed in an orchard (Matsuda et al., 2021b). Therefore, the effects of temperature on fertilization should be investigated by incubating detached pollinated spadices under various temperature conditions for 2 days to observe the morphological development of the early embryo and endosperm in salak.
In this study, to understand the effect of temperature on pollen-tube elongation and subsequent fertilization in the pistil after salak pollination, we incubated pollinated spadices at constant temperature conditions ranging from 18°C to 36°C with increments of 3°C, and pollen-tube elongation in the pistil was anatomically observed chronologically. In addition, to investigate the effect of high temperature after pollination on fertilization, pollinated spadices were incubated for 48 h at constant temperature conditions ranging from 24°C to 36°C, which simulated air temperature conditions during the hot rainy season. Next, the incubated pistils were sectioned, and the morphology of ovules was anatomically observed to estimate fruit set.
Salak trees of a female cultivar, ‘Sumalee’, cultivated in a private commercial orchard in eastern Thailand were used for the following experiments.
Experiment 1: The effect of temperature on pollen-tube elongation in pistilsPistillate spadices of ‘Sumalee’ salak on the day of flower opening were detached from the peduncle and immediately the peduncles were inserted into vials filled with water in November 2013. These spadices were pollinated using pollen from the spineless male ‘Sakam’ strain. The ‘Sakam’ pollen used for pollination was collected on the day of pollination from staminate spadices immediately after anther dehiscence, using the method described by Matsuda et al. (2020a). After 24-h incubation at 30°C on 1% agar medium containing 10% sucrose, 100 ppm boric acid, 300 ppm Ca(NO3)2·4H2O, 100 ppm KNO3, and 200 ppm MgSO4·7H2O, the pollen germination percentage was ≈70% according to the germination test described by Matsuda et al. (2020c).
Immediately after pollination, detached spadices were placed in a styrene foam box equipped with a Peltier effect temperature control device (cf. Matsuda et al., 2015) to regulate the internal temperature in a range of 18–36°C (± 0.2°C accuracy). Each box was opened 3, 6, 9, 12, 18 and 24 h after pollination, and 10–20 florets from each spadix were sampled and fixed in Copenhagen solution (ethanol: glycerol: distilled water = 10:1:8, v/v/v; Bridson and Forman, 1989). Although pollen tubes penetrating into the embryo sacs were observed 1 d after pollination for ‘Sumalee’ grown in the orchard condition (Matsuda et al., 2021b), florets incubated at temperatures below 24°C were collected and fixed after 36 and 48 h because low temperatures were expected to delay pollen-tube elongation. To maintain satisfactory water uptake from the peduncle, we removed the peduncle of each spadix from the water, cut it several centimeters from the tip to renew the cut surface 24 h after pollination, and immediately reinserted it into the water.
Petals and trichomes were removed from fixed florets, dehydrated with alcohol, and embedded in paraffin wax using the method described by Matsuda et al. (2011). The pistils were longitudinally sectioned into pieces ≈8 μm thick and stained with 0.1% aniline blue solution containing 0.1 M K3PO4. Pollen tubes in the pistils were observed under a fluorescence microscope (CyScope, Partec GmbH, Saarbrüchen, Germany). The pistil was visually divided into seven portions, i.e. 1) stigma, 2) upper style, 3) basal style, 4) upper ovary, 5) basal ovary, 6) micropyle, and 7) embryo sac (Fig. 1.A), and the deepest attained position of the pollen tube was recorded for each floret. Since each floret contained three ovules, the presence or absence of embryo sac malformation and the penetration of pollen tubes into the embryo sac were recorded for each ovule.
Fluorescence microscopic observations of pollen-tube elongation into a ‘Sumalee’ salak pistil. The pathway of the pollen tube (pt) is divided into seven portions (A): (1) pollen grains (pg) germinating on the stigma (stg); (2) pollen tubes penetrating the upper part of the style (stl); (3) penetrating the lower part of the style; (4) pollen tubes arriving at the upper part of the ovary (ovr) and then elongating down the locule (lc) along the ovule (ovl); (5) arriving at the basal part of the ovary; (6) pollen tube elongating into the micropyle (mp); (7) pollen tube penetrating one side of the synergids (sy) and reaching the embryo sac (es). (A) longitudinal section of a pistil; (B) transverse section of an ovary; (C) a normal ovule with embryo sac containing egg apparatus (ea) and antipodal cells (ac); (D) ovule with a degenerated embryo sac (des). Scale bars = 500 μm in A and B, and 100 μm in C, D, and 1–7. Image A was prepared by connecting five different pictures of the same section.
To determine the high temperature effect on embryo and endosperm development after pollination, the incubation of spadices in a range of 24–36°C in Experiment 1 was continued 48 h after pollination. After 48-h incubation, the boxes were opened, spadices were removed, and approximately 30 florets were collected for each temperature condition. These floret samples were immediately fixed in Copenhagen solution.
To observe the pistil morphology, the petals and trichomes were removed from the fixed florets, and the pistil was removed, rinsed in water, dehydrated in an alcohol series, and embedded in paraffin wax. After embedding, the pistils were sectioned into pieces 8–10 μm thick and stained with 0.05% w/v toluidine blue in a solution of 0.1 mol·L−1 citric acid and 0.2 mol·L−1 disodium hydrogen phosphate mixture (12:8, v/v) (pH 4.1). The stained sections were mounted to create permanent preparations by covering them with a cover glass, and the morphology of the ovules was observed under an optical microscope.
For each observed ovule, the morphology of the embryo sac was classified into two categories: degenerated (Fig. 2.1) or recognizable. For embryo sacs that were recognizable in the ovule, the developmental stage of the organs in the embryo sac was recorded separately for the embryo and endosperm. Embryo development was divided into five categories: 1) malformed embryo sac before fertilization (Fig. 2.2), 2) embryo sac containing an egg nucleus before fertilization (Fig. 2.3a), 3) zygote (Fig. 2.4), 4) early embryo (Fig. 2.5a–c), and 5) shriveled (Fig. 2.9b). Matsuda et al. (2021b) reported that after fertilization, salak endosperm first formed a large mononuclear primary endosperm nucleus, which was then dispersed in the embryo sac, followed by the development of numerous free endosperm nuclei as the endosperm nucleoplasm fused with cell groups on the embryo sac wall surface. Accordingly, endosperm development was classified into six categories: 1) malformed embryo sac before fertilization (Fig. 2.2), 2) embryo sac containing a polar nucleus before fertilization (Fig. 2.3b), 3) primary endosperm nucleus (Fig. 2.6), 4) dispersing nucleoplasm (Fig. 2.7a–c), 5) free endosperm nuclei (Fig. 2.8a and b), and 6) shriveled (Fig. 2.9a).
Microscopic observations of ‘Sumalee’ salak ovules. 1: ovule with a degenerated embryo sac; 2: an example of an ovule with a malformed embryo sac, for which the egg apparatus was formed not only on the micropylar end, but also on the side walls; 3a: egg nucleus (en) before fertilization; 3b: polar nucleus (pn) before fertilization; 4: zygote (zy) on the micropylar end; 5a: early embryo containing several nuclei (arrow heads); 5b: developing embryo with several cells. Inside cell walls were not clearly observed, and individual cells were not identified; 5c: Suspensor-like structure (sls) formed by one synergid, while the other synergid was degenerated (dsy); 6: embryo sac with a primary endosperm nucleus (pen); 7a: primary endosperm nucleus divided into several nuclei (arrow heads), and the nuclear membrane has disappeared; 7b: nucleoplasm dispersing toward the embryo-sac wall; 7c: nucleoplasm accumulated along the embryo-sac wall, and several cells on the wall surface opened; 8a: nucleoplasm and nucleus in opened cell on wall surface integrated into a free nucleus (fn); 8b: free nuclei (arrows) approaching the micropylar end; 9a: ovule with embryo sac shriveled from the chalazal end observed at temperature conditions higher than 30°C. Early embryo still developing; 9b: the entire embryo sac was shriveled. Scale bars = 50 μm.
In the evaluation of the high temperature effect after pollination on fruit set, when the embryo and endosperm nuclei (i.e. the early embryo and the dispersing endosperm nucleoplasm or free endosperm nuclei) were observed in the embryo sac, the ovule was considered to be fertilized. Such fertilized ovules were considered to develop a seed thereafter. Since a floret can succeed in fruit set when at least one of the three ovules is fertilized, the expected fruit-set percentage was calculated for each temperature condition. In the evaluation of the effect of post-pollination temperature on fruit size, a major factor for profitability, the estimated percentage of each fruit containing one, two, or three seeds was calculated for each temperature condition.
The observation results of pollen-tube elongation in pistils are shown in Figure 1. Germinated pollen grains were observed in the stigmas (Fig. 1.1). Pollen tubes penetrated through the upper part (Fig. 1.2) and then the basal part (Fig. 1.3) of the style to reach the locule (Fig. 1.4). The pollen tube elongated through the obturator in the locule to reach the basal part of the locule (Fig. 1.5) and then elongated through the micropyle (Fig. 1.6) to reach the embryo sac and penetrate one of the synergids (Fig. 1.7). In the three embryo sacs (Fig. 1.B) of the ovary on the day of pollination, normal ovules with the egg apparatus on the micropylar side of the embryo sac and antipodal cells on the chalazal side (Fig. 1.C) were observed, whereas ovules with a degenerated embryo sac (Fig. 1.D) were observed at an incidence of 14.4–29.2% (Table 1). The percentage did not change with incubation duration or temperature.
Incidence of an aborted embryo sac among observed ovules sampled after post-pollination temperature treatments within 24 h.
Pollen-tube elongation was retarded at low post-pollination temperatures (≤ 21°C), and pollen tubes did not reach the embryo sacs within 24 h after pollination (Fig. 3). Thereafter, the pollen tubes exhibited poor elongation; for less than 30% of the florets, the pollen tubes reached the embryo sacs within 2 days. No florets had pollen tubes reaching the embryo sacs in two or three ovules among the three observed ovules. At 24°C, the pollen tubes reached the embryo sacs in ≈30% of the florets within 24 h after pollination; however, 36 h after pollination, the pollen tubes reached the embryo sacs in more than 60% of the florets. In ≈20% of the florets, the pollen tubes reached the embryo sacs in two or three ovules.
Pollen-tube elongation in pistils of ‘Sumalee’ salak incubated at various temperatures (18–36°C) within 48 h after pollination. Pistils incubated at 27–36°C were observed within 24 h after pollination. Each circle represents an separate ovary. Open and closed circles at the position of the embryo sac indicate pollen-tube penetration into the embryo sac of one ovule and 2–3 ovules, respectively. Values in parentheses represent the percentage of florets with an embryo sac penetrated by pollen tubes on an hourly basis.
After pollination at 27°C, the speed of pollen-tube elongation increased, and pollen tubes reached the embryo sacs within 18 h. Pollen tubes reached the embryo sacs in more than 60% of the florets 18 h after pollination. Approximately 20% of the florets had pollen tubes that reached the embryo sacs in two or three ovules, similar to the results at 24°C. At 30°C, the pollen tubes elongated into the pistil the fastest, and they were observed to reach the embryo sacs within 9 h after pollination. Within 12 h of pollination, ≈40% of the florets had pollen tubes that reached the embryo sacs. Eighteen hours after pollination, ≈60% of the florets had pollen tubes that reached the embryo sacs, but less than 20% of the florets had pollen tubes that reached the embryo sacs in two or three ovules. The second-fastest elongation was observed at 33°C; in ≈60% of these florets, pollen tubes were observed to reach the embryo sacs after 18 h, and less than 20% had pollen tubes that reached the embryo sacs in two or three ovules. At 36°C, ≈60% of florets had pollen tubes that reached the embryo sacs 18 h after pollination the same as at 30–33°C, while in the remaining ≈40% of the florets, the elongation was slightly slower. Pollen tubes reached the embryo sacs in two or three ovules in 20–30% of the florets. The difference in elongation speed was the smallest at 27°C and tended to increase at higher temperatures.
Experiment 2: High-temperature inhibition of embryo and endosperm developmentIrrespective of the temperature conditions after pollination, 22.2–36.6% of the ovules were observed to have degenerated embryo sacs (Table 2). Even 48 h after pollination, less than 10% of the ovules were present before fertilization at temperatures ≥ 27°C, and the percentages were higher at lower temperatures than those at higher temperatures. At 24°C, ≈20% of the ovules were present before fertilization with some malformed (11.6%) and some containing an egg nucleus (8.7%). The percentage of ovules in which the zygotes and early embryos were completely developed after pollination was the highest at 27°C; zygotes and early embryos were found in 10.8% and 55.9% of these ovules, respectively. The percentage of ovules in which primary endosperm nuclei and developing endosperm were observed was also the highest at 27°C, and primary endosperm nuclei, dispersing endosperm nucleoplasm, and free endosperm nuclei were found in 8.6%, 38.7%, and 19.4% of these ovules, respectively. The percentage of ovules with developing embryos and endosperm nuclei was the second highest at 24°C, although fewer ovules developed into free endosperm nuclei at 24°C than at 27°C.
Effect of post-pollination temperature conditions within 24–36°C for 48 h on the development of egg apparatus and endosperm in ‘Sumalee’ salak.
In contrast, the percentage of ovules with developing embryos and endosperms tended to decrease at 30°C. At post-pollination temperatures ≥ 33°C, the percentage of ovules with developing embryos and endosperms decreased significantly. At 33°C, the percentage of ovules with early embryos was 28.9%, which was approximately half of that at 27°C. At 36°C, early embryos were observed in 11.8% of the ovules. At 33°C, endosperm nuclear division was observed in ≈10% of the samples, and at 36°C, endosperm nuclear division was rarely observed. At temperatures ≥ 30°C, shriveled embryo sacs were observed in more than half of the ovules, and the percentage of ovules with shriveled embryo sacs tended to increase with an increase in temperature. Shriveling was observed more frequently in the endosperm than in the egg apparatus.
The highest percentage of fruit set (80.6%) was estimated at 27°C (Table 3). The second-highest fruit-set percentage was ≥ 70% at 24°C. At ≥ 30°C, fruit set decreased significantly and was less than half of that at 27°C. A low percentage of fruit set was expected at 36°C. At higher temperatures (30–33°C), no fruits contained three seeds, 50% contained one seed, and 50% contained two seeds. At 36°C, 100% of the fruit was estimated to contain one seed.
Estimation of the fruit set and fruit type of ‘Sumalee’ salak obtained following post-pollination temperature conditions within 24–36°C for 48 h.
Salak pollen germinates well in a temperature range of 27–33°C (Matsuda et al., 2020c). In this study, the pollen tubes elongated the fastest in the pistil and reached the embryo sacs in 9–12 h at slightly higher post-pollination temperatures (30–33°C), but reached the embryo sacs optimally at 24–27°C (Fig. 3). Pollen tubes reached the embryo sacs at a high frequency 48 h after pollination at temperatures ≥ 24°C. Embryo development was significantly inhibited at temperatures ≥ 30°C (Table 2). Accordingly, our results suggest that temperatures ≤ 27°C, lower than the optimal temperature for pollen germination and fast pollen-tube elongation, are critical to achieve fertilization and begin embryo development in salak for subsequent fruit set.
Salak florets began to open in the late morning and were pollinated at around noon. Within the next 18 h, pollen tubes reached the embryo sacs (Fig. 3), and post-fertilization ovules were observed 1 day after pollination (Matsuda et al., 2021b). We found that salak requires low temperatures of 24–27°C for subsequent fertilization and embryo development, although it can tolerate high daytime temperatures of approximately 30°C during pollination.
During the hot period in the rainy season, when poor fruit set of salak is observed in eastern Thailand, midday temperatures often rise to approximately 35°C, and nighttime temperatures often remain at approximately 30°C. Two continuous days with a high temperature of ≥ 30°C after pollination significantly inhibited salak embryo development (Table 2), resulting in a decrease in the number of seeds per fruit and a significant reduction in potential yield (Table 3). According to our observations (Fig. 3 and Table 2), the optimal post-pollination night temperature for salak ranges from 24–27°C. A long duration of temperatures ≥ 30°C is therefore considered an important cause of poor fruit set in the rainy season.
Similar to salak, temperate and subtropical fruit species require lower temperatures for fruit set than the optimal temperature for pollen germination and fast pollen-tube elongation. Pollen germination and pollen-tube elongation in sweet cherry were vigorous in a range of 15–25°C (Beppu and Kataoka, 1999), but fruit set increased at daytime temperatures ranging from 10–20°C, with the lowest fruit set at 25°C (Beppu et al., 1997). In peach, pollen germination and pollen-tube elongation were better at 30°C than at 10°C, but pistil receptivity was rapidly lost at high temperatures (Hedhly et al., 2005), and fruit set was better at temperatures below 20°C than above 20°C (Kozai et al., 2004). In avocado, pollen tubes grew faster at post-pollination day/night temperatures of 33/28°C, but pistil receptivity was lost early, and fruit set was better at 25/20°C (Sedgley and Annells, 1981). In cherimoya, the highest pollen germination and fastest pollen-tube elongation occurred in the range of 22–25°C (Yonemoto et al., 1999) and 20–27°C (Matsuda et al., 2011), respectively, and night temperatures ranging from 17–22°C were preferred for reliable fertilization and fruit set (Matsuda et al., 2015). Similarly, in passion fruit, the optimal pollen germination temperature ranged from 32–36°C (Matsuda and Ogata, 2020), and temperatures ranging from 24–32°C were optimal for fertilization and fruit set (Matsuda and Higuchi, 2020; Matsuda et al., 2020b). In many fruit species, pistils and post-fertilization embryo development seem to prefer lower temperatures than male organs, as well as for pollen germination. High temperatures, even if they promote pollen germination and pollen-tube elongation, can be detrimental to fertilization because the pistil is exhausted quickly and tends to lose receptivity (Hedhly et al., 2005, 2007; Snider and Oosterhuis, 2011).
Matsuda et al. (2021b) observed the pistil morphogenesis of ‘Sumalee’ salak and reported that the embryo sac degenerated in ≈30% of the ovules 3 weeks before flowering, and this percentage was constant until flowering. Although the percentage observed in this study was slightly lower than that observed in their report, we observed similarly degenerated embryo sacs (Tables 1 and 2), and ≈20% of the florets had pollen tubes that reached the embryo sacs in 2–3 ovules, even in a range of 24–33°C (Fig. 3), in which pollen-tube elongation was highly promoted. The larger the number of seeds, the larger the fruit, which benefits the postharvest marketing channel and increases marketability. The number of seeds per fruit in salak (S. wallichiana) ranged from one to three (Lim, 2012a). In contrast, the number of seeds per fruit in the closely related S. zalacca, also an important tropical fruit species in Southeast Asia, is stable at three (Lim, 2012b). The number of seeds per fruit of S. wallichiana tends to vary widely due to cultivation factors such as insufficient pollination and high temperature inhibition, and a high number of ovules with degenerated embryo sacs at the time of flowering has been observed.
Matsuda et al. (2021b) observed that in salak, 10–15% of ovules had embryo sacs but malformed egg apparatus on the day of flowering. At 48 h after pollination, under temperature conditions ≥ 30°C, there were few ovules with malformed egg apparatus (0–3%), and a high percentage of shriveled embryo sacs was observed (Table 2). Malformed egg apparatus was considered to be present in ≈10% of the ovules with shriveled embryo sacs.
ConclusionOur results indicate that high temperatures (≥ 30°C) negatively affect fertilization and embryo development after pollination of salak (S. wallichiana) cultivated in tropical lowland areas. Pollination when nighttime temperatures drop ≤ 25°C is recommended to produce salak with high market value with fruits containing three seeds.
The Authors are very grateful to the Sukjit Orchard and Chanthaburi Horticultural Research Center for their kind support and advice on the field experiment. We would like to thank Dr. Tatsushi Ogata of JIRCAS for providing hand-made temperature control devices.
