Effect of Night Temperature on Pollen Characteristics, Growth, and Fruit Set in Tomato

Lycopersicon esculentum Mill. ‘Laura’ plants were grown in the North Carolina State Univ. phytotron at 26C day temperature and 18, 22, 24, or 26C night temperatures to determine the effects of night temperature on pollen characteristics, growth, fruit set, and early fruit growth. Total and percentage normal pollen grains were higher in plants grown at night temperatures of 18 and 22C than at 24 and 26C, but germination was highest in pollen produced at 26C. Seed content was rated higher on the plants grown at 18C night temperatures than in any of the other treatments. Numbers of flowers and fruit on the first cluster were lower in the 26C night treatment than in the other night temperature treatments. Plant height was greatest but total shoot dry mass was lowest in the 22C night temperature treatments. Fruit fresh mass increased with night temperature, reflecting more rapid development, but the experiment was not continued to fruit maturity, so the effect of night temperature on final fruit size and total plant production could not be determined. Night temperatures of 26C reduced fruit number and percentage fruit set only slightly at a day temperature of 26C, even though these temperatures were above optimal for pollen production and seed formation. To separate temperature effects on pollen from direct or developmental effects on female reproductive structures, pollen was collected from plants in the four night temperature treatments and applied to stigmas of a male-sterile cultivar kept at 24–18C minimum temperatures in adjacent greenhouses. In the greenhouse-grown male sterile plants, no consistent effects of night temperature treatment given the pollen could be seen in fruit set, fruit mass, seed content (either on a rating or seed count basis), seedling germination, or seedling dry mass. High temperatures have been reported to limit tomato production wherever days are warmer than 32C, or nights are warmer than 21C (Moore and Thomas, 1952). Because of this high temperature sensitivity, poor fruit set is an important limitation to tomato production in the tropics (Villareal, 1980). Poor fruit set can also be a problem in the southern United States. In Louisiana, summer fruit set of six tomato genotypes ranged from 50% to 1% (Hanna and Hernandez, 1982). The importance of high temperatures in general, and high night temperatures in particular, as a production limitation for tomatoes and other crops may increase if current climatic trends continue. The National Oceanic and Atmospheric Administration reports that an examination of day and night temperatures at hundreds of climate stations in the United States, China, and areas included in the former Soviet Union, showed that average high temperature at night increased 0.7C during the last 40 years (Karl et al., 1991), but daytime temperatures were generally constant. They attributed virtually all of the warming in the northern hemisphere during the last four decades to increases in average high temperatures at night. The available literature on the sensitivity of tomatoes to high night temperatures is difficult to interpret because the temperaturedependence of fruit set has not been established, particularly in terms of separating night temperature effects from day temperature effects. Most studies are conducted at only two sets of conditions (high day/high night vs. low day/low night temperatures) or conducted in the field under constantly varying but largely undocumented conditions (e.g., summer vs. spring production in f publishing this paper was defrayed in part by the payment of page nder postal regulations, this paper therefore must be hereby marked ent solely to indicate this fact. raduate student. Hanna and Hernandez, 1982). In other cases, only day temperatures were varied (e.g., Levy et al., 1978). Thus, it is fairly clear that night temperatures as well as day temperatures affect tomato fruit set, but the cardinal temperatures have not been determined. Classical studies of Went and co-workers in the 1940s are often cited as the basis for the importance of night temperature effects in tomatoes. Based on studies in growth chambers (Went, 1945) and in air-conditioned greenhouses (Went, 1944; Went and Cosper, 1945) the critical factor in the setting of tomato fruit was reported to be night temperature, the optimum range being 15 to 20C (Went, 1944). There are several problems with these early studies, however, because of the inadequacy of the facilities available at the time. These included the necessity of moving plants to varying nighttime regimes after only 8 h, imposition of different nighttime treatments on plants kept in greenhouses during the day at different times of the year, mosaic disease, and the use of growth chambers with poor climate control, and low and spectrally unbalanced lighting. The use of an 8-h day was a particular problem in determining the importance of nighttime temperatures. Kinet et al., 1985, observed that studies where night temperature treatments were longer than day temperature treatments generally concluded that night temperatures were more important than day temperatures and vice versa. Studies by Schaible (1962) and Curme (1962) in air-conditioned greenhouses also showed the importance of night temperatures, but had limitations in terms of seasonally varying irradiance and 8-h days. Poor fruit set at high night temperature is not well understood in terms of the relative importance of these effects on male and female reproductive tissues. The most definitive work has been done on cowpeas by Hall and co-workers. Warrag and Hall (1984) attributed low podset in cowpeas grown at 33/30C compared to those grown at 33/24C, to reduced pollen shed and viability. Podset J. AMER. SOC. HORT. SCI. 121(3):514–519. 1996. Fig. 1. Field of pollen grains (100×) showing germinated (G), undersized (U), shriveled (S) and normal, but ungerminated (N). was increased in these plants when flowers grown at lower night temperatures were used as a pollen source. Mutters et al. (1989) associated temperature sensitivity with inhibition of proline accumulation in pollen grains and greater accumulation of proline in anther walls under high night temperatures. In tomatoes, there is little consensus on whether pollen or ovule effects are more important at high temperatures, and none of this work has targeted night temperature effects. Levy et al. (1978) used cross-pollination experiments on fertile plants to compare male and female effects. They reported that overall, with high daytime temperatures, (nighttime temperatures were the same), effects on male gametes seemed more important than those on female gametes. Reduction in pollen fertility at high temperatures have been noted by Dane et al., 1991 and a decline in pollen germination as germination temperatures increased from 20–27C by Charles and Harris (1972). It is not clear to what extent reduced pollen production at high temperatures will affect fruit set in tomatoes. Even though a relatively small percentage of pollen produced is required for fertilization, literature on pollen competition (Burd, 1994) suggests that reduced pollen quantity can depress fertilization and viability and vigor of seeds produced. Pollen limitation and pollen competition effects may also be variable with species, and are currently somewhat controversial in the literature (Caspar and Niesenbaum, 1993). In a study of pollen competition effects in apples and pears, Janse and Verhaegh (1993) found pollen density affected fruit set, seedset, and seedling vigor of an apple cross, but the effect was not consistent across the two cultivars and no effect on seedling vigor was seen in the pear cultivar used. In pear, a higher pollen density did increase the number of apparently viable seeds. However, Dane et al. (1991) concluded that more research was needed on the causes of the wide genotypic variation in heat tolerance and the underlying physiological processes. The objective of the present study was 1) to establish a night temperature dependence curve for fruit set in tomatoes and 2) separate male and female effects, at least on a preliminary basis. Materials and Methods Phytotron. Seeds of ‘Laura’, an indeterminate greenhouse tomato cultivar, (De Ruiter Seeds, Naaldwijk, the Netherlands), were germinated in mist beds starting 1 Sept. 1993. They were then transplanted on 14 Sept. into a 26/22C growth chamber with growing area of 3 × 2 m. On 30 Sept., 104 four-week-old ‘Laura’ seedlings were transplanted into 9-liter containers (25-cm diameter pots) using a 2:1 (v/v) mix of gravel and peat-lite, placed two per cart and assigned to four growth chamber treatments. Day temperatures (0700–1900 HR) in all chambers were 26C. Night temperatures (1900–0700 HR) were: 18C, 22C, 24C and 26C. Night temperature treatments began 1 Oct. 1993, corresponding to visible flowers in the first cluster. Chamber irradiance at the start of the experiment was about 650 mmol·s·m at the base of the plant, provided by a mixture of fluorescent (T-12, 1500-ma, cool-white) and incandescent lights (100 W). Lamps were replaced on a regular schedule, so the total decrease of chamber lighting attributable to degradation of light sources was no more than 12% of the initial intensity. Details of growth chamber design and operation, such as formulation of nutrient solution (half-strength Hoagland’s) and methods used to avoid CO2 drawdown in the chambers, are provided by Thomas and Downs (1991). Plants were trained to a single stem by removing sideshoots, supported on stakes placed in the pots and watered twice daily. Carts were rotated weekly in the chambers to reduce position effects. Fully opened flowers were J. AMER. SOC. HORT. SCI. 121(3):514–519. 1996. pollinated midday by vibration for 3 sec with a commercial electric vibrator. Pollen sampling and in vitro germination. On four dates (29 Oct., 2 Nov., 3 Nov., and 4 Nov. 1993), pollen samples were collected from open flowers in the first and second inflorescence of 10 plants in each of the four treatments. The pollen samples were collected midmorning (usually 1000 HR) by vibrating the inflorescence while holding a flower at anthesis in a 50-ml glass vial to collect pollen. Within 30 min of collection, 10 ml of germination solution (Abdul-Baki and Haynes, 1993) consisting of 20% lactose by mass, 50 ppm boron and pH adjusted with potassium hydroxide to 5.2 were added to the vial to begin the in vitro germination process. Samples were allowed to germinate for 2.5–3 h, then aliquots were examined at 100× using a light microscope. Pollen grains were classified into the following categories (Fig. 1) germinated (pollen tube longer than diameter of pollen grain); shriveled; underdeveloped or undersized (smaller, darker) grains; and normal in appearance (large, spherical, light-appearing) but ungerminated. Grains were counted by category in five regions on each of four slides per treatment for each of the four sampling dates. Harvest. At the conclusion of the experiment (4 Nov.), plant height, shoot dry matter, flower and fruit number, and fruit fresh mass were measured. All fruit were still green at this time. All clusters were harvested but only data from the first clusters were used in the statistical analyses and figures presented because not all flowers had reached anthesis in clusters 2 and above in the lower temperature treatments, preventing computation of fruit set. For each fruit, the number of seeds present was rated visually after cutting the fruit in half cross-wise. A fruit with empty spaces in place of gel and few seeds visible received a seediness index rating of 0, and a fruit with good gel development and many seeds visible received an index rating of 5. Pollen application to male-steriles. Pollen was collected as described previously from 10 plants in each treatment and applied with a small brush to cluster 1 (2–3 Nov.) and cluster 2 (10–11 Nov.) of male-sterile plants [NC8288 MS-1035, provided by R. Gardner, North Carolina State Univ. (NCSU)] kept in an NCSU Horticultural Science glasshouse (minimum temperature 24/18C). Two fertile plants of NC8288 (distinguished by purple stems and fully developed anthers) and two extra sterile plants were included in each experimental block as nonpollinated controls. Plants were arranged on greenhouse benches in four blocks,