HORTSCIENCE 27(8):885-887. 1992. Seedling age and duration of high-tern- perature exposure (Expt. 1). Seeds were sown on 9 Jan. 1989, at Bradenton. Seedlings 0, Rosetting of Lisianthus Cultivars 10, 17, 24, or 31 days following sowing were grown in 28 ± 1C soil for 0, 7, 14, 21, or Exposed to High Temperature 28 days. Only seedlings from the control (0 days from sowing) were exposed to the high Brent K. Harbaugh temperature during germination since seeds Gulf Coast Research and Education Center, University of Florida, 5007 germinated in 5 to 10 days. The plants were 60th Street East, Bradenton, FL 34203 grown in growth chambers during the high temperature exposure and cool-white flu- Mark S. Roh and Roger H. Lawson orescent and incandescent bulbs provided a U. S. Department of Agriculture, Agricultural Research Service, Florist photosynthetic photon flux (PPF) of 80 µmol·m-2·s-1 for 12 h. Air temperature was and Nursery Crops Laboratory, Beltsville, MD 20705 maintained in the range of ±2C of the soil temperature. Germination and seedling de- Brent Pemberton velopment occurred in a controlled environ- Texas A&M Agriculture Research and Extension Center, Over-ton, ment room when the plants were not being 7X 75684 subjected to high temperature. The control room was maintained at 22 to 24C with a Additional index words. Eustoma grandiflorum, flowering PPF of 30 µmol·m-2·s-1 for 12 h from cool- Abstract. Three lisianthus [Eustoma grandiflorum (Raf.) Shinn.] cultivars 0, 10, 17, white fluorescent lamps. When the last high- 24, or 31 days from sowing were grown in 28C soil for 0, 7, 14, 21, or 28 days to temperature treatment was completed, all determine the effects of high temperature during seedling growth on the development plants were transplanted into 0.5-liter (10- of rosetted plants. Increasing the duration of high-temperature exposure increased the cm diameter) pots and grown in a fan-and- percentage of rosetted plants for all cultivars. Such exposure for 28 days resulted in pad-cooled glasshouse where the air ranged 96%, 93%, and 18% rosetted plants for cultivars Yodel White, Yodel Pink, and GCREC- from 33/15C (day/night). Shading was pro- Blue, respectively. Seedling age did not affect percentage of flowering ‘Yodel Pink’ vided by exterior paint, and the midday PPF plants, but as seedling age increased to 31 days, the percentage of flowering plants ranged from 600 to 800 µmol·m-2·s-1. increased with ‘GCREC-Blue’ and decreased for ‘Yodel White’. In a second experi- ‘Yodel Pink’, ‘Yodel White’ (Sakata Seed ment, four lisianthus cultivars were grown at 22C for 3 weeks and then exposed for Co., Yokohama, Japan), and ‘GCREC-Blue’ 28 days to soil at 22, 25, 28, or 31C. Increasing soil temperature resulted in more (a heat tolerant, semidwarf, blue selection rosetted plants for all cultivars. With soil at 31C, 83%, 58%, 19%, and 2% of the made at the Gulf Coast Research and Edu- seedlings rosetted for the cultivars USDA-Pink, Yodel White, Little Belle Blue, and cation Center, Bradenton) were studied in GCREC-Blue, respectively. this experiment. There were three plants per experimental unit and four replications. The percentage of plants in each experimental unit One factor limiting lisianthus production 39°'N). Depending on cultivar, we observed that rosetted (basal cluster of leaves with no is the high percentage of plants that rosette that from 30% to 85% of the plants rosetted visible stem elongation), semirosetted (basal and fail to flower within an acceptable crop- for the June seeding date in all three loca- cluster of leaves but with stem elongation or ping period (»140 days). Lisianthus is de- tions. Fewer than 5% of the February or Sep- flowering from nonterminal shoots), or scribed as an annual or biennial plant (Bailey tember sown plants rosetted (unpublished flowered (open flowers from the terminal and Baily, 1976). Rosette plants have a basal data). Average greenhouse temperatures ex- shoot and the absence of a basal cluster of cluster of leaves and very short internodes, ceeded 28C at the three locations during pro- leaves) was recorded 172 days from seeding. similar to the first season’s growth of many duction of plants from June sowings. These Data were analyzed following arcsine trans- biennials. Many rosetted seedlings have been and other observations indicated that high formation for data based on percentages (Lit- observed in native stands of lisianthus in temperatures during seedling production may tle and Hills, 1975). Grimes County, Texas (lat. 30°'N), during cause or contribute to rosetting in some lis- Temperature regression (Expt. 2). Seed- August and September when day tempera- ianthus cultivars. lings (21 days from sowing) of cultivars Yodel tures frequently exceeded 35C. Lisianthus Information on the effects of temperature White; Little Belle Blue [released from U.S. cultivars seeded May through October that on lisianthus growth and development are Dept. of Agriculture (USDA)-Beltsville]; form rosettes typically will not flower until sparse. Halevy and Kofranek (1984) re- USDA-Pink (a heat sensitive, semidwarf, pink April or May of the following year (Har- ported that plants flowered 11 to 23 days selection made at the USDA-Beltsville); and baugh, 1988). One or more side shoots of earlier at 26 to 30/18C (day/night) than plants GCREC-Blue lisianthus were grown in 22, rosetted plants may elongate and flower. grown at 20 to 24/13C (day/night). How- 25, 28, or 31 ± 1C soil for 28 days. En- These semirosetted plants flower unpredict- ever, temperature treatments were initiated vironmental conditions during germination ably and are of poor quality as cut flowers 110 days from sowing. Several reports have and seedling production were the same as in and potted plants. indicated that temperatures > 21C during Expt. 1, except lights were on for 18 h. After As part of a national cooperative research seedling development will reduce quality of the temperature treatments, seedlings were project to study flowering in lisianthus, seeds flowering plants (Roh and Lawson, 1987; transplanted into 3.5 × 3.5 × 6-cm pyram- were sown in February, June, and September Roh et al., 1989). Data on the effect of tem- idal cell transplant flats (35 ml soil per cell). in Bradenton, Fla. (lat. 27.5°'N); Overton, perature on rosetting were not presented. Continued development occurred under Texas (lat. 32°'N); and Beltsville, Md. (lat. Ohkawa et al. (1991) reported that air tem- greenhouse conditions as in Expt. 1, except peratures >25C induced rosetting in ‘Fu- daylength was extended from twilight to 0200 -2 -1 Received for publication 31 Oct. 1991. Accepted kushihai’ lisianthus seedlings. High- HR with 2 µmol·m ·s of incandescent il- for publication 11 Mar. 1992. Florida Agr. Expt. temperature treatments for >14 days com- lumination. Sta. J. Ser. no. R-01971. Appreciation is&tended pletely inhibited bolting. The research re- The experiment was terminated 110 days to Sakata Seed Co.. Yokohama. Japan. for sup- plying lisianthus seeds. The cost bf publishing this ported here was conducted to determine the from sowing, » 1 month after the first visible paper was defrayed in part by the payment of page relationship between high temperature ex- flower buds were observed in all cultivars. charges. Under postal regulations, this paper posure during seedling development and the Data on the size of seedlings at transplant- therefore must be hereby marked advertisement incidence of rosetting in several lisianthus ing, days to visible flower buds, and per- solely to indicate this fact. cultivars. centage of plants that were rosetted, semi-

HORTSCIENCE, VOL. 27(8), AUGUST 1992 885 Table 1. Percentage of rosetted, semirosetted, and flowering lisianthus ‘Yodel Pink’, ‘Yodel White’, and ‘GCREC-Blue’ cultivars as a result of exposing 0 to 31-day-old seedlings (from sowing) to 0, 7, 14, 21, or 28 days of 28C soil temperature.

zDuration × plant age interactions were not significant. The main-effect values represent the means of 20 observations for duration treatments (four replications averaged over five seedling age treatments) and plant age (four replications averaged over five duration treatments of high temperature). NS,*,**Nonsignificant or significant at P = 0.05 or 0.01, respectively.

Table 2. Growth responsesz of four lisianthus cultivars grown at 22C for 21 days from sowing and ing the duration of 28C exposure increased then grown in 22, 25, 28, or 31C soils for 28 days. the percentage of rosetted plants for all cul- tivars (Table 1). However, the relative per- centage of rosetted plants resulting from high temperature exposure was different for each cultivar. ‘Yodel White’ had 96% rosetted plants after seedlings were exposed to high temperature for 28 days, while ‘Yodel Pink had 93% and ‘GCREC-Blue’ had 18%. The percentage of semirosetted ‘Yodel Pink’ plants increased as the duration of high temperature exposure increased from 0 to 14 days. In ‘Yodel White’, the percentage of semiroset- ted plants increased with increased high- temperature exposure from 0 to 21 days, and in ‘GCREC-Blue’, the percentage or semi- rosetted plants increased with increased high temperature exposure from 0 to 28 days. As the duration of high temperature exposure increased, the percentage of normal flower- ing plants decreased for all cultivars due to the combined increase in rosetted and se- mirosetted plants. Seedling age at time of high-temperature exposure did not affect the percentage of flowering ‘Yodel Pink’ plants (Table 1). In- creasing the seedling age at the time of high temperature exposure from 0 to 31 days in- creased the percentage of flowering plants for ‘GCREC-Blue’ but decreased the per- centage for ‘Yodel White’. Experiment 2. For all cultivars except ‘GCREC-Blue’, the percentage of rosetted plants was higher the higher the temperature, ZValues for seedling width, leaves, and days to visible bud represent the mean of 64 observations (four from 22 to 31C (Table 2). As in Expt. 1, replications of 16-plant experimental units). Percentage values represent the mean of four replications ‘GCREC-Blue’ appeared to be less sensitive for percentage of 16-plant experimental units which were rosette, semirosette, or flowering. y to high temperatures, with 97% flowering Data recorded after temperature treatments, 64 days from seeding. NS,*,**,Nonsignificant or significant at P = 0.05 or 0.01, respectively. plants at 31C. ‘USDA-Pink’ was the most sensitive cultivar, with 34% rosetted plants even at 22C. ‘Little Belle Blue’ had < 5% rosetted, or had visible buds by day (110) two replications were sown 16 Jan. and for rosetted or semirosetted plants at £ 28C, but were recorded. The arcsine transformation the two remaining replications 18 Feb. 1990. 31C resulted in 19% rosetted and 25% se- was used for regression analyses of percent- Experiment. 1. Interactive effects between mirosetted plants. ‘Yodel White’ had similar age data. There were 16 plants per experi- duration of high temperature exposure and percentages of rosetted (20%) and semiro- mental unit and four replications in time for plant age were not significant, so only main setted (24%) plants after exposure to 28C. temperature treatments. Seeds for the first effects are presented and discussed. Increas- In general for all cultivars, as temperature

886 increased to 28C, seedling width and number cultivars were the most sensitive in both ex- Harbaugh, B.K. 1988. Scheduling lisianthus of leaves increased (Table 2). The mean periments, with 34% rosetted plants for (Eustoma grandiflorum) production. Gulf Coast number of days to visible bud increased as ‘USDA-Pink’ at the lowest temperature of Res. and Educ. Ctr. Res. Rpt. BRA1988-25. temperature increased to 31C. These results 22C. The cultivar GCREC-Blue was se- Little, T.M. and F.J. Hills. 1975. Statistical meth- indicated that higher temperatures resulted in lected for heat tolerance from a tissue cul- ods in agricultural research. Univ. of California larger transplants. However, the production tured line of ‘Blue Poppy’ at the Gulf Coast Press, Davis. of larger transplants would be of little value Research and Education Center. The low Ohkawa, K., A. Kano, K. Kanematsu, and M. commercially since many of the large trans- tendency to form rosetted plants by ‘GCREC- Korenaga. 1991. Effects of air temperature and plants did not flower on schedule. Blue’ at high temperature exposure (2% ro- time on rosette formation in seedlings of Eus- In summary, the results of both experi- toma grandiflorum (Raf.) Shinn. Scientia Hort. setted plants with 31C soil for 28 days) in- 48:171-176. ments support the hypothesis that high tem- dicates that breeding efforts can reduce the perature exposure of lisianthus seedlings high temperature sensitivity in lisianthus. Roh, M.S., A.H. Halevy, and H.F. Wins. 1989. Eustoma grandiflorum, p, 322-327. In: A.H. causes development of rosetted and semiro- Halevy (ed.). Handbook of flowering. vol. VI. setted plants. The percentage of rosetted plants Literature Cited Crit. Rev. Plant Sci. Nutr. Press, Boca Raton, increased as the duration of exposure to 28C Bailey, L.H. and E.Z. Baily. 1976. Hortus Third. Fla. soil increased from 0 to 28 days and as soil MacMillan, New York. Roh, M.S. and R.H. Lawson. 1987. Research and temperature increased from 22 to 31C. Cul- Halevy, A.H. and A.M. Kofranek. 1984. Evalu- development on new crops in the United States tivars varied in their sensitivity to high tem- ation of lisianthus as a new flower crop. Department of Agriculture. Acta Hort. 205:39- peratures during the seedling stage. The pink HortScience 19:845-847. 48.

HORTSCIENCE 27(8):887-889. 1992. facilitates the natural transfer of pollen grains to the stigmatic surface. The second study Role of Self-pollination and Reciprocal was conducted in the United States on ‘Le Grand, derived from peach-almond inter- Stigma/Anthers Position on Fruit Set specific hybridization and possessing a flower morphology similar to that of ‘Truoito’ (Weinbaum, 1985). ‘Le Grand’ was incap of Eight Self-compatible Almonds able of achieving optimal fruit set using either A. Godini, L. de Palma, and M. Palasciano artificial and natural selfing of flowers. Given the different results of the above Istituto di Coltivazioni Arboree, Università di Bari, 70126 Bari, Italy studies and the few genotypes concerned, Additional index words. Amygdalus communis, Prunus dulcis, Apulian almonds, hand further documentation involving a wider range self-pollination, flower morphology, self-fruitfulness of self-compatible almond cultivars seemed necessary. Therefore, we conducted a study Abstract. A study to ascertain the highest possible fruit set following self-pollination to determine if self-pollination may allow self- of eight self-compatible cultivars of almond (Amygdalus communis L. = Prunus dulcis fruitful almond cultivars to bear sufficient Mill) was carried out in Apulia (southern Italy). Fruits set from daily hand-selling crops to be planted in solid blocks i) without were high, compared to those from unassisted self-pollination. The results support provision for cross-pollination (Kester and growing most of the self-compatible Apulian almonds in solid blocks, without need for Asay, 1975) and/or ii) without insect vectors cross-pollination, provided that self-pollination is optimized by insect vectors. The (Weinbaum, 1985). To answer these ques- relative capability of the cultivars to set fruits by unassisted self-pollination was inde- tions, the capability of eight Apulian self- pendent of the reciprocal stigma/anthers position within the same flower. compatible almond cultivars to reach optimal fruit set after artificial and natural self-pol- About 30 sweet and bitter almond culti- artificial selfing on the fruit set of single self- lination was tested in 1990 at Valenzano vars native to Apulia (southern Italy) have compatible almond cultivars have been com- (Apulia, Italy). been identified as self-compatible. Their fruit pleted. The first study, undertaken in Greece, Cross-pollination was prevented by cov- set after natural self-pollination averaged demonstrated the capability of the Apulian ering single mature trees throughout bloom <15% and never >26% (Godini, 1977b; cv. Truoito (syn. “Tuono”) to exceed the with insect-proof cheesecloth cages, thus Godini et al., 1990a; Reina et al., 1985). optimal fruit set by both artificial and natural eliminating any possible injury to flowers by Since 25% to 40% fruit set is considered selfing of flowers (Vasilakakis and Porlin- branch-bagging (Grasselly and Olivier, 1976). optimal for almond, depending on the bloom gis, 1984). These successful results have been A single tree was chosen for each cultivar. density of the cultivar (Kester and Griggs, attributed to the intimate stigma/anther con- On each tree, 20 flowering branches, each 1959a), the fruitfulness of Apulian self-com- tact within single ‘Truoito’ flowers, which bearing 80 to 120 blossoms, were divided patible almonds by natural self-pollination is biologically interesting but horticulturally Table 1. Bloom date and duration, number of hand self-pollinations, and weather conditions through- inadequate. out the pollination period of Apulian almonds. Two studies on the effect of natural and

Received for publication 8 July 1991, Accepted for publication 18 Feb. 1992. Research funded by the Ministero Agricoltura e Foreste (Italy), special Grant ISMEZ (Paper no. 386). We are grateful to V. D’Alessandro and P. Simeone for their invaluable cooperation in field trials. The cost of publishing this paper was defrayed in part by the payment of page charges. Under postal regulations, this paper therefore must be hereby marked advertisement solely to indicate this fact.

HORTSCIENCE, VOL. 27(8), AUGUST 1992 887