Freezing Tolerance of Imbibed Winterfat Seeds: Possible Mechanisms and Ecotypic Differences

D. Terrance Booth Yuguang Bai James T. Romo

Abstract—We found survival of fully hydrated winterfat seeds 1978; Keefe and Moore 1981). Winterfat seeds imbibe greater (Eurotia lanata (Pursh) Moq.) in diaspores harvested from two amounts of water at 0 to 5°C than at warmer temperatures –1 U.S.A. and one Canadian location, and cooled to –30°C at 2.5°C h (Bai and others 1999). Therefore, we wondered if seed was similar to that of uncooled seeds. Seed and diaspore morphology hydration at snowmelt increases winterfat seed mortality appeared to contribute to freezing tolerance. The Canadian collec- from a freezing event. tion germinated more slowly at low temperatures and was more sensitive to imbibition temperature x freezing stress interaction, suggesting habitat correlated differences among the seed collections. Materials and Methods ______Seed Sources

Warm winter and early spring temperatures and melting Winterfat diaspores were hand collected in October, 1994, snow often saturate the soil and rehydrate seeds capable of from Matador, Saskatchewan, Canada; Sterling, Colorado, absorbing water. This includes the seed-containing dis- U.S.A.; and Pine Bluffs, Wyoming, U.S.A. Detailed site de- persal units (diaspores) of winterfat (Eurotia lanata (Pursh) scriptions were reported by Bai and others (1998a,b). Har- Moq.; Krascheninnikova lanata (Pursh) Mueese & Smit; vested diaspores were stored in paper bags at room tempera- Ceratoides lanata (Pursh) J. T. Howell), which often begin ture until used in experiments 4 to 6 months after harvest. germination with snowmelt (Springfield 1972; Hilton 1941; The mean dry weights of threshed seeds from Matador, Woodmansee and Potter 1971; Booth 1987). However, snow- Sterling, and Pine Bluffs were 25, 23, and 18 mg per 100 melt is frequently followed by a return to winter-like condi- seeds, respectively. tions that can expose fully hydrated, germinating seeds to –20°C or colder temperatures (Becker and Alyea 1964). Our knowledge of freezing in hydrated seeds has been Seed Imbibition and Differential Thermal largely limited to studies of lettuce (Lactuca sativa L.) Analysis (DTA) (Junttila and Stushnoff 1977; Stushnoff and Junttila 1978; Diaspores were imbibed at 0, 5, 10, and 20°C in darkness Keefe and Moore 1981, 1983; Roos and Stanwood 1981). to full hydration. We defined full hydration as the water These studies have identified two mechanisms of adaptation content 8 h before germination began. Full hydration re- to freezing stress. The first, supercooling, occurs when water quired about 24 h at 20°C and ≥ 120 h at 0°C (Bai and others in the seed is cooled below the freezing point without the 1999). Ice crystal formation is an exothermic process where formation of ice crystals. This can result from rapid cooling heat is released (heat of fusion) at the rate of 80 cal/g (20oC h–1) (Junttila and Stushnoff 1977; Ishikawa and Sakai (Masterton and Slowinski 1978). Differential thermal analy- 1982) or from the absence of nucleators. The second toler- sis (DTA) detects the exotherm created by the heat of fusion. ance mechanism is freeze-desiccation of the embryo. This is Differential thermal analysis of fully hydrated diaspores a redistribution of water inside the embryo to external ice was accomplished by sealing the diaspores in 0.25 ml tin crystals that form during slow cooling (1 or 2°C h–1) (Keefe capsules (LECO Corp., St. Joseph, MI), one diaspore per and Moore 1981; Ishikawa and Sakai 1982). Seed mortality capsule, with a thermocouple contacting the diaspore sur- due to freezing is generally the result of ice forming in the face and another thermocouple outside the capsule (Bai and embryo when the cooling rate exceeds the freeze-desiccation others 1998b). Diaspore and air (outside the capsules) tem- rate (Junttila and Stushnoff 1977; Stushnoff and Junttila peratures were recorded with a datalogger at 1 min. inter- vals. The resolution provided by the equipment was 1.66 nV or 0.004°C. Exotherms were detected by comparing the In: McArthur, E. Durant; Ostler, W. Kent; Wambolt, Carl L., comps. 1999. difference in temperatures inside and outside the capsules. Proceedings: shrubland ecotones; 1998 August 12–14; Ephraim, UT. Proc. The encapsulated diaspores were held at 0°C for 1 h, then RMRS-P-11. Ogden, UT: U.S. Department of Agriculture, Forest Service, –1 Rocky Mountain Research Station. cooled at 2.5°C h from 0 to –30°C over a 12 h period. D. Terrance Booth is with USDA, Agricultural Research Service, High Additionally, a sample of diaspores was cooled to –50°C, and Plains Grasslands Research Station, Cheyenne, WY 82009, U.S.A.; Yuguang Bai is with Kamloops Range Research Unit, Agriculture and Agri-food paired samples of diaspores, seeds, and embryos were cooled Canada, 3015 Ord Road, Kamloops, BC V2B 8A9, Canada; and James T. Romo to –30°C. Our experimental design was a randomized com- is with Department of Plant Sciences, University of Saskatchewan, plete block with three replications arranged in blocks over Saskatoon, Sask. S7N 5A8, Canada. This paper is adapted from Bai and others 1998a. time and 10 diaspores per replicate.

USDA Forest Service Proceedings RMRS-P-11. 1999 97 Seed Germination and Seedling Vigor When hydrated diaspores, threshed seeds, and embryos from the Matador collection were cooled together, we ob- Twenty diaspores per experimental unit were imbibed served both an HTE and an LTE for diaspores, but only one and cooled as described above, retrieved from the freezer at exotherm for seeds and for embryos. Mean temperatures for 0, –6, –10, or –30°C, incubated at 0°C for 24 h, then at 5 or diaspore HTEs and LTEs were –6.2±0.16 and –20.5±0.77°C, 20°C under 12 h light. Germination was counted at daily respectively. Mean temperatures for seed and embryo intervals up to 14 d. Diaspores were considered germinated exotherms were –21.3±0.79 and –14.9±0.82°C, respectively if the radicle was ≥ 2 mm. Seedling axial lengths after 14 d (data not shown). No diaspores were observed to have more were obtained using a digitizing tablet, and the lengths used than two exotherms, including those cooled to –50°C (data as an indicator of seedling vigor (Booth and Griffith 1994). In not shown). this paper we will concentrate on seedling axial length in Although imbibition temperatures influenced the tem- response to temperature treatments. peratures at which exotherms occurred, they did not affect freezing-related seed mortality. This is a significant con- trast with lettuce where greater seed water reduced freezing Data Analysis tolerance (Roos and Stanwood 1981; Keefe and Moore 1983) Data were first analyzed with ANOVA or general linear and where the formation of embryo ice is a fatal event model (Snedecor and Cochran 1980) over the three seed (Junttila and Stushnoff 1977; Stushnoff and Junttila 1978; collections, and were also analyzed in each collection where Keefe and Moore 1981). interactions occurred between seed collection and treat- To better understand freezing tolerance in hydrated win- ment. Data were further analyzed within each imbibition terfat seeds as compared to that of lettuce seeds, we present temperature or cooling temperature. Statistical significance a possible tolerance mechanism, a tentative model, embody- was assumed at P≤0.05 and means were separated by using ing probable reasons why lettuce seeds are killed at –30°C LSD. and winterfat seeds are not. The important points of logic supporting our model are as follows: 1. Winterfat diaspores supercooled to –4 to –6°C, as indi- Results and Discussion ______cated by the HTEs. In lettuce the HTE is caused by water freezing inside the pericarp but outside the seed endosperm DTA Results and Freezing Tolerance (Junttila and Stushnoff 1977). Both an HTE and an LTE Between 0 and –30°C two exotherms were detected, indi- occurred in winterfat diaspores, but only one exotherm was cating two separate ice-forming events at different tempera- observed for threshed seeds or embryos. This implies that ice tures. The first, or warmer, event is called the high tempera- crystals forming between the pericarp and the bract wall or ture exotherm (HTE), and the second, or colder, event is the between the testa and the pericarp wall, generate winterfat low temperature exotherm (LTE). Temperatures at which HTEs. HTEs and LTEs occurred in fully hydrated seeds were 2. When an HTE occurs depends on the size of a water similar among collections. They differed by imbibition tem- body (among other things) and the duration of cooling (Salt perature (P<0.01) with seeds imbibed at 0 or 5°C having 1961). The winterfat diaspore bract-pericarp interface, with HTEs and LTEs at warmer temperatures than those im- its abundance of small hairs (Booth 1988), is an area where bibed at 10 or 20°C (table 1). The average temperature for all a relatively large body of water is likely to exist when the HTEs and LTEs was –4.6 and –17.6°C, respectively. The LTE diaspore is hydrated. An accumulation of water at this hairy range was –3.7 to –26.8°C with 12% of all LTEs occurring surface is likely to promote the formation of ice crystals at equal to or warmer than –10°C. The pattern of exotherm the relatively warm –4 to –6°C observed for the HTE. occurrence is consistent with imbibition-temperature-based 3. The hairy surface of the winterfat pericarp appears to differences in seed moisture of fully hydrated diaspores that be a safe (for the embryo) place for ice crystals to form. We we have previously observed (Bai and others 1999). speculate that the ice crystals evident by an HTE initiate

Table 1—Temperatures (mean±SD) at which exotherms occurred for hydrated winterfat diaspores. Diaspores were collected from Matador, Saskatchewan; Pine Bluffs, Wyoming, and Sterling, Colorado, and were cooled at 2.5°C h–1. HTE: high temperature exotherm; LTE: low temperature exotherm.

Imbibition temperature (°C) Exotherm Collection 0 5 10 20

HTE Matador –4.3±0.6 –4.2±0.5 –5.1±0.7 –5.1±0.5 Sterling –4.1±0.4 –4.1±0.6 –5.0±0.8 –4.8±0.7 Pine Bluffs –4.0±0.4 –4.3±0.5 –4.7±0.6 –5.1±0.8 LTE Matador –15.1±5.4 –18.4±4.3 –19.9±4.6 –18.4±4.4 Sterling –14.6±5.5 –16.8±6.1 –20.0±4.8 –18.7±5.9 Pine Bluffs –14.0±6.5 –17.6±4.7 –17.9±6.1 –19.2±5.6

98 USDA Forest Service Proceedings RMRS-P-11. 1999 freeze-desiccating conditions by creating an osmotic gradi- We noted that critical tissues of winterfat seeds are ent in which liquid water in the embryo moves to extra- adjacent to the highly permeable testa. This is not the case embryo ice crystals. Freeze desiccation is the means by for lettuce and privet that have linear and miniature embryo which hydrated seeds tolerate cooling below about –15°C morphologies (Atwater 1980) and more cylindrical and spheri- (Keefe and Moore 1981; Ishikawa and Sakai 1982), and we cal gross morphologies. This means that lettuce and privet suggest diaspore morphology promotes freeze desiccation of embryos may not freeze desiccate as quickly as winterfat, the embryo. and that ice in lettuce or privet embryos may form larger 4. Membrane permeability is a fundamental aspect of crystals than are formed in winterfat embryos. Winterfat freeze desiccation (Lyons and others 1979). The winterfat embryo ice may form between cells or within cells. Evidence seed testa is membranous and highly permeable (Booth and from Differential Scanning Calorimetry studies of pea McDonald 1994) and is unlikely to interfere with emigrating (Pisum sativum L.) and soybean (Glycine max Merr.) using water molecules. isolated seed tissues indicates that water may freeze in the 5. The flat, peripheral-linear morphology of winterfat cell with no apparent damage to seed tissue (Vertucci 1988, seeds allows the highly permeable testa to surround critical 1989). [In animals, intracellular freezing was also thought tissues—cotyledons, hypocotyl, and radicle and in a manner fatal, but an exception was found in an Antarctic nematode that creates greater testa surface area compared to a sphere Panagrolaimus davidi (Timm) (Wharton and Ferns 1995).] or cylinder (fig. 1). Whether winterfat embryo ice forms in cells, between cells, Therefore, we speculate that winterfat’s HTE represents or between tissues and organs, it is reasonable to expect that the initiation of a freeze-desiccation process that reduces the large crystals are more likely to inflict greater damage. amount of water in the seed embryo before an LTE occurs Thus, if winterfat diaspore and seed morphology does pro- (Ishikawa and Sakai 1982). Thus, it appears that winterfat mote extra-embryo ice formation, and dehydration of critical diaspore bracts and the peripheral-liner morphology (Atwater tissues, it would help explain why hydrated winterfat seeds 1980) typical of chenopod seeds may contribute to freezing tolerate –30°C. tolerance in hydrated winterfat seeds. The difference be- tween lettuce and winterfat’s ability to survive an LTE event may also be in where LTE-producing ice forms. The place of Ecotypic Differences ice formation in hydrated seeds varies by species (Ishikawa Germination, with some exceptions in the Matador collec- and Sakai 1982) and extra-organ freezing, as occurs be- tion, was ≥ 70% for all collections and was not significantly tween the cotyledons and endosperm of common privet reduced by cooling to –30°C. For all collections, germination (Ligustrum vulgare L.) (Gazeau and Dereuddre 1980), may was not affected by imbibition temperature if diaspores were cause mortality.

Figure 1—Winterfat seed (A), and a transverse cross-section (B) (as indicated by the white line in (A)), with a scanning electron microscope. The cross-section shows the relationship of surface area and critical tissues. The testa is thin, transparent, and not visible as a separate entity in dry seeds such as these. (Photograph (A) is x20. Photograph (B) is x60 and courtesy of Dr. William Wergin, USDA-ARS Electron Laboratory, Beltsville, Maryland, U.S.A.)

USDA Forest Service Proceedings RMRS-P-11. 1999 99 not cooled below 0°C (data not shown). However, germina- Summary and Conclusions ______tion of the Matador and Sterling collections decreased at the higher imbibition temperatures after the diaspores were We accept the hypothesis that fully hydrated winterfat cooled to –6, –10, or –30°C when post-cooling incubation was seeds are tolerant of –30°C events. We found that post- 5°C. freezing seedling vigor varied by imbibition temperature; Axial lengths from seeds cooled to –30°C was significantly therefore, we conclude that imbibition temperature influ- less than for other cooling regimes; but, there were several ences freezing tolerance. We postulate that diaspore struc- exceptions (table 2). The U.S.A. collections appear to have ture and the peripheral-linear morphology of the winterfat had a wider range of imbibition temperatures within their seed contribute to freezing tolerance of hydrated seeds. We ability to tolerate –30°C without reduced growth. Axial have documented habitat-correlated differences among win- length after incubation at 5°C was similar among seed terfat populations with respect to post-freezing vigor and the collections and imbibition temperatures where seeds were effect of imbibition temperature on that response. Some not cooled below 0°C (table 2). With freezing stress, axial winterfat seeds are not negatively affected by a sequence of length was greater for the Matador and Sterling collections melting snow resulting in seed hydration, followed by freez- than for the Pine Bluffs collection (Bai and others 1998a). ing stress (–30°C), then cool (5°C) temperatures. Other Under the 20°C incubation temperature where seeds were seeds were found to survive this sequence, but freezing cooled below 0°C, seedling axial length was influenced by the damage was reflected in reduced germination rates, and interaction of seed collection and imbibition temperature. reduced seedling vigor such that these injuries may limit Seeds from the Matador collection had similar axial lengths field establishment or seedling survival. among imbibition temperatures (table 2), while the Sterling collection axial length was greatest for seeds imbibed at 20°C, and lowest when imbibed at 5°C. The Pine Bluffs References ______collection had no consistent trend among incubation tem- Atwater, B.R. 1980. Germination, dormancy and morphology of the peratures. Matador seeds were most sensitive of the three seeds of herbaceous ornamental plants. Seed Science and Tech- seed collections to imbibition temperature in the presence of nology. 8: 523-573. freezing stress. This suggests habitat correlated differences Bai, Y.; Booth, D. T.; Romo, J. T. 1998a. Winterfat (Eurotia lanata among seed collections and supports previous work on the (Pursh) Moq.) seedbed ecology: low temperature exotherms and ability of shrub populations to evolve site-specific character- cold hardiness in hydrated seeds as influenced by imbibition temperature. Annals of Botany. 81: 595-602. istics (Meyer and others 1989, 1990; Stutz 1982, 1989).

Table 2—Seedling axial lengths (mm) from hydrated winterfat diaspores of the Matador, Pine Bluffs, and Sterling collections imbibed at four temperatures then exposed to four cooling regimes. Seeds were incubated at 5 or 20oC for 14 days after cooling.

5°C20°C Cooling incubation incubation regime (°C) Matador P.B. Sterling Matador P.B. Sterling 0°C Imbibition 0 34 b B 36 a A 38 a A 49 a A 47 a A 50 ab AB –6 38 a A 32 b A 38 a A 56 a A 47 a A 51 b A –10 40 a A 33 b A 34 b A 56 a A 48 a A 46 b A –30 32 b A 30 b A 32 c A 36 b B 36 b A 38 c A

5°C Imbibition 0 36 a B 33 a A 33 a A 52 a A 46 a A 41 a C –6 36 a A 36 a A 37 a A 56 a A 46 a A 49 a A –10 37 a A 35 a A 39 a A 55 a A 49 a A 47 a A –30 30 b A 33 a A 34 a A 37 b B 39 b A 40 a A

10°C Imbibition 0 40 a B 34 a A 38 a A 52 a A 39 a B 46 bc BC –6 37 a A 34 a A 40 a A 54 a A 43 a A 50 a A –10 35 a A 37 a A 37 a A 55 a A 45 a A 55 a A –30 28 a A 34 a A 33 a A 43 a A 39 a A 42 c A

20°C Imbibition 0 43 a A 36 a A 39 a A 54 a A 48 a A 53 a A –6 43 a A 34 a A 38 a A 55 a A 49 a A 50 a A –10 40 ab A 37 a A 40 a A 53 a A 49 a A 53 a A –30 34 b A 30 b A 33 a A 33 b B 35 a A 41 a A

1Means with the same lower case letter are not significantly different at P≤0.05 within a imbibition temperature across cooling regimes (columns); means with the same capital letter are not significantly different at P≤0.05 within a cooling regime across imbibition temperatures (column intervals).

100 USDA Forest Service Proceedings RMRS-P-11. 1999 Bai, Y.; Booth, D. T.; Romo, J. T. 1998b. Developmental stages of Meyer, S. E.; McArthur, E. D.; Jorgensen, G. L. 1989. Variation in winterfat germinants related to survival after freezing. Journal germination response to temperature in rubber rabbit brush of Range Management. 51: 709-713. (Chrysothamnus nauseosus: Asteraceae) and its ecological impli- Bai, Y.; Booth, D. T.; Romo, J. T. [In press]. Imbibition temperature cations. American Journal of Botany. 76: 981-991. affects winterfat (Eurotia lanata [Pursh] Moq.) seed hydration Meyer, S. E.; Monsen, S. B.; McArthur, E. D. 1990. Germination and cold-hardiness response. Journal of Range Manage. 52. response of Artemisia tridentata (Asteraceae) to light and chill: Becker, C. F.; Alyea, J. D. 1964. Temperature probabilities in Patterns of between-population variation. Botanical Gazette. Wyoming. Agricultural Experiment Station Bulletin 415, Uni- 151: 176-183. versity of Wyoming, Laramie. Quamme, H. A. 1991. Application of thermal analysis to breeding Booth, D. T. 1987. Contributions to the reproductive autecology of fruit crops for increased cold hardiness. HortScience. 26: 513-517. winterfat (Eurotia lanata [Pursh] Moq.) with notes on direct Roos, E. E.; Stanwood, P. C. 1981. Effect of low temperature, cooling seeding methods. Ph.D. Dissertation, University of Wyoming, rate and moisture content on seed germination of lettuce. Journal Laramie. of The American Society for Horticultural Science. 106: 30-34. Booth, D. T. 1988. Winterfat diaspore morphology. Journal of Range Salt, R. W. 1961. Principles of insect cold-hardness. Annual Review Management. 41: 351-353. of Entomology. 6: 55-74. Booth, D. T.; Griffith, L. W. 1994. Technical note: Measuring post- Snedecor, G. W.; Cochran, W. C. 1980. Statistical methods. 7th germination growth. Journal of Range Management. 47: 503-504. edition. Ames, Iowa: Iowa State University Press. Booth, D. T.; McDonald, M. B. 1994. Cation concentration in post- Springfield, H. W. 1972. Winterfat seeds undergo after-ripening. imbibed winterfat seeds as influenced by imbibition tempera- Journal of Range Management. 24: 395-397. ture. Journal of Range Management. 47: 485-488. Stushnoff, C.; Junttila, O. 1978. Resistance to low temperature Gazeau, C.; Dereuddre, J. 1980. Effects de refroidissement prolonges injury in hydrated lettuce seed by supercooling. In: Li PH, Sakai sur la teneur en eau et la resistance au froid des tissus de la A, eds. Plant cold hardiness and freezing stress: proceedings. Troene (Ligustrum vulgare L. Oleacees). CR Acad. Sci. Paris Ser New York: Academic Press: 241-247. D. 290: 1443-1446. Stutz, H. C. 1982. Broad gene pools required for disturbed lands. In: Hilton, J. W. 1941. Effects of certain micro-ecological factors on the Aldon, E. F.; Oaks E.F., eds. Symposium reclamation of mined germinability and early development of Eurotia lanata. North- lands in the Southwest. New Mexico Chapter, Soil Conservation west Science. 15: 86-92. Society of America, P.O. Box 2142, Albuquerque, New Mexico: Ishikawa, M.; Sakai, A. 1982. Characteristics of freezing avoidance 113-118. in comparison with freezing tolerance: a demonstration of extra Stutz, H. C. 1989. Evolution of shrubs. In: McKell, C. M. ed, The organ freezing. In: Li P. H., Sakai A., eds. Plant cold hardiness biology and utilization of shrubs. Academic Press Inc., San Diego, and freezing stress Volume 2: proceedings. New York: Academic California: 323-340. Press: 325-340. Thompson, P. A. 1981. Ecological aspects of seed germination. In: Junttila, O.; Stushnoff, C. 1977. Freezing avoidance by deep super- Thompson J. R., ed. Advances in research and technology of cooling in hydrated lettuce seeds. Nature. 269: 325-327. seeds: proceedings. Center for Agricultural Publishing and Docu- Keefe, P. D.; Moore, K. G. 1981. Freeze desiccation: a second mentation, Amsterdam: 9-42. mechanism for the survival of hydrated lettuce (Lactuca sativa Vertucci, C. W. 1988. Thermal transitions associated with viability L.) seed at sub-zero temperatures. Annals of Botany. 47: 635-645. of seeds exposed to subfreezing temperatures. In: Earnest C. M., Keefe, P. D,; Moore, K. G. 1983. Freeze tolerance in hydrated ed. 17th North American Thermal Analysis Society Conference: Lactuca sativa L. seed: A model to explain observed variation Proceedings. Lake Buena Vista, Florida: 229-223. between seed lots. Annals of Botany. 51: 373-383. Vertucci, C. W. 1989. Relationship between thermal transitions and Lyons, J. M.; Raison, J. K.; Steponkus, P. L. 1979. The plant freezing injury in pea and soybean seeds. Plant Physiology. 90: membrane in response to low temperature; an overview. In: J.M. 1121-1128. Lyons and others (eds.). Low temperature stress in crop plants - Wharton, D. A.; Ferns, D. J. 1995. Survival of intracellular freezing the role of membrane: proceedings. Academic Press, New York: by the Antarctic nematode Panagrolaimus davidi. Journal of 1-24. Experimental Biology. 198: 1381-1387. Masterton, W. L.; Slopwinski, E. L. 1978. Chemical principles with Woodmansee, R. G.; Potter, L. D. 1971. Natural reproduction of qualitative analysis. W. B. Saunders Company, Philadelphia. winterfat (Eurotia lanata). Journal of Range Management. 24: 803 p. 24-30.

USDA Forest Service Proceedings RMRS-P-11. 1999 101