Heritability of Cold Tolerance (Winter Hardiness) in Gladiolus xgrandiflorus

Gladiolus(-i) are herbaceous perennials with long, sword-like leaves and tall spikes of showy, colorful flowers (Goldblatt et al., 1998). Numerous cultivars (>10,000) have been bred (Sinha & Roy, 2002) with extended vase life, floral novelty, or extended flowering periods (Kumar et al., 1999; Takatsu et al., 2002). Recent focus has included transformation for potential creation of a genetically modified organism (GMO) cultivar (Kamo, 2008).

There are several environmental factors that affect the winter hardiness trait, including low temperatures, variable snow/ice cover, low light periods, and secondary invasion by pathogens (Blum, 1988;Tcacenco et al., 1989;Walker et al., 1995). Winter hardiness is a necessary trait for herbaceous perennials growing in northern climates and is important for floriculture crops as well as consumers (Kim & Anderson, 2006). Underground storage organs in geophytes, e.g., corms, bulbs, tubers, rhizomes, etc., allow herbaceous perennials to survive cold winters. The underground structure of perennial Gladiolus is a corm or fleshy storage stem from which shoots and roots grow. Gladiolus is a genus that has not been studied to any great extent in the area of winter hardiness. Bettaieb et al., (2007) found that low temperature stress of 8°C caused increased catalase (CAT) activity and lower hydrogen peroxide (H 2 O 2 ) levels in gladiolus, but such information has not resulted in breeding for winter-hardy gladioli. Most or all cultivars are 'non-hardy' in Minnesota and other northern latitudes (Anderson, unpublished data).
The gladiolus breeding program at the University of Minnesota is part of a larger project in the Herbaceous Perennial Breeding Program to revolutionize geophytes. Gladiolus which are winter hardy or cold tolerant in USDA Zones 3-4 would allow this crop to overwinter in northern growing conditions and eliminate the need to dig the corms each fall and replant the subsequent spring. In recent years, due to the lack of adequate snow cover and cold temperatures the breeding program has had to supplement field overwintering over successive winters with laboratory freezing tests-a routine procedure widely used for woody and herbaceous perennial plants (Kim & Anderson, 2006). Our studies with USDA Z4 winter hardy chrysanthemums and gaura have shown that herbaceous perennial crowns and root systems must tolerate temperatures of -10°C (Z4) or -12°C (Z3) to survive Minnesota winters (Kim & Anderson, 2006;Pietsch et al., 2009). Most likely this is the case for gladiolus, although to the best of our knowledge, there have not been prior laboratory freezing tests of gladiolus corms for this purpose.
There are three inter-related research objectives for this study. First, selected cultivars and selections of Gladiolus xgrandiflorus will be tested to determine the range of cold tolerance at temperatures of 0°C to -10°C for all corm tissues and their subsequent regrowth potential. The second objective will be to determine the nuclear DNA genetic variation (using intersimple sequence repeats, ISSRs) of the tested genotypes in comparison with wild species and other hybrids. Third, the heritability (h 2 ) of cold tolerance will be assessed in hybrids derived from crossing tested winter hardy and non-hardy parents.

Programmed freezing tests
There were 5 reps/treatment (corms) x 4 treatments (freezing temperatures) for a total of 20 experimental units for each genotype included in the programmed laboratory freezing tests (Kim & Anderson, 2006). Freezing tests were conducted using a Tenney environmental growth chamber (Model No. T20S, Series 942; Tenney Environmental Lunaire Co., Williamsport, PA, USA) with a programmable Series 942 Ramping Controller (Watlow Controls, Winona, MN). This created precision in the "profile control" of pre-programmed multi-step ramp (linear temperature change) and soak (constant temperature) times (Waldron, 1997). Precision temperature data loggers (Veriteq SP-1000; Veriteq Instruments, Inc., Richmond, British Columbia, Canada) were buried in the center (adjacent to the planted corms) of randomly selected pots throughout the chamber to ensure the ramp and soak temperatures were actually obtained. Once the pots were placed in the growth chamber (+2°C) in a randomized complete block (RCB) design, 15.25 h elapsed after transfer for the soilless medium to return to +2°C. Cold tolerance was assessed at 0°C, -3°C, -6°C, and -10°C with 5 h ramp time periods (5 h each for +2°C to 0°C, 0°C to -3°C, -3°C to -6°C, -6°C to -10°C) and a 13.25 h soak time at each temperature treatment (0°C, -3°C, -6°C, -10°C) following completion of each ramp down period (Kim & Anderson, 2006). After each treatment, samples were removed for cold damage assessment upon completion of each soak time and placed into a cooling chamber at 2°C for 2-3 d (dark) until the soilless medium was completely thawed (>0°C).

Regrowth assay
Once thawed, potted corms for each temperature treatment were moved to glasshouse conditions of 20/17 °C (day/night) temperature with a 16 h photoperiod (0800-2000 HR) to assay freezing damage to corms and regrowth potential of roots from the basal plate and shoots from the apical meristem. Three weeks later, plant and corms samples were harvested; soilless medium was washed away from the roots. Root number, root lengths (cm), shoot number, and shoot lengths (cm) were recorded. Root and basal plate cell damage (live=white, dead=brown to necrotic) was also determined. Corm damage was assessed after cutting each corm exactly in half with a vertical cut. Tissues were scored on a 1-5 scale, with 1=dead (completely brown or black), 2=brown coloration, 3=green or yellow coloration, 4=slight discoloration, and 5=completely undamaged (Fuller & Eagles, 1978;Perry & Herrick, 1996;Kim & Anderson, 2006;Waldron, 1997).

Data analyses
The determination of LT 50 (the temperature at which 50% of the population sample is killed) was calculated for each genotype (Pomeroy & Fowler, 1973;Tcacenco et al., 1989).

www.intechopen.com
Quantitative data were analyzed by Analysis of Variance (ANOVA) and mean separations using Tukey's Honestly Significant Difference (HSD) test at α=0.05 using the Statistical Package for the Social Sciences (SPSS), Ver. 9.0 (SPSS, Chicago, IL, USA).

Experiment 2: molecular analysis of potential parents
Twenty genotypes of Gladiolus species (wild species accessions, hybrid cultivars, and numbered selections), including those used in Experiment 1, were used in this study. These represented a diversity of non-winter-hardy and winter-hardy accessions pre-screened (2005)(2006)(2007)(2008)

Polymerase Chain Reaction (PCR) conditions
Six UBC primers (University of British Columbia Vancouver, Canada, http://www.biotech.ubc.ca/frameset.html) with a high level of polymorphism and scorability (UBC 808, 810, 811, 813, 814, 818)-which have been successfully used with another monocot and geophyte (Lilium longiflorum; Anderson et al., 2010) and modifications of Roy et al. (2006) were used for PCR amplification and ISSR analysis. Each PCR contained 0.25 units Flexitaq™ DNA polymerase, 20µM of a single primer, 10mM dNTP, 1.0mM MgCl 2, 2µL diluted DNA solution and 5x Flexibuffer™, which was supplied with the polymerase, for a total volume of 25 µL in each reaction (Yamagishi et al., 2002). Amplifications were carried out in a thermocycler (PTC-100, MJ research Inc., Hayward, CA 94545, U.S.A.). Amplification conditions were 7 minutes of denaturation at 94°C followed by 50 cycles of temperatures [94°C for 30 seconds, 43°C for 70 seconds, 72°C for 120 seconds and a final 7 minute extension at 72°C]. Each sample was replicated thrice.

Gel electrophoresis
A 1.5 % agarose (mixed with 100 ml of electrophoresis, 1XTris Acetate) buffer was made and heated in a microwave at high power for ~2 mins. until completely melted. Ethidium bromide was added to the gel (final concentration 0.5 ug/ml) at this point to facilitate visualization of DNA after electrophoresis. Electrophoresis chambers were filled with 1XTris Acetate EDTA plus 5 µg/ml ethidium bromide (Sambrook et al., 1989). Sample DNA volumes (12 ml) were then loaded and a current of 75 Volts for 2~2.5 hours was applied. Gels were visualised under UV light and recorded using a Fluro Chem 500 camera (Alpha Innotech Corp., San Leandro, CA, USA). www.intechopen.com

Experiment 3: heritability of cold hardiness
In this experiment laboratory freezing tests were used to determine the level of heritability (h 2 ) (Kim & Anderson, 2006). Three population groups of segregating interspecific F 1 hybrids, derived from crossing G. xgrandiflorus hardy (H) or non-hardy (NH) parents (H x H, H x NH, NH x NH), were analyzed for cold hardiness. Since most of the F 1 hybrids are derived from interspecific hybrid parents, the F 1 generation is most likely to segregate (rather than the F 2 ; Anderson & Ascher, 1996). Seeds from 122 crosses or self pollinations (8,833 seeds) between the G. xgrandiflorus genotypes tested in Experiment 1 and an additional hardy numbered selection (VT03) or its inbred seedlings (VT03-2, -3, -5) made in 2006 were germinated. Due to low germination rates and seedling numbers, a total of 375 genotypes/population group were analyzed (Table 1).
The F 1 hybrids were grown in the glasshouse in 7.6 cm 2 containers filled with pasteurized gladiolus soilless mix (40% peat moss, 60% sand) under long day photoperiods (0600-2200 HR light) at 19/17°C (day/night) temperatures for 6 wk., followed by a dry down period of 4 wk at 21°C and darkness (to mimic summer dormancy). At the end of the dry down period, the corms were repotted in the gladiolus soilless mix and acclimated at 2°C for 1 wk prior to freezing tests.
A total of 375 corms/group were selected for the study (Table 1). The experimental design was a completely randomized design (CRD), consisting of 5 temperatures x 3 population groups x 5 freezing runs x 15 corms/population group/temperature, for a total of 1125 experimental units (corms). The five temperatures tested (0°C, -3°C, -6°C, -10°C, -12°C) included the four from Experiment 1 with an additional lower temperature (-12°C) in case the H x H genotypes were heterotic and tolerated lower temperatures than their parents. Corms from each of these three population groups were moved to the Tenney environmental growth chamber and re-acclimated at 2°C for 15.25 hours (cf. Experiment 1). Since the containers were smaller than those used in Experiment 1, the ramp times were shortened to 5 hr with a 2 hr soak time at each test temperature. Otherwise, the experiment protocols for laboratory freezing, thawing, reforcing in the glasshouse, as well as data collection, were the same as Experiment 1. Broad sense heritability (h 2 ) estimates and confidence intervals were also calculated (Knapp et al., 1985;Fehr, 1987). These h 2 estimates are a ratio of the total genotypic variance (additive, dominance, and epistasis) to the phenotypic variance.
Most mean shoot lengths were at 0.0 cm, three weeks after the freezing tests, due to complete death of the apical meristem ( Table 2). The significantly longest shoots occurred at the -3°C freezing temperature for 'Great Lakes' (0.84 cm) and Sel'n. 98-29 (1.74 cm) ( Table 2). Example www.intechopen.com shoot regrowth (Fig. 1D) later demonstrated that some stems had floral meristems and would flower (Fig. 1D, 0°C and -3°C) while others were thinner-stemmed and vegetative (Fig. 1D, -6°C). Genotypes were significantly different (p=0.026) whereas both freezing temperature treatments and genotype x freezing temperature interaction were highly significant (p≤0.001). (which was the best). These LT 50 s are similar in response to cold tolerant (LT 50 =-10°C) to non-cold-tolerant (LT 50 ≤-6°C) garden chrysanthemum and gaura rhizomes (Kim and Anderson, 2006;Pietsch et al., 2009). Since there were far fewer treatment and genotype combinations with shoots, compared to roots, apical meristems are more sensitive to cold and freezing than root initials. Thus, for the six tested gladiolus genotypes, the apical meristem (shoots) is more sensitive to freezing than roots which are likewise more sensitive than stem tissue (corms) to freezing temperatures. One could expect corms to remain in the ground after a freeze/thaw cycle until they have respired to death and/or been invaded by pathogens.

Genotype
Genotype tested (cultivar, sel'n.) Root No. Root length (cm) Corm damage Table 3. Lethal temperatures for root number, root lengths (cm), corm damage, and shoot lengths (cm) at which 50% (LT 50 ) of the population of gladiolus corm tissues for six genotypes were killed in the chamber freezing tests.

Experiment 2: molecular analysis of potential parents
Previous researchers only used two primers to determine clonal fidelity in gladiolus (Roy et al., 2006). Four (UBC 808,810,811,818) out of the six primers tested produced scorable, unequivocal bands, giving increased stringency. In the present study, 62 well-defined (clear and unequivocal bands) and scorable markers across replications were obtained. A previous gladiolus ISSR study (Roy, et al., 2006) had its own synthesized primers, rather than those readily available from the University of British Columbia (UBC). Several of these UBC primers were previously used in our laboratory to test genetic variation in clonal Lilium longiflorum (Anderson et al., 2010). A total of 38 (61.29%) primer pairs were polymorphic. Replications did not differ significantly in their banding patterns. Total numbers of scorable loci/primer ranged from 16 (Primer pair 808) to 19 (Primer pair 811) for all tested gladioli whereas the number of scorable, polymorphic loci/primer ranged from 8 (810, 818) to 12 (808).

Experiment 3: heritability of cold hardiness
Crossing groups, freezing temperatures, and their interaction were all highly significant (p≤0.001). In all crossing groups for -10°C, -12°C freezing temperatures, no living roots, root initials, apical meristems (shoots, shoot lengths) occurred (Fig. 3) and mean corm ratings for these temperatures (ranging from 1.1 to 1.2) barely exceeded the completely dead rating of 1.0 (Table 4). Thus, no transgressive segregants for corm ratings occurred with greater hardiness than the parents (Experiment 1, Table 2). In general, crossing groups involving at least one hardy parent (hardy x hardy, non-hardy x hardy) had significantly greater numbers and lengths of living roots and shoots than the non-hardy x non-hardy group.
Mean number of roots ranged from 0.0 (all three crossing groups at -10°C and -12°C) to 4.9 (non-hardy x hardy, -3°C) ( Table 4). The significantly greatest number of roots occurred in hardy x hardy crosses at 0°C, -3°C (Fig. 3A) and non-hardy x hardy at -6°C (Fig. 3B). In all crossing groups and freezing temperatures, the number of roots was significantly lower than that found for the parents (Table 2). Average root lengths varied from 0.0 (all three crossing groups at -10°C and -12°C) to 6.89 (non-hardy x hardy, -3°C; Table 4) with the significantly greatest root lengths found in hardy x hardy at -3°C and non-hardy x hardy at -3°C, -6°C (Fig. 3). Root lengths, in some cases (6.89 for non-hardy x hardy at -3°C, Table 4), exceeded parental values (6.36 for Sel'n. 98-29 at -3°C, Table 2). Overall, roots in the progeny averaged longer lengths than the parents. Root number heritability ranged from h 2 = 0.08 (hardy x hardy) to h 2 =0.67 (non-hardy x non-hardy) ( Table 4). Root length heritability was both negative (h 2 =-0.14, non-hardy x non-hardy crosses) to positive (h 2 =0.37) ( Table 4).
Corm ratings of the hybrids ranged from 1.1 (hardy x hardy and non-hardy x non-hardy at -12°C; non-hardy x hardy at -10°C) to 2.9 (hardy x hardy, 0°C) ( Table 4). Heritability of corm ratings was low, ranging from h 2 =-0.04 (hardy x hardy crosses) to h 2 =0.15 (non-hardy x nonhardy) ( Table 4). All corm ratings in all three crossing groups were significantly higher for freezing temperature treatmets of 0°C to -6°C than for lower temperatures (-10°C and -12°C) ( Table 4). The parental values for 0°C to -6°C corm ratings (Table 2) were significantly higher than any of the progeny. These differences could be attributed to the significantly smaller corm size and stem tissue volume/density since the hybrids were only two-year-old corms (Fig. 3) and ~1/5 the size of the tested parental corms of commercial size (Fig. 1). The average number of shoots among progenies ranged from 0.0 (all three crossing groups at -10°C and -12°C) to 1.1 (hardy x hardy at -6°C; non-hardy x hardy at 0°C, -3°C) ( Table 4). In this case, shoot numbers >1.0 indicated transgressive segregation over the parents which all had 1.0 shoots/corm (Experiment 1). In all cases, a significantly greater number of shoots were found at the 0°C, -3°C, and -6°C temperatures for all crossing groups. Mean shoot lengths ranged from 0.0 (all three crossing groups at -10°C and -12°C) to 10.7 cm (non-hardy x hardy at -6°C, Table 4). Non-hardy x non-hardy crosses for 0°C, -3°C, and -6°C had significatly longer shoots than those with 0.0 cm lengths. Shoot lengths at 0°C, -3°C, and -6°C for both hardy x hardy and non-hardy x hardy groups were significantly longer than any other freezing temperatures within these groups or freezing temperatures for all nonhardy x non-hardy crosses (Table 4). When shoot lengths exceeded 0.0 cm for all crosses at 0°C, -3°C, and -6°C (Table 2), these were significantly longer than any parents ( The total number of progeny with all living tissues after freezing ranged from zero (all three crossing groups at -10°C and -12°C) to 20 (26.7%; hardy x hardy at -3°C (Table 4). In general, a low range of hardy x hardy, non-hardy x hardy, and non-hardy x non-hardy progeny survived at 0°C, -3°C, and -6°C freezing temperatures (Table 4). Thus, while select progeny are hardy to -6°C, this is not within the minimum range required for herbaceous perennial survival in USDA Z4 and Z3 (-10°C and -12°C, respectively). Further breeding and testing will need to be done to determine whether or not the corms can be bred to survive lower temperatures.

Conclusion
A range in response was found among tested gladiolus genotypes for tissue damage after laboratory freezing (Experiment 1). 'King's Gold', for instance, was 'non-hardy' at 0°C to -10°C (dead roots/shoots). 'Great Lakes' was intermediate with living root/shoot tissues at 0, -3°C only whereas 'Lady Lucille' had living roots (0 to -6°C) and apical meristems (-3°C). LT 50 s = -10°C for stem tissues in most genotypes; all were severely damaged or dead at -12°C. The apical meristem is more sensitive to freezing than roots, which are likewise more sensitive than stem tissue (corms) to freezing. The genetic variation (ISSRs) for the tested genotypes ranged across a wide spectrum of the gladiolus genome (Experiment 2); no correlation with ability of tissues to survive cold temperatures was found, except Sel'n. VT-03 (USDA Z3) and G. dahlenii (GD=0.0.86). Two principle groups separated at GD=0.53 with only one monophyletic singleton ('Great Lakes'). Sel'n. 95-49 formed a distinct group with 'Lady Lucille' and 'Gemini' (GD=0.63). No transgressive hybrid segregants for corms occurred with greater hardiness than the parents (Experiment 3). In general, crosses with ≥1 hardy parent (hardy x hardy, non-hardy x hardy) had significantly greater numbers / lengths of living roots/shoots than non-hardy x non-hardy hybrids. In all crosses at -10°C, -12°C, no living roots, root initials, apical meristems occurred. Hybrids with ≥1 hardy parent had greater numbers / lengths of living roots and shoots than the non-hardy x non-hardy group. The highest number of roots occurred in hardy x hardy crosses (at 0°C, -3°C) and non-hardy x hardy (-6°C); root lengths, in some cases (6.89 for non-hardy x hardy at -3°C), exceeded parental values. Root number is barely heritable (h 2 = 0.08) for hardy x hardy hybrids but more so (h 2 =0.67) with non-hardy parents. Root length had a wider range of heritability (h 2 =-0.14 to 0.37). Heritability of corm ratings is likewise low (h 2 =-0.04 to 0.15).
A significantly greater number of shoots in progeny (all crossing groups) than the parents were found at the 0°C, -3°C, and -6°C temperatures, although heritability remained low (h 2 =-0.43 to 0.19). While select progeny are hardy to -6°C, this is not within the minimum range required for herbaceous perennial survival in USDA Z3-4. Further breeding and selection for increased cold tolerance would be required for gladiolus to reliably survive winter conditions in northern latitudes.