Materials and Methods

Phylloclades were harvested from vegetative cacti on 15 to 20 Apr. 2000 in Expt. 1; 15 to 20 Apr. 2001 in Expts. 2, 3, and 5; or as specified in Expt. 4. Phylloclades were rooted under natural daylengths [lat. 42°45′N, 13.4 h on 15 Apr. (Astronomical Almanac for the Year 2002, 2000)] in a 50 peat:50 perlite (by volume) mix in 55-cell (23 mL per cell) plug trays. Air and bench temperature were set at 23 and 25 °C, respectively. Plants were grown under intermittent mist until transplant. Mist duration was 4 s every hour or when accumulated light integral reached 0.2 mol·m⁻², whichever occurred first. Rooted plugs were transplanted into 10.2-cm (0.46-L) plastic pots containing 70 peat:30 perlite (SureMix Perlite; Michigan Grower Products, Galesburg, MI). Plants were irrigated with well water (containing 95, 34, and 29 mg·L⁻¹ Ca, Mg, and S, respectively) supplemented with water-soluble fertilizer to provide the following (mg·L⁻¹): 125 N; 12 P; 125 K; 13 Ca; 1.0 Fe, B, and Mo; and 0.5 Mn, Zn, and Cu (MSU Special; Greencare Fertilizers, Chicago) acidified to a titratable alkalinity of 140 mg·L⁻¹ CaCO₃. Fertilization was ended during vernalization treatments and resumed during forcing.

Plants were maintained in a greenhouse under 16-h LD from propagation until the beginning of SD or vernalization treatments. In Expt. 1, LDs were provided by pulling blackout cloth over the plants from 1700 to 0800 hr and lighting with incandescent lamps [(5 μmol·m⁻²·s⁻¹ photosynthetic photon flux (PPF)] from 1700 to 0000 hr. Long-day photoperiods in Expts. 2 through 5 were provided by high-pressure sodium (HPS) lamps (80 μmol·m⁻²·s⁻¹ PPF) for 16 h as needed from 0600 to 2200 hr to maintain a minimum of 80 μmol·m⁻²·s⁻¹. Prevernalization SD photoperiods (10-h) were achieved by pulling blackout cloth over the plants from 1700 to 0800 hr and lighting (5 μmol·m⁻²·s⁻¹ incandescent irradiation) from 1700 to 1800 hr. Plants were grown under natural daylengths during vernalization in Expts. 2 through 5 and as described in Expt. 1. Long days during forcing were provided by pulling blackout cloth over the plants from 1700 to 0800 hr and providing incandescent irradiation (5 μmol·m⁻²·s⁻¹) from 1700 to 0000 hr (16-h photoperiod). Temperature in each greenhouse was measured in an aspirated chamber every 10 s, and hourly averages were recorded by a CR-10 datalogger (Campbell Scientific, Logan, UT). Plants were grown at 21.5 ± 2.7 °C (setpoint = 20 °C) from propagation until vernalization and during forcing.

A randomized design was used in all experiments. Data were collected on percentage of plants flowering, days from the beginning of forcing (after vernalization) to visible anthers (time to flowering), number of apical phylloclades, number of apical phylloclades with flowers, and number of flower buds. Percentage of plants flowering was analyzed using 1-sample binomial tests (SPSS, 1999) in each treatment to test for difference from 99% flowering. Data for PFAP and BFAP were transformed [arcsin√PFAP and log(BFAP), respectively] and analyzed with a univariate general linear model in SPSS (SPSS, Chicago; SPSS, 1999). Days to flowering was analyzed using survival analysis (Cox regression; Dubois et al., 2003; SPSS Inc., 1997) treating flowering as a failure event with a cutoff of 80 d. Backward stepwise selection was used to include covariates for Cox regression analysis. Reported data for all experiments are raw means.

Results

Expt. 1

Temperatures from 7.5 to 15 °C and low DLI (≈4 mol·m⁻²·d⁻¹) during the 8 weeks of vernalization, combined with SD before vernalization, produced the greatest flowering response (Fig. 1). High DLI during vernalization reduced the percentage of flowering plants (Fig. 1A versus Fig. 1B), PFAP (Fig. 1C versus Fig. 1D), and BFAP (Fig. 1E versus Fig. 1F) when vernalization temperature was 15 °C or greater. The PFAP was greater on average when plants were vernalized under low DLI compared with high DLI (vernalization DLI P < 0.001 in global analysis of variance). Optimal vernalization temperature under the low vernalization DLI was 7.5 to 15 °C (100% of plants flowering, greater than 70% of apical phylloclades flowering, and greater than 1.2 BFAP), but the optimal range under the high vernalization DLI was 7.5 to 12.5 °C (86% or more of plants flowering, 60% or more of apical phylloclades flowering, and 1.2 or more BFAP). Short-day treatment (10-h) before vernalization generally increased the PFAP (Fig. 1C–D) and BFAP (Fig. 1E–F) for plants vernalized at 7.5 to 12.5 °C.

View Figure

• Download as Powerpoint
• Download Figure

Plants given prevernalization under LD and vernalized for only 4 weeks flowered poorly (58% to 73% plants flowering after 7.5 to 12.5 °C vernalization) compared with plants given SD before 4 weeks of vernalization (93% to 100% of plants flowering after 7.5 to 12.5 °C treatment, data not shown). A maximum of 83% PFAP and 1.45 BFAP were recorded after 4 weeks of vernalization for plants given a prevernalization treatment of high DLI and SD and then vernalized under low DLI at 7.5 °C. Because of the reduced flowering of plants provided with 4 weeks of vernalization, data for these treatments are not provided.

Expt. 2

Flowering uniformity generally increased as duration of prevernalization SD treatment and vernalization increased (Fig. 2). The maximum flowering index was typically found in plants given 6 weeks of SD in combination with 8 weeks of vernalization. Vernalization at 10 or 12.5 °C for a duration of 8 weeks elicited the greatest flowering uniformity regardless of prevernalization photoperiod treatment. ‘Jan’ vernalized at 12.5 °C for 8 weeks with no prevernalizaion SD flowered with uniformity similar to plants given 4 or 6 weeks of SD before 6 weeks of vernalization at 12.5 °C. Flowering index was low for plants vernalized at 15 °C, especially for plants not previously receiving an SD treatment. ‘Jan’ flowered more uniformly than ‘Rood’, especially at 15 °C or at 12.5 °C with fewer than 8 weeks of vernalization.

View Figure

• Download as Powerpoint
• Download Figure

Number of BFAP increased as the duration of vernalization increased at all three temperatures (Fig. 3). The effect of SD on BFAP before vernalization was significant only on 10 °C vernalization treatments. Plants given SD for 6 weeks and vernalized for 6 or 8 weeks at 10 °C had ≈0.4 more BFAP than if no SD treatment was provided before vernalization.

View Figure

• Download as Powerpoint
• Download Figure

Time to flower was influenced by vernalization duration and temperature (Table 1). ‘Jan’ vernalized at 10 or 12.5 °C for 6 or 8 weeks flowered on average 3 d earlier than plants vernalized at 15 °C for 6 or 8 weeks. ‘Jan’ given 2 to 6 weeks of SD before vernalization flowered an average of 2 d earlier than plants given no SD before vernalization, but there was a significant interaction with duration of vernalization. Plants vernalized at 10 °C for 8 weeks flowered 8 to 21 d earlier than plants vernalized at 15 °C for 4 weeks.

Table 1. Effect of short-day (SD) duration before vernalization and vernalization duration and temperature on days to flower from the beginning of long days in Hatiora gaertneri Jan and Rood.

View Fullscreen

Expt. 3

All ‘Jan’ and ‘Evita’ plants flowered after 8 weeks of vernalization at 12.5 °C with a higher PFAP and more BFAP than plants vernalized for 4 weeks (Table 2). Plants prevernalized in SD for 6 weeks had a higher PFAP than plants prevernalized for 4 weeks. Plants vernalized for 8 weeks tended to flower earlier than plants vernalized for 4 weeks, although there were significant cultivar-specific interactions. ‘Evita’ flowered more than 3 d earlier when prevernalization photoperiod was 10 h compared with 11 h.

Table 2. Effects of prevernalization photoperiod and vernalization duration (at 12.5 °C) on flowering of Hatiora gaertneri Jan and H. ×graeseri Evita.^z

View Fullscreen

Expt. 4

Plants propagated earlier in the year generally had higher flowering uniformity as shown by a larger flowering index (Fig. 4). Plants not leveled or leveled before SD flowered most uniformly. Uniformity was very low for plants leveled during vernalization. ‘Jan’ flowered more uniformly than ‘Evita’.

View Figure

• Download as Powerpoint
• Download Figure

The number of buds per flowering apical phylloclade (BPFAP) was generally highest for plants propagated in April (Table 3). ‘Evita’ had fewer buds per flowering phylloclade than ‘Jan’, although the difference was small. Number of buds per flowering phylloclade was generally highest for plants leveled before SD or not at all. Plants propagated in July flowered later than plants propagated earlier for comparable leveling treatments.

Table 3. Effects of propagation date and leveling date on buds per flowering apical phylloclade and days to flower in flowering Hatiora gaertneri Jan and H. ×graeseri Evita.^z

View Fullscreen

Expt. 5

Dark vernalization at 0 or 2.5 °C killed the plants (data not shown) or did not substantially promote flowering (Table 4). ‘Jan’ had a higher PFAP (48%) than ‘Evita’ (32%) and ‘Jan’ took an average of 2 d longer to flower. Three, and especially 6 weeks, of short days before dark vernalization increased the percentage of plants flowering in most treatments, PFAP, and BFAP. Vernalization temperature did not affect the percentage of phylloclades flowering or BPFAP. Maximum flowering (100% of plants flowering and 38% to 46% of apical phylloclades flowering) was achieved with 3 weeks of SD followed by 56 d of vernalization at 5 °C or 6 weeks of SD followed by 56 d of vernalization at 10 °C.

Table 4. Flowering of Hatiora gaertneri Jan and H. ×graeseri Evita vernalized in a dark cooler at 0, 5, 7.5, or 10 °C for 23 or 56 d.

View Fullscreen

Discussion

Our first objective was to study the effects of photoperiod before vernalization and DLI before and during vernalization on flowering of Hatiora. We accomplished this by treating H. gaertneri ‘Jan’ with LD or SD photoperiods for 6 weeks before vernalization, five vernalization temperatures (from 7.5 to 17.5 °C), and low or high DLI before and during vernalization. Our research supports previous reports of Hatiora flowering as a short–long-day plants (Boyle, 1991; Peters and Rünger, 1971; Rünger, 1960), but flowering was enhanced after a 10 to 15 °C treatment during or instead of the SD phase (Fig. 1).

Phylloclade temperature may exceed vernalizing air temperature under high irradiance. A thermocouple placed in the apical areole of a single H. ×graeseri plant recorded average hourly temperatures consistently above the measured air temperature throughout an 8-d period in June 2002. The difference between plant and air temperature increased as PPF increased with a maximum difference of 14.7 °C (data not shown). The shoot temperature of many cacti has been recorded to be 15 °C or greater above ambient air temperature because of the closed stomata of crassulacean acid metabolism plants during the daytime and high-heat storage capacity caused by succulence, among other factors (Nobel, 1988). In our experiments, high DLI during vernalization, provided with HPS lamps, reduced all flowering responses at 15 °C compared with 12.5 °C or less (Fig. 1). The results of Expt. 5 show that light is not an absolute requirement during vernalization, and we have shown that supplemental light during vernalization does not increase horticultural quality (Fig. 1). Therefore, reducing the PPF intercepted by the plants during vernalization may improve flowering by limiting the increase of plant temperature above air temperature caused by increased solar gain.

Our second objective was to identify combinations of SD treatments before vernalization and vernalization treatments that increase flowering uniformity and BFAP. We accomplished this by treating H. gaertneri ‘Jan’ and ‘Rood’ with 0, 2, 4, or 6 weeks of SD followed by 2, 4, 6, or 8 weeks of vernalization at 7.5, 10, or 12.5 °C. Our research supports previous reports (Boyle, 1991; Peters and Rünger, 1971; Rünger, 1960) that extending the duration of inductive conditions, either with SD or low temperatures, enhanced flowering uniformity in Hatiora (Fig. 2). Marginal thermoinductive conditions (15 °C) yielded poor flowering uniformity unless an extended SD prevernalization treatment was given (Fig. 2). This illustrates the necessity of SD before vernalization under natural SD photoperiods if temperatures are higher than those for optimum vernalization (12.5 °C or greater).

The response to vernalization is considered quantitative in many plants (Lang, 1965), including Hatiora. Seventy-five percent of apical phylloclades flowered on H. gaertneri vernalized for 80 d at 10 °C, and only 46% flowered on plants vernalized for 60 d (Peters and Rünger, 1971). An increase in the duration of vernalization at 10 °C from 50 to 70 d increased the percentage of plants flowering from 79% to 96% and increased the PFAP from 8% to 55% in H. ×graeseri (Rünger, 1960). Within a Hatiora crop, flowering response to vernalization is certainly quantitative. If a plant flowers, it may flower on a single phylloclade or it may flower on all phylloclades. If a phylloclade forms flowers, it may have one flower or it may have numerous flowers. The horticultural quality of a plant improves as the PFAP and the number of BFAP increase. Increased horticultural quality increases the value of the plants, and increasing vernalization or inductive treatments increases horticultural quality of Hatiora.

Prevernalization SD treatment under optimal vernalization conditions improved flowering by increasing 1) flowering uniformity (Figs. 1 and 2); 2) the number of BFAP (Fig. 3); and 3) the number of buds above the minimum possible (data not shown). In Expt. 2, plants with more than 90% of apical phylloclades flowering and more than two BFAP (one more bud than the minimum, 37 plants total) were vernalized for 6 or 8 weeks at 10 or 12.5 °C with the exception of one plant vernalized at 15 °C (data not shown). The majority (62%) of these plants with favorable horticultural flowering characteristics were given 4 or 6 weeks of SD before vernalization, and 86% were given 2, 4, or 6 weeks of SD.

Our third objective was to observe differences in flowering after exposure to two different photoperiods of SD treatment before vernalization. We accomplished this by treating H. gaertneri ‘Jan’ and H. ×graeseri ‘Evita’ with 10- or 11-h SD before 4 or 8 weeks of vernalization. Differences in flowering responses were mainly the result of differences in duration of vernalization. Numerous treatment interactions affected time to flower.

It is likely that Hatiora responds to low PPF during part or all of civil twilight, effectively extending natural daylengths beyond sunrise to sunset. Photon fluence rates have been measured from ≈13 to 0.007 μmol·m⁻²·s⁻¹ during civil twilight (the period starting in the morning or ending in the evening when the sun reaches 6° below the horizon) depending on weather conditions (Kishida, 1989). Similar low fluence rates cause photoperiodic responses in many plants (Cockshull, 1984; Hughes et al., 1984; Whitman et al., 1998). Daylength, including civil twilight, does not reach 12 h until ≈16 Oct. or 11 h until ≈8 Nov. at lat. 42°45′N (East Lansing, MI) (Astronomical Almanac for the Year 2002, 2000). Given that the critical photoperiod for photoinduction is 11 to 12 h (Boyle, 1991), naturally photoinductive conditions for Hatiora are probably not reached until mid-October to early November in East Lansing, MI. If vernalization is to begin before this time, i.e., for early-season crops, or if temperatures are insufficient for vernalization, there are apparent benefits of SD treatment before vernalization to extend the induction period.

Our fourth objective was to identify propagation and leveling treatments that increase uniformity and BFAP. To accomplish this, we propagated H. gaertneri ‘Jan’ and H. ×graeseri ‘Evita’ over a span of 4 months and leveled the plants during different developmental stages. We then prevernalized the plants under SD for 6 weeks and vernalized them for 4 weeks. The plants were leveled before or during prevernalization SD or during vernalization.

It has been recommended to level Hatiora before induction for uniformity or a more compact shape if necessary (Boyle and Stimart, 1989; Nell, 1988), allowing a new set of phylloclades to reach 75% maximum size before vernalization begins (Boyle, 1997). Our data support the recommendation that additional vernalization is necessary if plants are leveled during vernalization (Nell, 1988). The low uniformity observed in ‘Jan’ and ‘Evita’ leveled during 4 weeks of vernalization (Fig. 4) suggests that apical phylloclades respond to vernalization more quickly than subtending phylloclades or that the process of leveling significantly damages meristems in subtending phylloclades. Flower primordia do not form on photoinduced H. gaertneri until after the SD phase of photoinduction (Boyle et al., 1994), so leveling during vernalization probably does not remove flower buds directly.

Hatiora gaertneri ‘Jan’ that were leveled 14 d after vernalization began and were vernalized for 8 weeks at 10 °C showed similar uniformity (Fig. 2) to ‘Jan’ leveled 47 d before prevernalization SD were begun or not leveled at all and vernalized for only 4 weeks (Fig. 4). We can conclude that fewer than 8 weeks of vernalization would be necessary in growing conditions similar those in Expts. 1, 2, and 3 to achieve a strong flowering response similar to that in Expts. 1, 2, and 3 if the plants are not leveled or are leveled before SDs are begun.

Our fifth objective was to determine the effects on flowering of SD treatment followed by dark vernalization in a cooler. To accomplish this, we treated H. gaertneri ‘Jan’ and H. ×graeseri ‘Evita’ with SD for 3 or 6 weeks and then we vernalized the plants from 0 to 10 °C in darkness for 23 or 56 d. Data presented here show that Hatiora flowered marginally when cooled in darkness (Table 4) in contrast to previous work that showed H. gaertneri failed to flower if vernalized for 60 to 80 d in constant darkness at 10 °C (Peters and Rünger, 1971). Prevernalizing the plants with SD increased the PFAP and BFAP after dark vernalization (Table 4). Increasing the prevernalization SD and vernalization duration also increased the percentage of plants flowering in most dark treatments. Vernalization at 5, 7.5, and 10 °C was equally ineffective at creating horticulturally desirable plants. Although many plants flowered, the horticultural quality of all plants was low after dark vernalization in a cooler. Future research may show that flowering of Hatiora vernalized in a cooler is improved by vernalization in the light of a greenhouse before or after the dark vernalization period or by providing a minimum PPF in the cooler.

Poor flowering in coolers at 0 to 10 °C vernalization may be a result of suboptimal temperatures in combination with no light. In previous research, a vernalization temperature of 11 or 14 °C was more effective than 5 or 8 °C (Peters and Rünger, 1971). It is possible that 5 to 10 °C is suboptimal for dark vernalization of Hatiora, although ‘Jan’ vernalized in a greenhouse at 7.5 °C under low DLI (≈4 mol·m⁻²·d⁻¹; Fig. 1) flowered completely and uniformly, especially when grown under SD before vernalization. It was previously shown that H. ×graeseri vernalized at 5 °C flowered (Rünger, 1960), but this treatment was preceded by 50 d of 15 °C, which may have vernalized or preconditioned the plants for vernalization at 5 °C. The differences between our research and previous reports are likely the result of cultivar differences and experimental conditions, including light intensity and photoperiod.

High irradiance or availability of adequate nutrition and sucrose may substitute for or act similarly to vernalization in some plants (Napp-Zinn, 1984; Roldán et al., 1999). In some plants, flower initiation is profoundly influenced by availability of accumulated assimilates (Bernier et al., 1993; Bodson and Bernier, 1985). For example, onion (Allium cepa L.) with high carbohydrate levels before vernalization initiated flowers more rapidly after vernalization than plants with low prevernalization levels of carbohydrates (Brewster, 1985). It has been suggested that new vegetative growth in H. gaertneri is supported by translocation of carbon from older tissue (Boyle, 1992). We have shown that high irradiance in combination with SD before vernalization increases PFAP and BFAP in Hatiora (Fig. 1). Flowering after vernalization in darkness could be improved by high irradiance before vernalization, after it, or both, but additional studies are warranted to test this theory. SDs before vernalization improved the flowering response in a greenhouse or in a dark cooler. Short days may act as a preconditioning treatment by directing assimilates to storage rather than growth processes.

Increased duration of vernalization and prevernalization SD, to a lesser extent, generally hastened flower initiation, evocation, or development under LD (Tables 1, 2, and 4). Plants given 6 weeks of 10-h prevernalization SD flowered 4 d faster than plants given 11-h SD (Table 2). These data show that subsequent forcing time may be reduced by extended induction duration and sufficiently short photoperiods.

Hatiora gaertneri and H. ×graeseri performed differently in our experiments. Hatiora gaertneri ‘Jan’ flowered more uniformly or with a higher PFAP than H. ×graeseri ‘Evita’ in Expts. 4 and 5 (Table 4; Fig. 4) and with more BFAP in Expt. 4 (Table 3). Hatiora ×graeseri is an interspecific hybrid between H. gaertneri and H. rosea (Barthlott and Taylor, 1995). Hatiora rosea is native to higher altitudes than H. gaertneri (Barthlott and Taylor, 1995) and has a greater requirement for vernalization than H. gaertneri (Boyle, 1990, 1995). Three late-flowering H. ×graeseri cultivars (‘5805’, ‘Phoenix’, and ‘Capella’) also flowered poorly under nonideal inductive conditions (data not presented). The H. rosea parentage of ‘Evita’ helps explain the relatively weaker flowering of ‘Evita’ compared with that of ‘Jan’ and supports previous research indicating strong genotype × environment interactions for flowering in Hatiora (Boyle, 1995).

Flowering requirements of Hatiora are comparable to a number of noncactus plants. Other plants with similar short–long-day flowering activity in which SD may be replaced or enhanced by vernalization include Campanula medium (Wellensiek, 1960, 1985), Coreopsis grandflora (Ketellapper and Barbaro, 1966; Runkle, 1996), Echinacea purpurea (Heide, 2004), Dactylis glomerata (Heide, 1987), Festuca rubra (Heide, 1990), and other temperate grasses (Heide, 1994). The critical temperature for heading in Festuca pratensis was 15 °C under SD and 12 °C under LD (followed by LD) (Heide, 1988), which is very similar to Hatiora. Trifolium repens (Thomas, 1961) and Echeveria harmsii (Rünger, 1962) also have shown short–long-day flowering requirements.

These data, in combination with previous research, confirm that uniform flowering of Hatiora requires a rather specific sequence of environmental conditions and cultural practices. First, the flowering response of older plants is greater than younger plants, although Hatiora are vegetatively propagated and juvenility should not be an issue. Short days for 4 to 6 weeks followed by vernalization at 7.5 to 12.5 °C promote flower induction. Flower development is then favored by forcing plants under LD at moderate temperatures of 17 to 20 °C for 7 weeks. Sale of uniform flowering plants requires 17 to 21 weeks from the start of the SD inductive process.