Seed Propagation Methods for Ratibida columnifera (Nutt.) Wooton & Standl. (2024)

General conditions.

Seeds collected from five open-pollinated genotypes (TX RC 8, TX RC 12, TX RC 30, TX RC 44, and TX RC 48) grown in trial plots (Somerville, TX; lat. 30.522288N, long. –96.429397W) were used for germination studies. The trial plots were in a full-sun location in the nursery in a native soil of silty clay loam (0–6 inches) and silt loam (6–80 inches) (U.S. Department of Agriculture Natural Resources Conservation Service, 2022). Midday photosynthetically active radiation (PAR) averaged 1638 µmol⋅m^–2⋅s^–1 photosynthetic photon flux density (PPFD) during midday (Fieldscout Quantum Foot-Candle Meter; Spectrum Technologies, Inc., Aurora, IL). These plants were irrigated on an as-needed basis and were fertigated to the leaching point weekly with 20–20–20 water-soluble fertilizer (Peters Professional 20–20–20 General Purpose; Everris, ICL, Dublin, OH). Because of the nonuniform maturation of inflorescences on R. columnifera, seed was harvested from fully dried seed heads when mature. Seeds were harvested in the late summer and were placed into a paper envelope and fully sealed. The seed heads remained in the envelopes until dry, and were then placed into sealed plastic bags and put into storage at 3 °C. Seed for germination experiments was used within 6 months of harvest (excluding storage condition and seed maturity experimentation).

To estimate whether germination requirements varied among a population of seedlings, seed from open-pollinated parent plants with varied phenotypic morphology was tested. These were used for the following germination experiments.

Before each germination experiment, a percent viability test using a triphenyltetrazolium chloride (TTC; Sigma-Aldrich, St. Louis, MO) staining method (Riebkes et al., 2015) was conducted on the seed from each maternal genotype. Soaking in 1% TTC solution for 48 h caused the embryo within the seed to stain red if viable (Riebkes et al., 2015). This seed viability test was performed on a subsample of 100 seeds from each treatment combination at the beginning of each germination trial. Germination data were normalized to the proportion of seed that was staining viable for each maternal genotype.

Effect of overnight hydration.

The first germination experiment examined effects of soaking seed in water before germination (Dec. 2018). Nine hundred seeds consisting of 300 seeds for each of three genotypes (TX RC 30, TX RC 44, and TX RC 48) were soaked overnight in 50 mL double-distilled water and then surface-dried. Nine hundred additional seeds remained dry until placed into petri dishes (polystyrene disposable sterile, 100 × 15 mm, catalog no. 25384-302; VWR International, Radnor, PA). One hundred seeds were placed into each sterile petri dish on moist filter paper (9 cm, Qual. 415, catalog no. 28320-041; VWR International), for a total of 18 petri dishes. There were three replicates of each treatment. There were nine dishes with 100 soaked seeds in each, and nine dishes with 100 unsoaked seeds in each, for a total of 1800 seeds tested. Seeds were placed under grow lights, with 16-h on/8-h off photoperiod increments (60 W) in laboratory conditions (23.3 °C). PAR under the grow lights under laboratory conditions averaged 39.1 µmol⋅m^–2⋅s^–1 PPFD. Seed was checked for germination every 48 h for 4 weeks after treatment initiation. Germination was defined as the emergence of the radicle (>1 mm).

A completely random design was used for this experiment. Factorial combinations of three genotypes (TX RC 30, TX RC 44, and TX RC 48) and two hydration treatments (control and overnight hydration). Analysis of variance and Tukey’s honestly significant difference (hsd) were used for the interactions among treatments and genotypes, with P ≤ 0.05 for significance, using JMP Pro 15 for continuous variables.

Effects of storage conditions on viability.

To examine optimal storage temperatures for maintaining viability, a cold-storage vs. room-temperature storage experiment was conducted. This experiment was replicated every 6 months to express the maintenance or degradation of viability, with continuation until depletion of available seeds after 18 months (began Dec. 2018). Seed was harvested in late Summer 2018 (August), was placed into a paper envelope, and was fully sealed. The seed heads remained in the envelopes until dry, and were then placed into sealed plastic bags and put into storage at either 23 °C or 3 °C. Three genotypes were used (TX RC 30, TX RC 44, and TX RC 48), at two different storage temperatures, replicated three times for a total of 1800 seeds. There were 300 seeds in each treatment combination at storage temperatures of 23 °C and 3 °C. Seeds were stored away from light and were kept dry in sealed, semipermeable plastic bags. At the predetermined storage length, seeds were placed into sterile petri dishes (VWR International) on filter paper (VWR International) moistened with double-distilled water. Germination conditions and data collection were as described for the hydration experiment. Three replicates of each genotype were placed in culture conditions in a completely random design. Analysis of variance, a generalized linear model, and Tukey’s hsd were used for the interactions among treatments and genotypes, with P ≤ 0.05 for significance, using JMP Pro 15 for continuous variables.

Effects of scarification and cold stratification pretreatments.

Cold stratification and scarification were examined in the same experiment (May 2019). Treatments included control (no stratification and no scarification), three moist stratification durations (30, 60, and 90 d at 3 °C stratification), and three sulfuric acid (51%, Macron Fine Chemicals, Center Valley, PA) scarification durations (5, 10, or 15 min acid scarification). Seed was harvested in Aug. 2018 and placed in semipermeable bags at 3 °C. Sufficient seeds from storage were placed into stratification at 30, 60, and 90 d before germination tests to allow scarification and stratification germination effects to be tested simultaneously with the control and scarification treatments. There were 300 seeds per treatment. Seeds were placed into sterile petri dishes on moist filter paper. Seeds were checked for germination every 48 h for 4 weeks after planting. Three replicates of 100 seeds per petri dish for each stratification and scarification treatment were arranged in a completely randomized design. Analysis of variance and Tukey’s hsd were used for the interactions among treatments and genotypes, with P ≤ 0.05 for significance, using JMP Pro 15 for continuous variables.

Effects of seed maturity on germination.

Requirements for seed maturity were examined by harvesting the columnar inflorescences at three stages of development: immature in appearance (green with ray flowers still attached), mid-age (ray flowers senesced, just beginning to change from green to gray-brown), and fully mature (ray flowers senesced, inflorescence brown to black-brown) but not shattered. Each seed head was divided into thirds and was named based on the region of the inflorescence: basal, middle, and apical. Seeds were aggregated for inflorescences of a given genotype and maturity stage. This was repeated with three different genotypes (TX RC 8, TX RC 12, and TX RC 30). Three hundred seed subsamples from each maturity stage × inflorescence segment × genotype were stained with TTC to estimate viability. This experiment took place in June 2020. Seed was harvested in June 2020 over the course of several weeks because of the nonuniform maturation of seed heads. Harvested seed was allowed to air-dry in paper envelopes, and was then transferred into semipermeable plastic bags for dry storage at 3 °C. Three petri dishes each contained 100 seeds that were placed on moistened filter paper for 24 h and placed under grow lights in laboratory conditions. Seeds were checked for germination every 48 h for 4 weeks after planting. Germination was defined as the emergence of the radicle (>1 mm). Factorial combinations of treatments were used in this completely randomized design. Analysis of variance and Tukey’s hsd were used for the interactions among treatments and genotypes, with P ≤ 0.05 for significance, using JMP Pro 15 for continuous variables.

Effect of overnight hydration.

Application of overnight hydration did not have a significant effect (P ≤ 0.05) on percent germination for the interaction among germplasm × soaking, nor for either of the main effects of soaking or germplasm (data not presented). This suggests that soaking R. columnifera seeds in double-distilled water before planting is not required.

Effects of storage conditions on viability.

No significant interactions among germplasm, duration of storage, or storage temperature were found. The main effect of storage temperature was also not significant (P ≤ 0.05). However, germplasm had a significant main effect on percent germination (Table 1). TX RC 30 had the highest percent germination at 82.5%, followed by TX RC 44 at 74.5%, with the lowest percent germination being TX RC 48 at 68.2%. All three germplasms differed from each other statistically (P ≤ 0.05). Storage duration also had a significant main effect on germination. The data show no significant difference in percent germination up to 12 months of storage (Table 1). By 18 months of storage, there was a slight but statistically significant reduction in percent germination, resulting in about a 10% loss in viability (Table 1). This minimal reduction in viability after 18 months suggests that seed can potentially be stored for longer periods if needed.

TX RC 30 had the highest percent germination of the three selections, followed by TX RC 44, and then TX RC 48 (Table 1). This indicates that the ability to retain viability over time can depend on germplasm selection. Seed longevity for R. columnifera as shown in our study is in concurrence with the observation that seed longevity can vary among accessions within a species because of differences in genotype and provenance (Hong and Ellis, 1996). In addition to genetic influences, differences in germination among accessions can be the result of the cumulative effect of the environment during seed maturation, harvesting, drying, the time of seed harvest, duration of drying, and the period before the seed is placed in storage (Hong and Ellis, 1996). Each of the germplasms in our experiment had maximum germination percentages well over the 60% observed in the study by Romo and Eddleman (1995), in which their seed was collected from the University of Saskatchewan’s Matador Research Station, a native environment. Our seed was collected from well-fertilized and irrigated stock in a nursery setting. The differences in cultural protocols of the stock plant could be why our seed had greater germination percentages overall. This may also be why our seeds were able to maintain a percent germination of 71.2% after 18 months of storage regardless of the temperature of storage. Storage duration also had a significant effect on germination (Table 1). There was no significant difference in germination for the first year of storage. It was not until 18 months that we began to observe statistically significant reductions in germination. This implies that seed may not need to be subjected to cold storage to retain viability in the short term. Future studies should be performed to determine if temperature has significant effects on retaining viability in long-term, multiyear storage durations and what impact freezing seeds may have on viability.

Effects of scarification and cold stratification pretreatments.

Application of stratification affected significantly the percent germination of the three R. columnifera genotypes (Table 2). There was a significant interaction effect among the three genotypes and the pretreatments (Table 2). Thus, the effect that the seed pretreatment had on percent germination depended on which genotype was examined. Stratification proved to be an important factor in germination success, as the literature suggests (Romo and Eddleman, 1995). TX RC 30 had a significant decrease in percent germination compared with the control when exposed to a 90-d stratification. TX RC 44 and TX RC 48 had no statistical evidence of enhanced germination with any of the stratification treatments. Increased stratification to 90 d either reduced percent germination or had no benefit (Table 2). Sulfuric acid treatments of 5 to 15 min had substantial deleterious effects on germination of all three genotypes (data not presented).

None of the three genotypes exhibited a statistically significant increase in percent germination with stratification (Table 2). All three genotypes exhibited statistically significant negative effects from sulfuric acid scarification in this experiment. Our results indicate that cold stratification of up to 90 d failed to improve germination percentage, and sulfuric acid scarification was not beneficial (data not presented, no significant effects). Although this is true for these regional selections, stratification may be necessary for selections from other geographic locations. For this reason, we suggest a 30- to 60-d cold, moist stratification screening for R. columnifera selections. Future studies could examine additional stratifications times between 0 and 60 d to refine recommendations further for individual maternal genotypes. Of perhaps greater need is the investigation of potential variation in chilling requirements for germination across the large geographic range of R. columnifera. Although not promising from the results of our study, sulfuric acid scarification using a more dilute solution or shorter duration could be tested.

Effects of seed maturity on germination.

Developmental stage of the inflorescence and seed location on the inflorescence had a significant effect on percent germination (Table 3). There was a statistically significant two-way interaction for germination percentages among germplasms and developmental stage, and germplasms and region on inflorescence (Table 3). Mid-age and mature developmental stages had increased percent germination for all three germplasms. TX RC 12 had the lowest percent germination out of the three germplasms, except in reference to the mature inflorescences. Basal and middle portions of the inflorescence produced greater germination percentages than the apical portions, except for TX RC 30. Basal and middle portions of the inflorescence produced greater germination percentages than the apical portions, except for TX RC 30. In TX RC 12, the immature and mid-age developmental stages were not statistically different from one another. The mature developmental stage of TX RC 12 germinated more readily than the less mature stages and was comparable in germination percentages to the greater levels observed for TX RC 30 on mid-age and mature sections. There was no statistical evidence of differences in percent germination when looking at location on inflorescences of TX RC 12. TX RC 30 had increased percent germination in the mid-age and mature developmental stages, with the immature stage being significantly less, but greater than similar maturity stages on TX RC 12 or TX RC 8 (Table 3). The basal and mid-inflorescence seed locations outperformed the apical portions of the inflorescences for most stages of development. This was most pronounced on immature inflorescences, which is consistent with the apical portions containing largely immature seeds, whereas those in basal portions of the inflorescence would be more mature based on earlier chronological flowering as the inflorescence matured. Thus, for commercial seed production, harvest should be from fully mature inflorescences to encourage maximum and uniform germination. If collection cannot wait for full inflorescence maturity, then only the basal portions of the inflorescence should be retained.

Developmental stage of the inflorescence and seed location on the inflorescence had a significant effect on percent germination (Table 3). There was a statistically significant interaction effect between the germplasm and developmental stage, and germplasms and region on the inflorescence. In all germplasms, mid-age to mature inflorescences had an increase in percent germination when compared with the immature inflorescences. Because of this, it is recommended to plan seed harvests for when a majority of the inflorescences have dropped their ray petals and turned brown. Two of the three germplasms had significantly higher percent germination when looking at the basal and middle portions of the seed head, with the third germplasm having no differences among locations. This makes sense, because R. columnifera disk florets open first on the basal portion of the inflorescence, and finish opening at the apical portion. This would mean that seed at the basal portion of the inflorescence has had a longer time to mature than apical seed. This phenomena happens in other Asteraceae species as well, such as Artemisia annua L., where among genotypes there were differences ranging from those having capitula with open ray and disk florets, to those with capitula that are tightly enclosed within involucral bracts (Wetzstein et al., 2014). For this reason, knowledge of the timing of flower development in key genotypes, as well as the relative developmental timing of ray and disk florets, are of critical importance for successful crossing (Wetzstein et al., 2014). All of these data imply that R. columnifera is consistent with the more generalized recommendation for wildflowers that seed should be harvested near a stage when they would disperse naturally for optimal germination yields (Hong and Ellis, 1996). The impact of genotype on total germination percentage was evident in our experiment (Table 3) as well as those discussed earlier (Tables 1 and 2). In our experiment, TX RC 30 had the greatest germination percentage on middle and fully mature seeds just as it had in earlier studies, whereas TX RC 12 had comparable germination only on basal portions of fully mature inflorescences, and TX RC 8 had much reduced germination even on basal portions of fully mature inflorescences.