Relative light germination

Seeds of all three batches failed to germinate in the dark and reached 96% (B2) or 98% (B1, B3) germination in 12 h d⁻¹ light. RLG is therefore 1 in all cases.

The triggering of seed germination in cacti by light, that is, positive photoblastism, is well known and is associated with a range of plant traits including shorter stature and lighter seeds. C. repandus is cylindrical and has relatively light seeds (3.2 mg). Other species in the tribe Cereeae also have light seeds, and a RLG averaging 0.89 (n = 12 species) (Yang, Reference Yang1999; Seal et al., Reference Seal, Flores, Ceroni Stuva, Dávila Aranda, León-Lobos, Ortega-Baes, Galíndez, Aparicio-González, Castro Cepero, Daws, Eason, Flores Ortiz, del Fueyo, Olwell, Ordonez, Peñalosa Castro, Quintanar Zúñiga, Ramírez Bullón, Rojas-Aréchiga, Rosas, Sandoval, Stuppy, Ulian, Vázquez Medrano, Walter, Way and Pritchard2008; Flores et al., Reference Flores, Jurado, Chapa-Vargas, Ceroni-Stuva, Dávila-Aranda, Galíndez, Gurvich, León-Lobos, Ordóñez, Ortega-Baes, Ramírez-Bullón, Sandoval, Seal, Ullian and Pritchard2011). Small (0.62–0.77 mg) seeds of Trichocereus terscheckii also germinate in the light but not in the dark (Ortega-Baes and Rojas-Aréchigo, Reference Ortega-Baes and Rojas-Aréchigo2007). These species thus shows a preference for germination close to the soil surface, as light only penetrates to a depth of a few millimetres (Tester and Morris, Reference Tester and Morris1987). The relationship between light preference for germination and smaller seeds has been reported previously for European herbs from a variety of habitats (Millberg et al., Reference Millberg, Anderson and Thompson2000; Jankowska-Blaszczuk and Daws, Reference Jankowska-Blaszczuk and Daws2007), pioneer trees from semi-deciduous tropical forest and cacti, with around 60% of a broad range of seed lots among 136 taxa having RLG of 1 (Seal et al., Reference Seal, Flores, Ceroni Stuva, Dávila Aranda, León-Lobos, Ortega-Baes, Galíndez, Aparicio-González, Castro Cepero, Daws, Eason, Flores Ortiz, del Fueyo, Olwell, Ordonez, Peñalosa Castro, Quintanar Zúñiga, Ramírez Bullón, Rojas-Aréchiga, Rosas, Sandoval, Stuppy, Ulian, Vázquez Medrano, Walter, Way and Pritchard2008; Flores et al., Reference Flores, Jurado, Chapa-Vargas, Ceroni-Stuva, Dávila-Aranda, Galíndez, Gurvich, León-Lobos, Ordóñez, Ortega-Baes, Ramírez-Bullón, Sandoval, Seal, Ullian and Pritchard2011). However, less is known for light-responsive cactus seeds about the spectral requirements for the trigger, which has implications for understanding of the importance of nurse plants, and shading, in the natural regeneration process in cacti (Ordoñez-Salanueva et al., Reference Ordoñez-Salanueva, Seal, Pritchard, Orozco-Segovia, Canales-Martínez and Flores Ortiz2015). This specific requirement for light might also serve to stop the seed from germinating within the developing fruit before endozoochorous dispersal (Flores et al., Reference Flores, Jurado, Chapa-Vargas, Ceroni-Stuva, Dávila-Aranda, Galíndez, Gurvich, León-Lobos, Ordóñez, Ortega-Baes, Ramírez-Bullón, Sandoval, Seal, Ullian and Pritchard2011), although this is not the case in all cacti as vivipary has been reported for at least 53 species (Cota-Sánchez, Reference Cota-Sánchez2004; Cota-Sánchez et al., Reference Cota-Sánchez, Reyes-Olivas and Abreu2011; Aragón-Gastélum et al., Reference Aragón-Gastélum, Flores, Yáñez-Espinosa, Reyes-Olivas, Rodas-Ortiz, Robles-Díaz and González2017).

Onset of RED light sensitivity is related to the seed imbibition period

The imbibition rates for B2 and B3 seeds were very similar, with MCs increasing rapidly to around 25% fresh weight (FW) basis by 24 h and remaining around 30% from 2 to >8 d. (Fig. 1). During water uptake, the sensitivity of seeds to RED light increased. Few seeds responded to RED when applied around 10% MC, about half responded at an average seed MC of 15%, and at 25% MC, all seeds capable of germination responded to RED. Thus, as the seeds were hydrated in the transitional region of sorption zones II and III, they became sensitive to light-stimulated germination, similar to the heat shock response of Carica papaya seeds (Wood et al., Reference Wood, Pritchard and Amritphale2000). What was noticeable was the difference in RED light sensitivity among seed batches with B2 requiring 100 min RED to achieve around 60% germination compared with only 25 min for B3 to be saturated with RED and achieve higher germination (around 90%) (Fig. 1A).

Germination is dependent on RED light

With the sufficient dose of RED light seed germination in all three batches of C. repandus reached 88–96% (Fig. 2). However, the single dose required to promote a certain level of germination varied between batches, indicating different initial competencies for germination. While B2 required a dose of around 100 min RED to attain around 50% germination, the other two batches needed <10 min RED. Thus, the requirement for RED light for the germination of B3 and B1 seed was more than one order of magnitude lower than B2. Pre-storage time could have impacted the germination response, as B1 had been stored the longest (2 years) and had the highest initial germination level without RED treatment (Fig. 2, closed circles). However, parallel RED dose-response lines for the three batches suggest that the variation in pre-storage time did not affect the overall sensitivity of seeds to increasing RED (Fig. 2).

Fig. 2. Effects of RED light dose (log min) on seed germination (probability scale) for C. repandus B1 (⬤), B2 (■), B3 (▴). Probit germination increased by 1.571 ± 0.059 per log dose (min) for each seed batch, which had zero dose germination levels. Further explanation for fitting parallel lines is given in the text. Binomial confidence limits (95%) are shown for the total responders based on two replicates of 25 seeds sown per light dose.

Intra-specific variability in seed lot quality with respect to the seed proportion that easily germinates is well known, and is often a function of the maternal environment during development, for herbaceous (Baskin and Baskin, Reference Baskin and Baskin2014) and tree (Daws et al., Reference Daws, Lydall, Chmielarz, Leprince, Matthews, Thanos and Pritchard2004) seeds. In addition, B2 seed was stored for the shortest period of time (1 year) prior to experimental use and post-harvest handling can affect the depth of any primary dormancy. In the case of Stenocereus queretaroensis seeds, storage in the dark at 25°C and 42% RH increased germination competence to 85% after around 15 months (De la Barrera and Nobel, Reference De la Barrera and Nobel2003), similar to that of B1 which had been stored dry for 2 years before use.

A linear relationship was observed between seed germination (on a probit scale) and exposure time to RED for the seed batches B1, B2 and B3 (Fig. 2). Fitting parallel lines did not significantly affect the goodness of fit compared to free-fitting [F = 1.22; tabulated F(2, 28)  3.34; r ² = 0.850], such that probit germination increased by 1.571 ± 0.059 per log dose (min). This suggests that although the initial competency levels for germination were different (intercepts 4.266 ± 0.101, 1.715 ± 0.124 and 3.752 ± 0.116 probit for B1, B2 and B3, respectively), the seed sensitivity (i.e. the distribution of responses per dose) to the RED dose was the same across seed batches of this species. Thus, germination capability increased by 1 probit per 0.64 log min of RED dose in all three seed batches.

High sensitivity to increasing the RED light level has been observed in the cactus S. queretaroensis, for which seed germination approximately doubled per log dose (De la Barrera and Nobel, Reference De la Barrera and Nobel2003). As C. repandus seed germination increased about fourfold per log dose (Fig. 2), it appears that these seeds are relatively more sensitive to RED light than S. queretaroensis. The fine-tuning of the seed response to light is modulated by PIF (phytochrome[phy]-interacting factor) transcription factors, which are pivotal in cellular signalling and promote shade-avoidance through regulated gene expression (Leivar and Quail, Reference Leivar and Quail2011). When phy molecules are activated by light, the rapid degradation of PIF proteins is induced. Other pathways converge to regulate PIF, however, including the gibberellin pathway which contributes to the lowering of seed physiological dormancy. In Arabidopsis, for example, PIF1 promotes the expression of two gibberellic acid (GA)-repressor (DELA) genes that decrease the sensitivity of the seed to GA (Leivar and Quail, Reference Leivar and Quail2011).

Dark incubation reduces light sensitivity

The long-term retention of light-sensitive seeds in the dark often results in the desensitization to light (or the induction of secondary dormancy; Bewley and Black, 1994). To see if this was also the case for C. repandus, B2 seeds were left in the dark for greater than the usual period (3 d prior to RED). Two conditions of RED stimulation, 100 min and 1000 min, were applied and dark incubation times of up to 18 d were used.

Lengthened dark imbibition decreased light sensitivity significantly (Fig. 3). For example, germination was reduced from around 92 to 20% when dark incubation time increased from 2 to 18 d prior to irradiation with 1000 min RED. When seeds were only exposed to RED for 100 min, germination fell from 52% after the shortest dark incubation time of 2 d to 20% after the seeds were incubated in the dark for 8 d. A negative linear relationship between probit germination and dark incubation time was observed for both RED doses. GLIM analysis showed an insignificant difference between the parallel model and free-fitting models (F = 0.79; tabulated F(1,10)  4.96; r ² = 0.863), such that probit germination = 7.331–0.242x (1000 min RED) and probit germination = 5.709–0.242x (100 min RED). The reduced sensitivity of C. repandus seeds to RED light after long-term (1–2 weeks) exposure to darkness suggests that buried seeds could enter the soil seed bank (Fig. 3).

Fig. 3. Effect of dark incubation time prior to RED light irradiation on the germination (probability scale) of C. repandus B2. Seeds were imbibed for up to 18 d prior to the application of RED light for 1000 min (■) or 100 min (▫). Binomial confidence limits (95%) are shown for the total responders based on two replicates of 25 seeds sown. The fitted parallel lines indicate a one probit reduction in germination per 0.242 dose of dark imbibition, that is, every 4.1 d.

The effect of multiple, short doses of RED light and long periods of darkness per day on seed germination was also assessed. B3 seeds were rehydrated in the dark for 3 d and exposed to 1 or 5 min of RED per day, cumulatively for 20 min, with germination proceeding in the dark. When 1 min of RED was provided daily, germination increased to around 40% by 15 d, but there was no further increase by the 20th day (Fig. 4). Germination promotion did not reach the full level for this seed lot. A germination increase was greater when 5-min daily exposures to RED were used. A single 5-min treatment on the first day resulted in 48% germination, with the trajectory for germination promotion similar to that of a 1-min daily dose (Fig. 4). However, four daily 5-min treatments (i.e. 20 min cumulatively) raised germination to 92%, close to the maximum for this seed lot (Fig. 4).

Fig. 4. Effect of cumulative exposure to RED light on the germination of B3 seeds. Seeds were imbibed in the dark for 3 d and then RED light applied for 1 min (▴) and 5 min (▵) per day to total exposure time of 20 min. Thus, the experiments ran for 20 and 4 d, respectively. Binomial confidence limits (95%) are shown for the total responders based on two replicates of 25 seeds sown.

A single dose of 100 min of RED light are sufficient to saturate the light requirement in B3 seeds, which then germinate to a high level in the dark (Fig. 2). The seeds are clearly highly sensitive to RED light, as the dose effect can be reduced to just 20 min at 5 min a day for 4 d. However, shortening the exposure to RED light to 1 min, a day only partially promotes subsequent germination (Fig. 4). Nonetheless, the seeds show signs of a low fluence response (Cone et al., Reference Cone, Jaspers and Kendrick1985; De Petter et al., Reference De Petter, van Wiemeersch, Rehty, Dedonder, Fredericq and De Greef1988), similar to that of S. queretaroensis seeds that respond over a 10-d period to a minimum light level of 0.15 μmol m⁻² s⁻¹ for 10 h d⁻¹ (De la Barrera and Noble, Reference De la Barrera and Nobel2003). Such sensitivity to high-quality light acts as a gap-sensing mechanism that also enables the rapid emergence of the seedling (Daws et al., Reference Daws, Burslem, Crabtree, Kirkman, Mullins and Dalling2002). When the daily RED dose is insufficient to stimulate all seeds to germinate (i.e. <5 min), the long periods of darkness between RED applications are presumed to be sufficient to permit the reversion of the active form of phytochrome (P _FR) to the inactive form (P _R), and some of the seeds fail to germinate potentially because of the induction of secondary dormancy (Cone et al., Reference Cone, Jaspers and Kendrick1985; De Petter et al., Reference De Petter, van Wiemeersch, Rehty, Dedonder, Fredericq and De Greef1988). Similarly, seeds of eight out of 11 species of cacti did not germinate to a significantly higher level when the same set of seeds was transferred from dark to light, suggesting that darkness had triggered secondary dormancy (Flores et al., Reference Flores, Jurado and Arredondo2006). The molecular framework for such rapid and time-lapse responses to light could be modulated by the interaction between de-etiolated 1 (DET1), which is an evolutionary conserved protein and central repressor of light-induced germination, constitutive photomorphogenic 10 (COP10), long hypocotyl in FR 1 (HFR1) and PIF1 (Shi et al., Reference Shi, Wang, Mo, Tang, Zhong and Deng2015).

As postulated for S. queretaroensis (De la Barrera and Nobel, Reference De la Barrera and Nobel2003), only with erosion of the soil surface or tilling would C. repandus seeds likely accumulate sufficient light to germinate. This response has the advantage that the seeds could await rainfall in a habitat, in which rainfall events are far apart. Eventual emergence therefore depends on the seeds having the ability to survive. Seeds used in this study, which were stored for up to 2 years before experimentation, nonetheless had initial (on receipt) germination competency of at least 96%, indicating, as with S. queretaroensis (De la Barrera and Nobel, Reference De la Barrera and Nobel2003), that the seeds have good longevity. Nonetheless, it would be interesting to know precisely how temperature and humidity affect seed longevity in C. repandus.

RED light effect is reversible by FR light

The involvement of phytochrome in the RED light response was confirmed by reversing the effect with FR light using seed lots B2 and B3.

For B2 seeds, the greater the dose of FR given to seeds immediately after 3-d hydration followed by exposure to RED for 1000 min, the less effect RED had on germination. While control (1000 min RED) seeds reached 80% germination, irradiation with FR for 1.5 min (0.18 log min), 15 min (1.176 log min) and 150 min (2.18 log min) reduced the germination to 48, 34 and 26%, respectively. GLIM analysis revealed a significant (r ² = 0.986, P < 0.05) dependency of (decreased) germination on the FR of −0.488 ((±0.215) probit per log min, suggesting that FR was very effective at increasing the requirement for RED. Numerous PIFs are known to repress light responses in plants, including PIF8 (Oh et al., Reference Oh, Park, Song, Bae and Choi2020). PIF8 protein accumulates more in Arabidopsis in FR light than in darkness or RED light through a mechanism that includes the inhibition of COP1 by phyA (Oh et al., Reference Oh, Park, Song, Bae and Choi2020). In seeds, the FR light response depends on phyA being transported from the cytoplasm to the nucleus by shuttle proteins FAR RED ELONGATED HYPOCOTYL 1 (FHY1) and FHY1-like (FHL) (Sheerin and Hiltbrunner, Reference Sheerin and Hiltbrunner2017). In the nucleus, light-activated phytochromes suppress the E3 ubiquitin ligase complex CUL4/DDB1^COP1/SPSA and inactivate PIF, although additional mechanisms must exist that shift the action of phyA to FR (Sheerin and Hiltbrunner, Reference Sheerin and Hiltbrunner2017). For example, in Arabidopsis seeds, SPATULA gene expression is repressed under FR light in a PIF1-dependent manner (Vaistij et al., Reference Vaistij, Barros-Galvao, Cole, Gilday, He, Li, Harvey, Larson and Graham2018).

As FR clearly desensitized the seeds to RED light, we next investigated how quickly seeds could escape the effect of FR completely (Fig. 5). B3 seeds were hydrated for 3 d in the dark and then given 60 min RED light, which in the control promoted germination to >90% and then subjected seeds to 10 min FR light up to 2 d afterwards. The application of FR within 10 min of RED was sufficient to reduce germination to <10% (Fig. 5). As the time between the RED and FR pulse increased the reversibility of RED by FR was reduced, such that applying FR 24 h after RED failed to stop germination in 60% of the seeds. However, when 48 h were allowed to elapse between RED and FR, FR failed completely to stop the progress of germination; the seeds reached 100% (data point not shown in Fig. 5). The dependency of germination on escape time since RED illumination was described for B3 seed as: Germination (probability)  3.41 (±0.019) + 0.078 (±0.011) dose, in hours (r ² = 0.945, P < 0.01) (Fig. 5).

Fig. 5. Effect of FR light on C. repandus B3 seed germination (probability scale). After rehydration for 3 d, seeds were exposed to RED for 60 min and, then after a range of escape times (hours), exposed to FR light for 10 min. Responsivity to FR light decreased by one probit of germination per 0.078 dose, that is, 12.8 h. Binomial confidence limits (95%) are shown for the total responders based on two replicates of 25 seeds sown. A data point at 48 h and 100% germination is not shown in the graph but was included in the probit analysis, by convention, as 99.9% germination.

Presumably, C. repandus seed burial or deep shading by a nurse plant cumulatively reduces seed sensitivity to light and supports the formation of a soil seed bank, as is relatively common in small seeds of many cacti, for example, 0.6 mg seeds of P. chende (Flores et al., Reference Flores, Jurado, Chapa-Vargas, Ceroni-Stuva, Dávila-Aranda, Galíndez, Gurvich, León-Lobos, Ordóñez, Ortega-Baes, Ramírez-Bullón, Sandoval, Seal, Ullian and Pritchard2011; Ordoñez-Salanueva et al., Reference Ordoñez-Salanueva, Orozco-Segovia, Canales-Martinez, Seal, Pritchard and Flores-Ortiz2017). However, if the seeds receive sufficient RED light and time to initiate the germination process before burial, the stimulatory effects of RED light are irreversible by FR light. This time interval is evidently around 1–2 d (Fig. 5). At 26°C in the light (12 h d⁻¹), C. repandus seed germination is visible within 3 d (Yang, Reference Yang1999), and germination processes prior to radicle emergence must be far advanced by 2 d, as the seeds will have been at full hydration from 24 h onwards (Fig. 1). The commitment to complete germination quickly is a feature of small seeds of desert plants in order to maximize the likelihood of seedling establishment (Baskin and Baskin, Reference Baskin and Baskin2014). This temporal control of the interplay between the effects of RED and FR light on cactus seed germination involving phytochrome might also be temperature-dependent. The biologically active Pfr state can convert to Pr in a light-dependent thermal relaxation process called dark reversion (Mancinelli, Reference Mancinelli1994). As phyB inactivation is proportional to temperature in the dark, phytochromes potentially also function as thermal timers that integrate temperature information overnight (Jung et al., Reference Jung, Domijan, Klose, Biswas, Ezer, Gao, Khattak, Box, Charoensawan, Cortijo, Kumar, Grant, Locke, Shafer, Jaeger and Wigge2016). Although much is known about how temperature influences cactus seed germination (Yang, Reference Yang1999; Seal et al., Reference Seal, Daws, Flores, Ortega-Baes, Galíndez, León-Lobos, Sandoval, Ceroni Stuva, Ramírez Bullón, Dávila-Aranda, Ordoñez-Salanueva, Yáñez-Espinosa, Ulian, Amosso, Zubani, Torres Bilbao and Pritchard2017; Barrios et al., Reference Barrios, Sanchez, Flores and Jurado2020), the possible modulating and concurrent role of phytochromes in both the thermal and photon responses of cactus seeds appears to be unexplored.