Study area

We collected seeds of the selected shrub species in the ‘‘Santuario del Bosque de Niebla’’ (SBN) reserve, located in Xalapa, Veracruz, Mexico (19°30'047.80´´N, 96°56´15.93´´W; 1250 m a.s.l.) (Fig. 1). The SBN is a fragment of TMCF (30 ha) that presents different successional stages including open areas, secondary vegetation, and remnants of old-growth forest (^{Atondo-Bueno et al., 2018}). During the first half of the last century, parts of the SBN were occupied as shade to coffee plantation, where coffee and citrus replaced the understory vegetation. Since 1975, however, the area was protected, and forest recovery took place through elimination of coffee and citrus and allowing natural succession (^{Atondo-Bueno et al., 2018}). The local climate is temperate humid with an average annual temperature of 18.7 °C and mean annual rainfall of 1671 mm, the soil is classified as Andesite (^{Williams-Linera et al., 2013}). The vegetation cover is dominated by Quercus xalapensis Bonpl., Q. germana Schtdl. & Cham., Carpinus tropicalis (Donn. Sm.) Lundell, Clethra macrophylla M. Martens & Galeotti, Liquidambar styraciflua L., and Turpinia insignis Tul. (^{Williams-Linera et al., 2013}).

Figure 1: Map of the study area. The polygon is the Santuario del Bosque de Niebla reserve, located in Xalapa, Veracruz, Mexico. 

Shrub species

We selected eight shrub species typical of TMCF and of ecological importance in these communities, given their high abundance and potential role in ecological restoration (cover, density, frequency, and sociability), availability of fruits, and location in forest clearings (^{Gelviz-Gelvez et al., 2020}). The species meeting these criteria were: Conostegia xalapensis D. Don., Hoffmannia excelsa K. Schum., Hybanthus glaber (Dowell) Standl., Miconia desmantha Benth., M. mexicana Naudin, Moussonia deppeana (Schltdl. & Cham.) Klotzsch ex Hanst., Hamelia patens Jacq. and Piper lapathifolium Steud. (Table 1).

Experimental design

We visited the SBN twice a month between May and June 2020 to collect mature fruits from eight native shrub species (Table 1). Overall, five fruits were collected from five healthy individuals per species; in the case of Hybanthus glaber, we collected 50 fruits per individual and transported them to the laboratory in labeled paper bags. To determine the seed extraction method, harvested fruits were classified as dehiscent and fleshy (^{Sierra-Escobar et al., 2020}). Thus, for dehiscent fruits, we allowed the natural release of seeds at room temperature, whereas for fleshy fruits, we cleaned the seeds in a container with water. After seed cleaning, viable seeds were selected according to the flotation method and submerged for 24 h (^{Varela and Arana, 2010}). Seeds obtained were stored in paper bags at room temperature for later use in germination tests.

We applied a pre-germination treatment to seeds, after a week of seed collection, by imbibing them for 12 h in tap water. Seeds of the eight selected species were placed in Petri dishes with three layers of filter paper (Whatman No. 1) as a substrate for germination. The experiment consisted of a control treatment (seeds with no pre-germination treatment) and an imbibed seed treatment. Ten replicates were conducted, with 30 seeds per petri dish, for a total of 300 seeds per species per treatment (600 seeds per species in both treatments) for a total of 1200 seeds per species in both germination conditions, and germination was compared between greenhouse (commercial greenhouse with a plastic cover to filter sun light) and germination chamber (Thermo Fisher Scientific, Model TFFU2065FWA, Marietta, USA) conditions.

Under greenhouse conditions, seeds were subjected to a fluctuating temperature of 15±1.85 to 48±3.25 °C and a photoperiod of 10 h of light and 14 h of darkness, while in the germination chamber, temperature was constant at 25 °C under a photoperiod of 12 h of light and 12 h of darkness. We assessed germination when the emergence of the radicle was >0.5 mm (^{Bandara et al., 2019}). The number of germinated seeds was recorded daily for 60 days, starting the day after setting up the experiment. This procedure involved using entomological tweezers and a stereoscopic microscope (Leica EZ4HD, Heerbrugg, Swiss) for precise observation and measurement.

Data analysis

We calculated the total percentage of germination for each species under each condition, the number of days for the start of germination, the number of days to reach 50% germination (t₅₀), and the germination rate for each species using the “germinationmetrics” package (^{Aravind et al., 2023}). Additionally, for each species we constructed germination accumulation curves.

We used analysis of variance (ANOVA) to evaluate differences in the germination of seeds with pre-germination treatment (imbibition) compared to control treatment, and a second ANOVA to evaluate differences in germination between both conditions (i.e., greenhouse and germination chamber). We ran a Tukey HSD test to detect differences between pairs of means. Data was transformed using arcsine to achieve normality and homogeneity (^{Burgos-Hernández et al., 2014}). Finally, cumulative germination curves of each species were used to compare the effect of temperature and light on seed germination using non-parametric Kaplan-Meier analysis. Data analysis was performed using the statistical program R v. 4.4.2 (^{R Core Team, 2022}).

Beginning of germination and t₅₀

The number of days for germination initiation differed between the greenhouse and germination chamber for all species (F _1,318 =18.74, P<0.001; Table 2). Overall, germination initiation time in the germination chamber was 12±1.17 days (mean ± standard error). Species that started germination faster after sowing in the germination chamber were M. deppeana (6 days) and H. patens (6 days), for seeds with and without treatment, while untreated seeds of H. glaber and treated seeds of P. lapathifolium needed more days to germinate (16 and 18 days, respectively). On the other hand, the start of germination of all species in the greenhouse was slower, with an average of 33 ± 3.33 days (Table 2). Hybanthus glaber started germination faster (27 days); otherwise, the treated seeds of C. xalapensis required the greatest number of days to germinate (45 days). Under both conditions (i.e., germination chamber and greenhouse), initiation of germination in all species was similar in seeds with and without treatment (F _1,158 =0.002, P=0.965).

In the germination chamber, some species reached t₅₀ a few days after sowing. For example, treated and untreated seeds of H. patens (7±0.1 days) and seeds with treatment of M. deppeana (10±0 days), whereas seeds without treatment of C. xalapensis and M. mexicana, needed more days after planting to reach t₅₀ (33±1.03 and 32±1.26 days, respectively; Table 3). Overall, the time to reach t₅₀ for seeds placed in the greenhouse was notably extended. For example, treated seeds of C. xalapensis reached t₅₀ at 51±8.58 days after sowing. Similarly, untreated seeds of M. desmantha reached t₅₀ at 37±5.30 days after sowing (Table 3). No differences were found in t₅₀ between greenhouse and germination chamber (F _1,318 =0.138, P=0.71).

Effect of environmental parameters

Temperature and light influenced the germination percentages obtained for all species. The temperature and light conditions provided by the germination chamber demonstrated efficacy in influencing both the onset of germination and the time required to achieve 50% germination (Temperature, (²=11230, P<0.0001; Light, (²=10290, P<0.0001). For example, seeds of Moussonia deppeana, Hamelia patens, and Hoffmannia excelsa germinated quickly, reaching the highest number of germinated seeds (germination peak) between 5 and 15 days (Fig. 3); in contrast, in the greenhouse, seeds of the same species germinated at a longer period and in less quantity, with germination peaks between 35 and 40 days (Fig. 4).

Figure 4: Cumulative germination curves (%) of eight shrub species collected at ‘‘Santuario del Bosque de Niebla’’ reserve, located in Xalapa, Veracruz, Mexico, by treatment (control, imbibition) and in each germination condition (i.e., germination chamber, greenhouse).  

Germination percentages

As expected, our results indicate that the eight shrub species evaluated exhibited improved germination response under the constant conditions of 25 °C provided by the germination chamber. The only exception was Hybanthus glaber (Violaceae), which displayed similar germination response in both the germination chamber and the greenhouse, with a small increase in seed germination in the greenhouse where temperature conditions ranged from 15 to 48 °C. Some previous studies have observed similar germination responses in species of the same botanical family. For instance, ^{Augspurguer (1979}) noted that seeds of Hybanthus prunifolius (Humb. & Bonpl.) Schuzle-Menz germinate favorably in conditions of low humidity and respond well to fluctuating temperatures. Additionally, ^{Godefroid and Van de Vyver (2020}) observed that seeds of Viola calaminaria Lej. (Violaceae) respond better to high temperatures and fluctuating temperatures. Conversely, all species exhibited low germination percentages in the greenhouse, with Miconia mexicana and Piper lapathifolium showing no germination.

Temperature is a key factor in the seed germination process, as it stimulates enzymatic activity and nutrient mobilization (^{Suárez and Melgarejo, 2010}). Therefore, all species have optimal temperatures for high germination rates (^{Funes et al., 2009}). For instance, Hamelia patens (99%), Hoffmannia excelsa (99%), and Moussonia deppeana (84%) reached high germination percentages in the germination chamber. This positive response in germination may also be related to the fact that all three species have seeds with soft testa (^{Jurado; 2013}, ^{Aguilar-Morales et al., 2022}) and short dormancy (^{Figueroa et al., 1996}). Germination studies for H. patens and H. excelsa (Rubiaceae) are scarce, no studies were found on germination under similar conditions evaluated in the present study and studies of H. patens have focused on asexual propagation (^{Elgimabi, 2009}; ^{Odusanya et al., 2019}). However, germination studies have been conducted under nursery conditions on Hamelia axillaris Sw., where seeds were placed in peat pots on the surface of sifted soil, and 68% of germinated seeds were observed (^{Garwood, 1983}). In the case of M. deppeana, our results are consistent with those obtained by ^{Brechú-Franco et al. (2018)} and ^{Jurado (2013)} who obtained germination of 93% for this species under laboratory conditions, both at a temperature of 25 °C and under different photoperiods, 16 h of light and 8 of darkness, and 12 h of light and 12 h of darkness, respectively.

Species of the Melastomataceae family, such as M. desmantha (77 %), M. mexicana (75 %) and C. xalapensis (63 %), showed high germination percentages at constant temperature of 25 °C. Similarly, some studies reported a better germinative response between 25 and 30 °C of some species in the Melastomataceae (^{Silveira et al., 2013}; ^{Escobar and Cardoso, 2015}; ^{Fernández-Sánchez et al., 2020}). The results obtained for M. mexicana (75 %) are consistent with those reported by ^{Pérez-Cadavid et al. (2018}), who obtained 64 % of germination under controlled conditions, 25 °C during the day and 22 °C at night and a photoperiod of 12 h of light and 12 h of darkness. Also, ^{Godoi and Takaki (2007)} obtained 94% of germination under conditions of 12 h of light and 12 h of darkness and 25 °C temperature for Miconia theizans (Bonpl.) Cong. Otherwise, our results for C. xalapensis (63%) differ from those obtained by ^{Sánchez et al. (2015)}, who observed 9% of germination in the same controlled conditions. Also, ^{Pérez-Cadavid et al. (2018)} obtained a maximum germination for C. xalapensis of 33%, under controlled conditions, 25 °C during the day and 22 °C at night and a photoperiod of 12 h light and 12 h darkness. Primary dormancy has been recorded in some species of Melastomataceae (^{Silveira et al., 2013}), which might promote the presence of inhibitors in the testa (^{Oliveira-Silveira et al., 2012}), a factor that may have influenced observed germination. On the other hand, the observed low germination percentages of H. patens, H. excelsa and C. xalapensis and the null germination in Miconia mexicana and Piper lapathifolium in greenhouse conditions might be related to the increase of the temperature, above the germination optimum, resulting in a delay or inhibit germination in some species (^{Silva et al., 2007}; ^{Silveira et al., 2013}; ^{Martínez and De la Barrera, 2020}). To our knowledge this is the first study that reports germination percentages for M. desmantha, which is abundant in forest clearings.

Contrary to our expectations, the imbibition treatment did not increase seed germination in any of the eight shrub species studied. Our results showed that the germination of all species could be determined by other factors, presumably light and temperature, and not by the presence or absence of pre-germinative treatment. Some species, such as H. patens and C. xalapensis, showed a slight increase in germination percentages in seeds under pre-germination treatment in the germination chamber. However, this increase was not significant, while M. deppeana and H. excelsa, higher percentages were recorded in seeds without treatment.

Seed germination and t₅₀

The initiation of germination was significantly higher in the germination chamber than in the greenhouse for all the species. For example, H. patens, M. deppeana, and H. excelsa started germinating a few days after planting, at 5, 6, and 8 days, respectively. In addition, it was observed that H. patens reached 50% of germinated seeds (t₅₀) 7 days after sowing, M. deppeana reached t₅₀ at 10 days, and H. excelsa reached t₅₀ at 15 days. The maximum germination for M. deppeana was reached in 16 days; this finding differs from those obtained by ^{Jurado (2013)} and ^{Brechú-Franco et al. (2018)}, who observed that M. deppeana began germination at 15 days, reaching the maximum germination at 45 and 60 days after planting, respectively. A soft testa is reported for H. patens, M. deppeana, and H. excelsa, which might allow the absorption of water and contributed to rapid germination (Jurado, 2013; ^{Aguilar-Morales et al., 2022}).

The initiation of germination was slightly slower for H. glaber and P. lapathifolium; however, the results show a favorable response for both species. H. glaber started at 16 days and reached t₅₀ at 23 days after planting. Some studies for species of Violaceae mention that they can take up to 41 days to reach maximum germination (^{Augspurguer, 1979}; ^{Godefroid and Van de Vyver, 2020}). We also observed the same pattern for the maximum germination of H. glaber, which was reached at 30 days on average. Specifically, P. lapathifolium required 16 days for germination initiation and 26 days to reach t₅₀, a pattern consistent with the findings reported by ^{Silva et al. (2007)}. The mean germination time for Piper aduncum L. at a constant temperature of 25 °C was 18 days.

The Melastomataceae was the best represented botanical family in this study with three species: M. desmantha, M. mexicana and C. xalapensis. Of the three species, C. xalapensis began germination in less time, 8 days after planting and reached 50% of germination at 29 days, while M. mexicana began germination at 15 days and reached t₅₀ at 34 days. There are studies that suggest that in species of the Melastomataceae, total germination can occur between 24 and 71 days (^{Pérez-Cadavid et al., 2018}). In addition, it has been documented that some species of this botanical family have a hard testa which could delay the germination process (^{Oliveira-Silveira et al., 2012}; ^{Silveira et al., 2013}). We acknowledge that the dominance of these species may be a consequence of the sampling method, since we sampled shrubs in forest gaps where these species are abundant.

Overall, in greenhouse conditions we observed a delayed germination for all species, this trend could be explained by the high temperatures observed in this condition (thermo inhibition; ^{Hills and van Staden, 2003}). For example, C. xalapensis started germination at 44 days after planting and reached t₅₀ at 49 days. Hoffmannia excelsa showed a similar response, initiating germination at 31 days and obtaining t₅₀ at 40 days after planting.

Effect of environmental parameters on germination

Our results confirmed that the imbibition treatment had little influence on the germination of the eight shrub species. In contrast, constant light and temperature of 25 °C provided by the germination chamber were favorable for the germination of most species; however, light variations and in particular alternating temperatures of 15 to 48 °C negatively affected germination.

These findings are consistent with previous studies on other species that occupy clearings and have seeds of similar characteristics, such as Trema orientale f. micrantha Kuntze (^{Montejo et al., 2002}), and Guazuma ulmifolia Lam. (^{Muñoz et al., 2004}). For these two species, satisfactory germination was observed in thermoperiods of 25 to 35 °C. However, as the temperature increased to 40 °C, the germination percentages decreased, particularly for G. ulmifolia, which lost 47 % viability of its seeds. This indicates that this temperature approaches the upper limit for germination of this species. Despite the high temperatures recorded in the greenhouse, M. deppeana obtained considerable germination percentages, which suggests that this species is tolerant to significant temperature fluctuations. This might be an advantage for this species to thrive in extreme environmental conditions (^{Baskin and Baskin, 2014}).

The results of this study showed an important difference in the germination response between both germination conditions of the eight species evaluated. The optimal conditions for germination for most species were found to be constant light and temperature provided by the germination chamber, as this was associated with high rates of germination and generally rapid seedling development. On the other hand, the conditions of the greenhouse, including the light and alternating temperatures ranging from 15 to 48 °C, were unfavorable for most species, leading to low rates of germination and significant delays in the process. However, M. deppeana showed significant germination under these conditions as well as of H. excelsa and M. desmantha showed high germination percentages, which might suggest that these species could be suitable candidates for restoration interventions in cloud forest in central Veracruz. Furthermore, the greenhouse conditions hindered the germination of P. lapathifolium and M. mexicana.

Finally, the imbibition treatment did not significantly improve the germination response of the species in either of the two germination conditions. Specifically, the treatment had no influence on the percentage of germination, the beginning of germination, or the number of days required to reach 50% germination. These results are consistent to those reported by ^{Viveros et al. (2015)}, who observed a slight increase in the germination of imbibed seeds of Enterolobium cyclocarpum (Jacq.) Griseb. To further investigate the germination response of these shrub species, it might be necessary to conduct field-based germination experiments that consider factors such as water availability, soil moisture, and the effects of varying light intensities. Additionally, examining the relationship between seeds and seedlings with nurse plants might provide valuable insights into factors (i.e., seed quality, seed dormancy, soil condition) that influence seedling survival and growth, which can be used to improve seedling establishment techniques in restoration projects (^{Chaneton et al., 2010}). Understanding the seed germination process can provide important information regarding the specific environmental conditions required for the successful propagation of the studied species. This knowledge can inform conservation efforts by allowing managers to create strategies for optimal seed storage and handling and improve germination rates in both natural and artificial settings. Furthermore, a thorough understanding of germination requirements can guide the development of more effective restoration techniques, potentially increasing the success rates of reintroduction programs and habitat rehabilitation projects (^{Gelviz-Gelvez et al., 2020}; ^{Toledo-Aceves et al., 2022}; ^{Martínez-Calderón et al., 2023}).