Results and Discussion
The average 100-seed weight was 134.1±6.8 g. Freshly collected seeds had a moisture content of 36.2±1.9% and 92.5±0.4% germination (Fig. 1). After 7-day drying, seed germination decreased significantly to 27.5% where moisture content declined to 22.5%. The viability of seeds dried to 20.7% moisture content (seeds dried for 14 days) was 5% and further drying to lower moisture contents resulted in complete loss of viability (Fig. 1). The sharp decline in germination when seeds moisture content reduced to c. 20% f. wb. indicates that the seeds are ‘recalcitrant’
Compared with other desiccation-sensitive seeds of Fagaceae species, e.g. Q. robur (Finchsavage, 1992), Q. coccifera L.,Q. pubescens Willd. and Quercus pedunculiflora K. Koch (Ganatsas and Tsakaldimi, 2013), C. sclerophylla seeds had a similar desiccation response, showing that the percent germination of seeds decreased with the progress of drying. The seeds lost their viability completely when the moisture content of the seeds dropped lower than 20% in 2-3 weeks. The result is also consistent with drying rates reported on Quercus seeds, where these studies reported that the drying rate of Fagaceae seeds is relatively slow. Further, it required 15-25 days to reach moisture levels that are lethal to seeds when dried in silica gel. However, Tian and Tang (2010) reported thatC. fissa and Q. fabri seeds can be dried to approximately 20% moisture content within 72 hours, which is contradicting with our results. This is not surprising, as we found in our earlier work (Li et al., 2018) thatQ. fabri seeds required a longer drying period of at least 20 days to reach approximately 20% when dried above silica gel, which is in agreement with Xia et al. (2012b). The contradicting result on the drying rate could be attributed to the differences in the drying method. Both whether or not the seeds were dried in a sealed environment and the amount of silica gel used was not provided by Tian and Tang (2010).
In this study, we show that X-ray computed tomography (CT) can be a useful tool to study the internal structures of seeds without altering the intact structure of seeds. CT results show that the pericarp ofC. sclerophylla is similar to other Fagaceae seeds, such asQuercus (Xia et al., 2012b). Pericarp in Quercus is divided into three areas: (1) the apex; (2) the scar; and (3) main pericarp (the remaining area between the apex and the scar), and our results indicate that these regions are clearly distinguishable inC. sclerophylla (Fig. 2.3a). Xia et al. (2012b) found that water loss during seeds desiccation of Quercus species occurs only through the scar instead of main pericarp or the apex in some species including Q. nuttallii, Q. suber and Q. palustris . Xia et al. (2012a) showed that the moisture content of Q. franchetiionly reached 23% after 164 days of drying in silica gel. For this species, a comparatively small scar (ca. 6% of the pericarp area) might be the mechanistic basis of limited water loss through the pericarp (Xia et al., 2012b). For Q. suber , Eduardo and Belén (2000) studied the morphological and micro-morphological differences of the scar and main pericarp. The main pericarp provides a protective barrier in slowing down water loss and it is mainly contained of a layer of epidermal wax, a thick cuticle covered epidermal layer and a single layer of palisade cells. However, the scar lacks this protective structure, which is comprised by vascular strand presumably facilitating moisture loss. Thus, this feature was reported to have ecological advantage, because the embryos of Quercus seeds cannot reach lower water content during brief exposure to drought conditions in natural environment.
Whilst investigating the route of water loss in C. schlerophylla might be arguably an important area worth exploring, we believe the results of photomicrographs and CT scan suggested otherwise. Clearly, in C. sclerophylla , embryo is on the opposite side of the apex and sits just below the scar region (Fig. 2). This is in marked contrast withQuercus embryo, which is located on the apex side and far away from scar (Finch-Savage and Blake, 1994; Xia et al., 2012b). Thus, during desiccation no matter water loss only occurs through scar or intact pericarp, embryo of C. sclerophylla seed will be the first part within the seeds to lose water due to embryo position and continuous drying may kill the seeds, so there is no practical significance to study the water loss pathway during seed drying. However, our drying results indicate that C. sclerophylla seeds lose viability only when the moisture content drops to c. 20% f.wb.
There appears to be strong difference in the germination pattern betweenQuercus and C. sclerophylla seeds. Notably, in C. sclerophylla seeds, embryo elongates and grew thicker initially before protrude through the scar, then elongated into root and radicle (see Fig. 2.3). For Quercus seeds, the radicle protrudes through the apex region during germination (Bonner and Vozzo, 1987; Löf et al., 2019). It is possible to suggest that this difference could be due to the position of embryo between these two genera. However, this standpoint requires detailed investigation given that seeds of manyQuercus have epicotyl physiological dormancy (Allen and Farmer Jr, 1977; Baskin and Baskin, 2014; Farmer Jr, 1977).