A survey of C. dalatensis and C. capitata was undertaken using the snowball sampling approach with 15 respondents (5 forest managers, 5 ethnic people living near the forest boundary, and 5 forest planters). Before completing the interview, all respondents were freely consented.

The interview questionnaire had three questions: (1) Have you ever seen these two plant species?; (2) Where did you see these two plant species?; and (3) How many people know about these two species at this location? Following the interviews, we established 5 survey transects per species, each ranging in length from 2 to 3 km, to locate mature trees.

The program http://geocat.kew.org/editor (Gardens, 2024) and the IUCN Red List Criteria and Classification Guidelines (Standards & Subcommittee, 2022)( http://www.iucnredlist.org/documents/ RedListGuidelines.pdf) were used to calculate the Extent of Occurrence (EOO) and Area of Occupancy (AOO).

The experiments were conducted in a randomized block design with three replications. The data measured were survival percentage (%), rooting percentage (%), number of roots, root length (cm), and rooting index (number of roots multiplied by root length). The data were analyzed using the Duncan test in SPSS 26.0.0 software (Statistical Package for Social Sciences version 26).

Camellia dalatensis has a narrow distribution in Phat Chi Village, Tram Hanh Commune, Da Lat City, Lam Dong Province. The number of mature individuals in the natural habitat was 205 plants. The Extent of Occurrence (EOO: 0.464 km2) and Area of Occupancy (AOO: 8,000 km2) for Critically Endangered (CR) C. dalatensis are

illustrated in Figure 1(upper panel). Camellia capitata occurs in Cat Tien National Park, an area managed by Lam Dong Province (Cat Tien District). There were 5 mature plants in the natural habitat. C. capitata is categorized as Critically Endangered (CR) with EOO of 0.050 km2 and AOO of 4,000 km2 (Figure 1; lower panel). Both C. dalatensis and C. capitata were shown to have a very small distribution area (Figure 2 ), which were threatened by logging, deforestation, planting of industrial crops, such as cashew, rubber and coffee, and the harvesting of non-timber forest products (Beech et al., 2017); (Orel et al., 2014). Recently, the forests are being exploited for medicinal herbs and ornamental plants with high polyphenol contents and beautiful flowers, including these two Camellia species (Figure 3) (Trinh, 2022).

Asexual propagation by using cuttings is the optimal solution for preserving genetic characteristics from mature plants when propagation by using seeds is limited by low fecundity. There are many factors affecting rooting, including the environment, the types of cuttings used for propagation, and the use of plant growth regulators (PGRs) ((Bhupathireddy et al., 2022); (Dewi & Sabhara, 2022); (Eed & Burgoyne, 2014); (, 2022); (Griffin et al., 1998); (Ray & Ali, 2017)). Cutting propagation efficiency is often increased by using exogenous PGRs, including IBA treatments (Bhupathireddy et al., 2022);(, 2022). However, the optimum concentrations and PGR forms (solution or solid form) can differ markedly between species. For example, 1,000 ppm IBA in solution gave the best results in Lycium barbarum L., but 1% IBA was optimal for Cornus mas L. (Çelik & Çetin 2021). In this study, 1,000 ppm IBA was suitable for C. dalatensis, but C. capitata required a higher concentration of 1,500 ppm IBA for obtaining the highest rooting percentage.

Differences in PGR requirements for the rooting of cuttings not only vary between plant species, but also depend on the position in the plant where the cuttings are sourced. Cuttings taken from mature trees often have reduced endogenous auxin content compared to cuttings taken from juvenile plants. Hence, it is necessary to undertake experiments on cuttings responses toward PGR dosages to ensure that optimal PGR dosages are being used for the correct cutting type. Also, at very high concentrations, the rooting efficiency can be significantly reduced, such as in Prunus laurocerasus L. at 2 g/L IBA (Sulusoglu & Cavusoglu, 2010). Thus, depending on the type of tree and the age of the tree, the rooting auxin treatment needs to be adjusted to the appropriate concentration, and it is necessary to have specific studies for each tree species.

The effectiveness of using IBA in propagation of threatened Camellia species are shown in Tables 1 and 2. There were significant (P < 0.05) effects of IBA rates on survival percentage, rooting percentage, number of roots, root length, and rooting index.

For C. dalatensis, the survival percentage ranged from 65.56% to 80%; the rooting percentage ranged from 14.45% to 45.56%; the number of roots ranged from 0.57 to 3.80; the root length ranged from 1.67 cm to 4.23 cm; and the rooting index ranged from 1.12 to 16.32. The most effective treatment was 1,000 ppm IBA. For C. capitata, the survival percentage ranged from 28.89% to 40.00%; the rooting percentage ranged from 7.78% to 31.11%; the number of roots ranged from 0.30 to 1.90; the root length ranged from 0.40 cm to 2.13 cm; and the rooting index ranged from 0.15 to 4.09. The most effective treatment was 1,500 ppm IBA.

For C. dalatensis and C. capitata, the liquid growth regulator was more effective compared to the charcoal powder in promoting rooting. Results on survival percentage and rooting percentage are compared in (Figure 4).

Liquid IBA solutions are easy to prepare, while the phloem and xylem of cuttings have better access to exogenous hormones than hormones combined with powders. However, powder containing activated charcoal is an effective adsorbent for many pollutants (organic, inorganic, microbial, and biological) (Mohammad-Khah & Ansari, 2009). Therefore, when the activated charcoal- contained powder is dry-mixed with a plant growth regulator, the dry mix can reduce infection. Therefore, in this research, we sought to identify the most effective treatment, whether the activated charcoal-contained powder or liquid plant growth regulator.

Cutting types are divided into softwood (non- lignified) cuttings, semi-hardwood (partly lignified) cuttings, and hardwood (lignified portion) cuttings (Griffin et al., 1998); (Islam et al., 2010). Most propagators of woody plants use semi-hardwood cuttings for higher rooting efficiency compared to softwood and hardwood cuttings (Alkaç et al., 2022).

In each cutting type, the cutting position, whether terminal shoot or stem cuttings, also has a great impact on rooting efficiency. In most species, cuttings from terminal shoot have higher rooting efficiency. However, the use of appropriate plant growth regulators can affect the rooting efficiency of stem cuttings and ensure propagation efficiency with the number of rooted cuttings higher than that of the terminal shoot cuttings (Solikin, 2019). Choosing the right cuttings for propagation and increasing the efficiency of propagation are necessary solutions.

For some Camellia spp., it is common to choose semi-hardwood cuttings because this cutting type provides the best results. In this study, semi- hardwood cutting was used and divided into terminal shoot cuttings and stem cuttings (Tables 3 and 4).

Based on Experiment 1, the IBA concentrations used were 1,000 ppm for C. dalatensis and 1,500 ppm for C. capitata. The parameters of survival percentage, rooting percentage, number of roots, root length, and rooting index were significantly different for C. dalatensis (P < 0.05). Terminal shoot cuttings provided better results compared to stem cuttings with the survival percentage of 90.00% compared to 67.78%, the rooting percentage reaching 58.89% compared to 31.11%, 5.10 roots per cutting compared to 1.97 roots per cutting, the root length reached 4.50 cm compared to 2.33 cm, the rooting index reached 23.08 compared to 4.62. Similar results have been reported by Wazir (2014) in propagating C. japonica by using terminal cuttings with 1,000 ppm IBA where the highest percentage of rooting was 84.96%, and the longest root length was 18.3 cm. For C. capitata, the parameters of survival percentage, rooting percentage, and number of roots per cutting were significantly different (P < 0.05). However, root length and rooting index were not statistically significantly different (P > 0.05). The survival percentage of terminal shoot cuttings compared to stem cuttings was 45.56% compared to 37.78%. However, the rooting index for terminal shoot cuttings was about 9% lower than that of stem cuttings (25.56% versus 34.45%), and the number of roots per cutting was 1.80 compared to 1.50. The root length ranged from 1.87 cm to 2.00 cm (average 1.93 cm) and the rooting index range was

3.00 to 3.36 (average 3.18) for both cutting types.

Comparing the rooting parameters, the effective cutting type for C. capitata propagation was semi- hardwood stem cuttings. At the time of gathering cuttings in the field, the C. capitata had begun to sprout. Therefore, the terminal shoot cuttings were very immature and weak, so they were easy to be infected by fungi. Besides, we collected the stem cuttings at the same time that the cuttings were partially lignifying, to have better auxin synthesis; thereby, the rooting percentage of stem cuttings was higher than that of terminal shoot cuttings in C. capitata.

The growing substrate environment should have good porosity to support smooth water draining and provide a suitable temperature and humidity regimes for root development. The substrate media can be combined with other factors, such as plant growth regulators and cutting types, to bring the highest efficiency to the propagation process (Johnson et al., 2005)(Silva et al., 2012).

The selected growing substrate for propagation in this study was a mixture of sand and coir dust. Sand ensures good drainage, and coir dust improves water retention in the rooting environment. In this experiment, 1,000 ppm IBA was used for semi- hardwood cuttings of C. dalatensis, while 1,500 ppm IBA was used for semi-hardwood cuttings of C. capitata. Rooting results are shown in Table 5(C. dalatensis), Table 6(C. capitata), and Figure 5.

For C. dalatensis, there was no statistically significant differences in survival percentage (P > 0.05) between treatments, with a mean value of 95.28%. However, rooting percentage, number of roots, root length, and rooting index were significantly different (P < 0.05). Rooting percentage ranged from 75.56% to 88.89% (average 82.22), number of roots per cutting ranged from 4.93 to 6.78 (average 5.76), root length ranged from 4.60 cm to 6.23 cm (average 5.44 cm), and rooting index ranged from 22.73 to 42.88 (average 31.82).

The treatment with the best results for C. dalatensis was 25% sand:75% coir dust, which may be related to the wet environment where this species occurs. C. dalatensis was distributed in Dalat City which has high humidity and high annual rainfall. For C. capitata, there was no significant difference between the survival and rooting percentages (P > 0.05), the average survival percentage was 65.56% and the average rooting percentage was 51.39%. The number of roots per cutting ranged from 2.03 to 2.83 (average 2.50), the root length ranged from 1.77 cm to 2.87 cm (average 5.44 cm) and the rooting index ranged from 4.54 to 8.12 (mean 5.53). Treatment with the best results for C. capitata was the treatment with equal parts of sand and coir dust. C. capitata occurs in Cat Tien District which has lower humidity and rainfall compared to that in Dalat City. Therefore, this species was able to grow better in the medium which had similar conditions to their natural habitat.

The effects of media types and composition on rooting efficiency have been reported in many plants (Eed & Burgoyne, 2014). Air content and oxygen diffusion rate in media are important for rooting (Ercisli et al., 2002). The characteristics of the species also influence rooting success. In particular, Camellia impressinervis grows well in high moisture and well drained soil (Tran, 2018), in which their cuttings had the best rooting results in a growth medium of 100% sand with high air content and oxygen diffusion capacity (, 2020).