*3.6. NormFinder Analysis*

To further determine the reliability of the results obtained by the geNorm algorithm, the NormFinder application was used to evaluate both the intra- and inter-group variation to calculate the stability of the candidate reference genes, with lower values of inter- and intra-group variation corresponding to increased stability of the candidate gene [48]. The results of the NormFinder analysis are shown in Figure 5. *UBC28* (0.055) and *helicase-15* (0.138) were the most stable genes in seedlings, while *UBC28* (0.063) and *UBC7* (0.123) were the most stable genes in adult trees (Figure 5A,B). The best combination of reference genes for different tissues was *helicase-15* (0.134) and *ubiquinone* (0.267) (Figure 5C). In the combined leaf developmental stages set, the optimal combination of candidate reference genes was *UBC7* (0.190) and *ACT7* (0.245), while over the entire growth cycle, the most stable candidate genes were *ubiqunone* (0.267) and *UBC36* (0.295) (Figure 5D,E). *TCTP* (0.030) and *ubiqunone* (0.035) were the most stable genes under cold treatment for 7 d, while *GAPDH* (0.062) and *UBC36* (0.088) were the most stable genes under cold treatment for 24 h (Figure 5F,G). Including both cold stress treatments, the candidate reference genes *EF2* (0.168) and *UBC36* (0.193) were the most stable (Figure 5H). Under heat treatment for 24 h, *PAB2* (0.089) and *CYP20-2* (0.414) were the most stable reference genes (Figure 5I). Under all temperature stresses, *PAB2* (0.185) and *EF2* (0.349) showed the highest stability (Figure 5J). The best combination of reference genes to compare all sample sets was *ubiquinone* (0.375), *EF2* (0.408), *PAB2* (0.413), and *GAPDH* (0.483) (Figure 5K). Notably, *TUA* and *PPR* were the least stable genes in most sets, similar to the results calculated by geNorm. In general, the results of the NormFinder analysis for the 14 candidate reference genes under different experimental conditions were similar to the results of the ΔCt and geNorm analyses.

**Figure 3.** Ranking of expression stability of the 14 candidate reference genes in *L. megaphylla* using geNorm analysis. Results from (**A**) different seedling tissues; (**B**) different adult tree tissues; (**C**) all seedling and adult tree tissues; (**D**) 16 leaf developmental stages; (**E**) entire growth cycle including all different seedling and adult tree tissues as well as 16 leaf developmental stages; (**F**) cold treatment for 7 days; (**G**) cold treatment for 24 h; (**H**) cold treatments including 7 days and 24 h; (**I**) heat treatment for 24 h; (**J**) different temperature treatments including cold and heat; (**K**) total samples.

**Figure 4.** Pairwise variation values calculated by geNorm were used to determine the optimal number of reference genes from the 14 candidates in *L. megaphylla*. The cutoff value of Vn/n+1 to determine the optimal number of reference genes for RT-qPCR normalization was 0.15.
