3 Results
3.1 Variations in leaf functional
traits
The intraspecific variation in majority of the leaf functional traits
was moderate, with a coefficient of variation (CV) ranging from
0.10–0.30 (Fig. 1). The variation in the structural traits was
relatively less, CV ranging from 0.02–0.55. Among the chemical traits,
LC exhibited the least variation,
while LP (CV = 0.07–0.83) and
LN:LP (CV = 0.09–0.62) showed the greatest variations (Fig. 1).
In contrast, leaf functional traits varied significantly among the
species (p < 0.05).Pinus massoniana had the
highest LT and the lowest LA and SLA. Alangium kurzii andQuercus fabri had higher LA, SLA and LN but lower LDMC and LC:LN.
The LN of the evergreen species was significantly lower than that of the
deciduous and coniferous species. Deciduous species had higher LA, SLA,
LN, and LP values than those of both evergreen and coniferous species.
Coniferous species had the lowest LA and SLA (p < 0.05)
but the highest LT (Fig. 1).
At the community level, the variations in most of the traits was
moderate (CV = 0.15–0.36), except LCC,
LNC, and LC:LNC, where the CV
< 0.10 (Table 2). Among the structural characters, the
greatest variation was seen in LAC (CV = 0.36) ranging
from 7.57–57.37 cm2. Among the chemical traits, LP
varied greatly, especially LN:LPC was 20.49 and the
values ranged from 10.39–46.35 (CV = 0.24) (Table 2).
3.2 Phylogenetic signals and correlations of leaf functional
traits
The functional dendrogram (Fig. 2) showed that the leaf functional
traits of the 18 coexisting woody species were phylogenetically
structured, and species of the same genus clustered and had similar
functional traits. Significant phylogenetic signals were detected only
for LT and LA (K > 1, p < 0.01)
among different species. The SLA showed a moderate phylogenetic signal
(K = 0.53, p < 0.05). However, the LDMC and
chemical traits (K = 0.05–0.40) were randomly distributed in the
phylogeny because there were no significant phylogenetic signals
(p < 0.05, Table 3).
At the species level, irrespective of phylogenetic influence, there were
significant negative correlations between SLA with LT, LDMC, and LC:LN;
and LN with LT. There was a clear positive relationship between SLA with
LA, LN; and LN, LP with LA respectively (Table 4). Interestingly, after
removing the influence of phylogeny, the correlations between SLA with
LT, LN; and LN with LA, LC; and LC with LT, LDMC; and LP with LC, LN
were significantly enhanced (Table 4). At the community level,
non-phylogenetic correlations based on CWMs of the nine leaf traits were
largely consistent with these results (Tables 5). However, after
controlling for phylogenetic variables, most of the relationships showed
a little weakening trend, and only the correlations between
SLAC with LCC; and LPCwith LAC, LTC increased slightly (Table
5).
3.3 Primary environmental factors affecting leaf functional
traits
At the community level, 23.19% of the variation in leaf functional
traits was explained by environmental factors. TK was a unique
significant edaphic factor (R 2 = 17.83%), and
altitude and aspect were the two significant topographical factors
explaining 9.58% and 8.94% of the variation, respectively (Fig. 3a).
For phylogenetic compositions, all factors showed significant effects
except slope, which together explained 38.56% of the total variation.
Among them, the most important environmental factors were TP, altitude,
and convexity, each with an interpretation rate of over 20% (Fig. 3b).