Effect of leaves damaged by Dendrolimus punctatus and insect frass on soil priming effect.

Insect herbivory in the forest canopy leads to a large amount of damaged leaves and frass input to soil, with consequence on soil carbon cycle. However, the influence of damaged leaves and frass from insect canopy herbivory on the soil priming effect is unclear. We examined the effects of leaf litter, leaf damage caused by Dendrolimus punctatus, and insect frass on soil priming effect by using the 13C natural abundance technique. The results showed that the addition of leaf litter, damaged leaves, and frass significantly increased native soil organic carbon mineralization, producing a positive priming effect. Moreover, significant differences were observed among treatments. The accumulative priming effect induced by frass was the largest, followed by damaged leaves, and that of leaf litter was the smallest. The priming effect was positively correlated with total P, condensed tannin, total phenolic content, and the ratio of condensed tannin to P (condensed tannin/P), and negatively correlated with C/N, lignin/N, C/P, and lignin/P in the early stage of incubation. There was a significant negative correlation between the priming effect and lignin content in the later stage of incubation. Our results indicated that damaged leaves and frass increased the magnitude of positive priming effect, which was influenced by different factors at different incubation stages. Our results would strengthen the understanding in the effects of insect herbivory on soil carbon cycling in forests, and improve the accuracy of the assessment of its effects on forest carbon sink.

 Sinoseneciominshanicus (Asteraceae, Senecioneae), a new species from south-eastern Gansu and northern Sichuan, China.

Sinoseneciominshanicus (Asteraceae, Senecioneae), a new species from south-eastern Gansu (Wenxian and Zhugqu counties) and northern Sichuan (Pingwu county), China, is described and illustrated. This species is similar to S.rotundifolius, a species locally endemic to Songpan county in northern Sichuan, in having a scapigerous habit, orbicular leaves and solitary capitula, but differs by the presence (vs. absence) of stolons and by having thinner rhizomes (ca. 2 mm vs. more than 5 mm in diameter), stems proximally sparsely fulvous arachnoid or glabrescent (vs. densely sericeous-villous) and obscure (vs. conspicuous) main veins on adaxial surface of leaves. The chromosome number of the new species is reported to be 2n = 60. Colour photographs of living plants in the wild and a distribution map are provided for the new species and S.rotundifolius. The geographical distribution of S.rotundifolius is also corrected, with the previous record of this species from south-eastern Gansu (Wenxian county) actually referring to S.minshanicus.

Sinoseneciopingwuensis (Asteraceae, Senecioneae), a new species from northern Sichuan, China.

Sinoseneciopingwuensis (Asteraceae, Senecioneae), a new species from Pingwu county in northern Sichuan, China, is described and illustrated. This species is distinguished in Sinosenecio by having leathery, glabrous, ovate or ovate-oblong leaves often pinnately-veined and solitary capitula 2.3-4.3 cm in diameter, a unique character combination hitherto never recorded in the genus. Two floral micromorphological characters (configuration of filament collar of stamens and anther endothecial cell wall thickenings) and achene surface features of the new species are reported. Color photographs of living plants and a distribution map are also provided for the new species.

[Application of indicator species in predicting forest management effect on forest species diversity and community composition].

By using indicator species analysis (ISA) method, this paper studied the feasibility of using indicator species to reflect the responses of species diversity and community composition of subtropical forests in Huitong of China to forest management. Ninety-four significant indicator species from 357 understory species were identified, and a new indicator species dataset (community level) was constructed to examine the association between indicator species dataset and original community dataset, and to evaluate the predictive potential of indicator species in reflecting forest management effect. There existed a strong association between the two datasets (Mantel r = 0.898). The indicator species dataset could well predict the management effect on species diversity (regression analysis, R2 > 0.74) and community composition (ANOVA, F >16.79). When the two datasets were applied to Nonmetric Multi-Dimensional Scaling (NMDS) ordination and K-mean cluster analysis, the indicator species dataset could well identify the forest types with different management treatments, as the original community dataset did. Also, the indicator species dataset nearly played the same role as the original community dataset in identifying the species diversity, community composition, and forest type. It was suggested that for saving costs in overall investigation of forest ecosystem, indicator species could be used as a surrogate of full community to predict forest management effect.