Previous research found that phylogenetic clustering increased with disturbance for tropical trees, suggesting that community assembly is mainly influenced by abiotic factors during early succession. Lianas are an important additional component of tropical forests, but their phylogenetic community structure has never been investigated. Unlike tropical trees, liana abundance is often high in disturbed forests and diversity can peak in old secondary forest. Therefore, phylogenetic structure along a disturbance gradient might also differ from tropical tree communities.

Together with Prof. Ferry Slik and other colleagues, Dr. Mareike Roeder of Xishuangbanna Tropical Botanical Garden (XTBG) conducted a study in a montane forest in a montane forest in permanent plots Mengsong township, Xishuangbanna, Yunnan, China (21°28–34°N, 100°26–31°E).They asked whether the phylogenetic structure of liana communities varied along a disturbance gradient defined by forest structure and whether the phylogenetic structure of liana communities correlated with environmental variables such as soil characteristics or topography. They then asked how incorporating all optimal solutions of the phylogeny affected the interpretation of community phylogenetic analyses. They finally asked whether different phylogenetic community structure measures produced corresponding outcomes and whether DNA-barcodes assist in liana identification for field studies.

The researchers used the net relatedness index (NRI) from 100 equivalent phylogenies with varying branch length that were constructed using DNA-barcode sequences to determine phylogenetic community structure of lianas. Over 3000 individuals (apparent genets) and over 3800 ramets (individuals and their clones) were measured and identified. They considered three additional phylogenetic indices for comparison.  They also checked if NRI changed when it was weighted by liana ramet or individual abundance: both NRIs were highly correlated (r < 0.98, p < 0.001), justifying the choice of just using ramet density.

When NRI was used as index, phylogenetic clustering of liana communities decreased with decreasing tree basal area, suggesting that liana competitive interactions dominated during early succession. Liana communities in mature forests, on the other hand, were phylogenetic clustered, which could be caused by dispersal limitation and/or environmental filtering.

The study entitled “Phylogenetic clustering increases with succession for lianas in a Chinese tropical montane rain forest” has been published in Ecography.