Lambir liana project

From BiodivBorneo09

Jump to: navigation, search

Correlation Between Tree Bark Characteristics and Liana Growth in Lambir

Andrew Brownjohn, Molly Rooney, & Douwe Yntema



Lianas play an important role in forest dynamics. Attempting to understand growth patterns of lianas can help us better understand forest ecology. The correlation between bark characteristics of trees in the Lambir Hills 52 hectare plot and the number of lianas found on those trees was analyzed. A significant difference was found between the number of lianas found on trees with peeling vs. non-peeling bark. Also, there was a significant difference between the number of lianas found on trees with non-peeling smooth vs. fissured bark. This suggests that we are more likely to find lianas on trees with non-peeling bark.

(See summary of argument)


Lianas, referring to woody climbing plants, are typically identified as pioneer species found in disturbed areas (Whitmore, 1988; Reddy & Parthasarathy, 2003). Lianas play an crucial role in forest dynamics by competing with trees for light and other resources (Londre & Schnitzer, 2006). They often amplify the affects of tree falls, landslides, and other forest disturbances (Perez-Salicrup & de Meijere, 2005). Woody climbers are estimated to represent 10-45% of the woody stem species and may be increasing in abundance due rising atmospheric CO2 levels, increased annual mean temperatures, and elevated levels of human disturbances (Londre & Schnitzer, 2006); lianas are currently estimated to hold 1-3% of the carbon in tropical rainforests (Davies, 2009). The importance of understanding role of lianas in forest dynamics cannot be understated. In order to better understand liana interactions with trees in Lambir Hills National Park, we studied tree bark characteristics and their influence on liana growth.


  • Do differences in host bark characteristics impact the presence of lianas in Lambir Hills National Park, Malaysia?

It has been suggested that early successional tree species have morphological characteristics that prevent liana growth. Considering lianas use tree trunks for support, bark differences may be an important tree characteristic to consider. Certain bark characteristics may be more conducive or hindering to liana growth. Fissures and grooves, for instance, could be used as grips for lianas. Peeling trees might “slough off” lianas when their bark peels. We hypothesize that there is a correlation between bark type and the number of lianas.


To observe whether bark type makes a difference in the growth of lianes on trees, we chose two characteristic types of bark, based on textural differences:

  • Smooth: Tree bark without cracks above 2mm deep. Barks with a presence of pimpling and other slight changes in texture still fall within this category.
  • Fissured: Tree bark specifically characterized by the presence of cracks of 2mm or greater.
    • We also noted whether a tree was peeling (scaling, flaking, or striping included). “Peeling” encompassed any and all bark types that either exhibited bark that was visibly peeling on its own off of the tree trunk, and any that could be sloughed off with minimal effort (to test this, approximately a two second pull or scratch of the bark was used, with all three members of the team coming to agreement over the peeling capability of the tree). With this additional observation, trees were characterized as either smooth and non-peeling, fissure and non-peeling, smooth and peeling, or fissure and peeling.

After standardizing our bark characteristics, we sampled trees within the Lambir 52 hectare plot using a transect method. We chose six 100 x 10 meter transects based on ease of navigation and with the goal of surveying all soil types. Any tree with a diameter at breast height (1.3m) of greater than 30cm was included within the survey. We measured tree diameter by tape when necessary and verified the diameters using the previous year’s Lambir plot data. We then categorized bark type and counted the number of lianas growing on the tree. Any liana clearly using a tree in our sample for support towards the canopy (on any segments of the trunk only) was counted towards a tree’s total liana count. Epiphytic plants of all kinds were not included in the sample, and epiphyte roots were carefully checked against lianas to maintain the integrity of our data. We counted a liana that split into several parts and either grew on different trees with separate ‘limbs’ or encirclied a single tree with multiple limbs as a single, individual liana.

We then tested for interactions between number of lianas and recorded bark types, using the following analytic methods:


First we compared the average number of lianas found on fissured and smooth bark types. Our hypothesis was that the average number of lianas found on smooth and fissured bark is different. The null hypothesis was that there was no difference in the number of lianas found on smooth and fissured bark. We used a Wilcoxon test of the general linear model form to compare the observed liana averages on two bark types.

We also compared the average number of lianas found on peeling and non-peeling trees. We used another Wilcoxon test (general linear model) to compare the observed liana averages of peeling and non-peeling types. The null hypothesis was that there was no difference in the number of lianas found on peeling and non-peeling bark.

Finally we compared the average number of lianas found on smooth and fissured trees which were non-peeling. Again, a Wilcoxon test was used to compare the observed liana averages of smooth and fissured non-peeling types. The null hypothesis was that there was no difference in the number of lianas found on smooth and fissured non-peeling bark.


  • Based on the general linear model, there was no significant difference between smooth and fissured bark. (p=0.344).
  • However, there was a significant difference (p=0.0358) between smooth bark and fissured bark excluding bark that was smooth and peeling or fissured and peeling.
  • There was also a significant difference in the number of lianas found on peeling bark versus non-peeling bark. (p=0.0345).
  • We discovered that the mean number of lianas found on peeling and non-peeling bark was 0.4000 and 0.8537 with standard deviations of 0.5 and 1.5093 respectively. The mean number of lianas found on smooth non-peeling bark and fissured non-peeling bark were 0.6429 and 1.3078 respectively.


The statistics most pertinent to our hypothesis, those dealing with smooth and fissured bark, were found highly insignificant by a general linear model of the Poisson family. In contrast, the peeling characteristic taken as a supplement to ‘bark type’ was shown by the same test to have significance in predicting the presence of lianes. This significance was found both when peeling was compared directly to smooth and fissured bark, as well as versus those grouped simply as ‘non peeling bark.’ Interestingly, a distribution applied only to non-peeling fissured and smooth barks found that fissured barks were more likely to have a liana than those with smooth bark.

The importance of tree bark texture – whether smooth or fissured – is not as significant compared to a bark’s capacity to peel in predicting the presence of lianes. The peeling statistics strongly suggest that a tree, regardless of bark texture, is less likely to be found with a liana growing on it if it possesses a bark that is easily peeled off. Our initial hypothesis was not entirely disproved when only non-peeling bark was taken into account. Our observations about fissured bark being more conducive to liana growth (or conversely, that smooth bark prevents liana growth) had statistical support when data was taken solely within those trees whose bark does not peel.

A statistical advantage such as that comparing trees with peeling barks versus those that do not peel suggests an adaptation that could allow a tree to shed lianes that attempt to climb upon it by simply shedding its bark, and in turn, the liana along with it. More comprehensive studies are needed to verify the conclusions of this paper; future studies should include a larger sample size, a standardized method of bark classification, attention to tree bark changes over time, and tree canopy structure.

Works Cited

  • Londre, R. A. & Schnitzer, S. A. (2006), ‘The distribution of lianas and their change in abundance in temperate forests over the past 45 years’, ECOLOGY 87(12), 2973-2978.
  • Perez-Salicrup, D. & de Meijere, W. (2005), ‘Number of lianas per tree and number of trees climbed by lianas at Los Tuxtlas, Mexico’, BIOTROPICA 37(1), 153-156.
  • Reddy, M. & Parthasarathy, N. (2003), ‘Liana diversity and distribution in four tropical dry evergreen forests on the Coromandel coast of south India’, BIODIVERSITY AND CONSERVATION 12(8), 1609-1627.
  • Whitmore, T. C. (1988), Tropical Rain Forests of the Far East, Oxford University Press, USA.
  • Davies, Stewart. (2009, June 16). ‘Ecology and evolution in CTFS plots’, [Lecture to Harvard Summer School Course: Biodiversity of Borneo.] Sarawak, Malaysia.