Liana diversity and density in Ghana show positive association with precipitation, precisely opposite to pantropical trends |

Bismark Ofosu-Bamfo, University of Energy and Natural Resources in Sunyani, Ghana, discusses his article: Patterns and drivers of liana community structure across five forest ecosystem types in Ghana

In recent years, ecologists in the tropics have dedicated a good amount of research to understand the pattern and drivers of community assemblages of woody climbing plants, from local to global scales. This understanding is important because woody climbing plants, also known as lianas, are very common in tropical forests and their assemblage patterns and drivers may provide insights into the factors that govern assembly of tropical plant species. A key challenge to developing a unified narrative on patterns and drivers of species assembly lies in the limited available data from many important ecological zones. When available, it is also necessary to determine how well those patterns reflect local and subregional settings. To learn more, we studied liana assembly patterns in Ghana in relation to climate, soil, and forest structure.

The forests in Ghana are primarily distributed along a rainfall gradient with variation in forest physiognomy in different forest types. This allowed us to test how well the current understanding of liana community ecology applies locally to Ghanaian forest ecosystems (e.g., if lianas are more abundant in seasonally dry forests than in aseasonal or wet forests, and more abundant in disturbed than undisturbed forests). We did this by conducting an inventory of lianas and trees, and collecting soil samples in seventy-five 40 x 40 m plots evenly distributed in five forest ecosystem types in Ghana, ranging from dry semideciduous forests to wet evergreen forests. We also measured canopy openness, light intensity, slope, and elevation in plots, and obtained climate data at the forest level.

We observed interesting outcomes regarding the drivers of liana species richness, abundance, and basal area. For example, tree species richness supported liana species richness, abundance, and basal area, but not tree abundance. Rainfall and liana species richness, liana abundance relative to tree abundance, and basal area also showed a positive association, in contrast to previous global analyses, such as Schnitzer (2005) and DeWalt et al. (2014).

There were also distinct liana species compositions across various forest ecosystems, highlighting the unique assemblages present within each forest ecosystem type. Indeed, this was clearly demonstrated in the distribution of species along various environmental gradients as shown in the canonical correspondence analysis of our study.

These findings will be useful to resource managers who need to make silvicultural decisions that include liana management or control. Managers can decide on how much effort should be dedicated to a particular ecosystem relative to others based on the outcome of our study. Local businesses and individuals who harvest specific non-timber forest products for medical, fibre, and construction will also be better informed as to the ecosystem types their liana of interest will prefer most and is likely to be found.