New study could improve monitoring of tropical trees in face of climate change

0
IMAGE

Credit: University of Stirling

Experts have challenged the principle that tropical ecosystems are aseasonal – after discovering regular cycles in fruiting, flowering and leafing in such climates.

Researchers from the University of Stirling made the unexpected observations after studying 30 years of data from 851 tropical trees in the Lopé National Park in Gabon, Central Africa.

The team – led by Stirling PhD researcher Emma Bush – also uncovered the reasons behind inaccuracies in previous monitoring of tropical tree behaviour and made recommendations to enhance future approaches. This will improve monitoring and, in turn, the data that is available for organisations that manage tropical ecosystems and their products, which support livelihoods and the survival of wildlife.

The research is published in a special section of the Biotropica journal, guest edited by Dr Katharine Abernethy, Reader in Tropical Ecology in Stirling's Faculty of Natural Sciences.

Ms Bush's study – entitled Towards effective monitoring of tropical phenology: maximising returns and reducing uncertainty in long-term studies – assessed the Lopé dataset, which is the longest, unbroken record of Central African rainforest behaviour in the world. The data, which includes unique data on phenology patterns, local weather, forest productivity, and animal populations, allows scientists to analyse climate change effects on the landscape.

The academics identified regular cycles in 36 per cent of samples and found flowers are much more likely to occur on a regular, annual cycle than fruits and leaves. The relative unpredictability of fruiting events underscores how complex these tropical ecosystems are, with a lot of variation at individual tree and species level.

They also analysed the processes used to monitor cycles to understand why regular cycling activity is observed in some species and not others.

"We found that, on average, new leaves are easier to spot in the tropical forest canopy than flowers and that, both the visibility of the event and how long it lasts, are really important influences on whether we can detect this behaviour," explained Ms Bush.

"We also found evidence that supports the approach of long-term monitoring to understand these highly complex tropical plant communities. Significantly, we found that the chance of detecting regular cycles doubled when monitoring lasted 20 years, compared to 10."

Dr Abernethy, Professor Nils Bunnefeld and Kathryn Jeffery, Research Fellow, all from Stirling, also collaborated on the research.

Dr Abernethy, Ms Bush and Ms Jeffery are also co-authors on a related paper, entitled Annual cycles dominate reproductive phenology of African tropical trees, also published in the special section, which focuses on plant behaviour – such as fruiting, flowering and leaf fall – in tropical ecosystems.

Dr Abernethy said: "The special section is about cycles of productivity and natural cycles in tropical ecosystems – and how these are changing with the age of trees; seasonal changes related to climate change; and unpredictable climatic effects.

"It's also about how poorly these patterns are understood, even though they are critical for feedback to local weather, for humans using forests – such as for timber or for food – and for animals relying on plants for fruit as food.

"The research underlines how long it takes to understand the behaviour of such diverse systems as tropical forests, or such long-lived organisms as rainforest trees, and how essential it is that this research sector expands."

The University of Stirling oversees the science programme in Lopé National Park for the Gabon National Parks Agency, which collaborated on the Bush et al's paper. The International Medical Research Centre in Franceville, Gabon, were also involved in the work.

###

Media Contact

Greg Christison
[email protected]
01-786-466-687

http://www.stir.ac.uk

Related Journal Article

http://dx.doi.org/10.1111/btp.12543

Leave a Reply

%d bloggers like this: