Plants at risk: a study for a sustainable management

Posted on

Pterocarpus angolensis is a deciduous tree species that grows in southern and eastern Africa over a wide range of localities and environmental conditions. It is threatened by overharvesting due to its valuable timber (Blood wood, Kiaat) and by land use changes.
Species distribution models (SDM) could provide more accurate information on distribution and environmental requirements and thereby assist sustainable management of this tree species.

In a new study recently published on Forest Ecology and Management, a team of scientists (among them, CMCC researcher Antonio Trabucco from IAFENT Division) tried to estimate the potential, realised and future distribution of P. angolensis. A distribution map with good discrimination of Pterocarpus angolensis was obtained that highlights how its occurrence is mainly influenced by the amount of summer rainfall, by the minimum temperature in winter and by temperature seasonality. Moreover, climate change can decrease the species range considerably, especially in the west, threatening species existence in Namibia and Botswana.

The abstract of the paper:
The deciduous tree species Pterocarpus angolensis occurs in the dry woodlands of southern Africa and grows under a broad range of environmental conditions. It is threatened by overharvesting due to its valuable timber (Blood wood, Kiaat) and by land use changes. Information on the most suitable environmental conditions for the species is often old and anecdotal, while available data on its occurrence refer to range extent and not to distribution. Species distribution models (SDM) could provide more accurate information on distribution and environmental requirements and thereby assist sustainable management of this tree species.

Maxent models were developed to estimate the potential, realised and future distribution of P. angolensis and to identify detailed environmental requirements. Occurrences data of the species were sourced from herbaria and other published sources; environmental data from global GIS databases. Relevant environmental predictors were selected through a jack-knife test of the first model runs. The addition of information on competing species, fires and deforestation was tested to determine realised distribution. Model quality was evaluated with an independent presence-absence dataset. The model was projected with two different climate change scenarios to study their effect on the distribution by 2080.

Results show that a potential distribution map can be obtained with good discrimination of the presence of the species (AUC 0.83) and fairly good calibration (correlation coefficient 0.61). Range extent and environmental requirements are more detailed than those described in literature. The distribution of the species is mainly influenced by the amount of summer rainfall, by the minimum temperature in winter and by temperature seasonality. Potential and realised distributions are very similar, with Madagascar as major exception where the species can grow but does not occur. Adding the fire history of the last 13 years or the distribution maps of potentially competing species as predictor variables did not improve the distribution model. It did illustrate that P. angolensis is mainly found in areas with annual fire frequency below 45% and that only a few of the tested species show signs of competition. Using a forest cover map improved the realised distribution slightly (Kappa coefficient 0.64). Climate change can decrease the species range considerably, especially in the west, threatening species existence in Namibia and Botswana. On the other hand, the species’ occurrence is predicted to increase in Zambia.

Read the integral version of the paper:

De Cauwer  V. , Muys B., Revermann  R. , Trabucco A.
Potential, realised, future distribution and environmental suitability for Pterocarpus angolensis DC in southern Africa
2014, Forest Ecology and Management, Volume 315, Pages 211–226, DOI: 10.1016/j.foreco.2013.12.032

Start typing and press Enter to search

Shopping Cart