Updated: Feb 7, 2020
Ecosystem services are the benefits that humans get from nature.
When there is not much available information on ecosystem services, like in many developing countries, models can be a solution by providing additional information.
It is important to check that models are reliable so they can be used by decision-makers.
We looked at five ecosystem services (carbon, water, firewood, charcoal, grazing and water use) in sub-Saharan Africa.
The models were checked using data from 16.7 million km^2 across 36 countries.
We found that models for carbon and water were better than models for firewood, charcoal, grazing and water use. This is because it is harder to make reliable models of ecosystem service use than its physical supply.
We have shown that it is possible to check the reliability of ecosystem service models in areas where only small amounts of data are available. Ecosystem models can now help meet the information demand from policy-makers in sub-Saharan Africa.
Ecosystem services are the benefits that humans get from nature. Ecosystem services are divided into four categories depending on the types of benefits they provide. These are; supporting services (e.g. pollination), provisioning services (e.g. water supply), regulating services (e.g. breakdown of waste), and cultural services (e.g. recreation).
Human activities that damage the environment can also reduce the availability of ecosystem services. To continue benefiting from nature, many governments are trying to manage their natural resources (i.e. forests) in a more ecosystem-service-focused way. To do this effectively, governments need to know where all the ecosystem services are found.
When there is not much available information on ecosystem services, like in many developing countries, computer generated models can help by providing additional information. There have not been many scientific studies on how to check how realistic these ecosystem service models are, which means they are sometimes overlooked by policy- and decision-makers.
The suitability of ecosystem service models is usually lowest in the areas they are needed most – developing countries. To try and tackle this, we looked at multiple ecosystem service models across sub-Saharan Africa to see how realistic they were.
What we did
We looked at five ecosystem services that are important in sub-Saharan Africa (carbon, water, firewood, charcoal, and grazing). The potential supply of an ecosystem service relates to the availability of a natural resource in the environment. In our study we looked at the potential supply of carbon and water, and modelled these using existing models. The realised supply of ecosystem services takes into account not just the availability of the resource, but also its accessibility to humans. For example, in this study we looked at the realised supply of firewood, charcoal, grazing and water. Models for these realised services were built using the output of existing models alongside human population density.
The models chosen for testing had to be able to estimate some of the potential ecosystem services (carbon and water) as well as provide an input to the new models of realised ecosystem services (firewood, charcoal, grazing and water). The models had to include a range of complexity and we had to have enough data to test the model.
The models were tested using data from 16.7 million km2 across 36 countries in sub-Saharan Africa. This included 1675 data points from 16 independent data sets (carbon: 214, water: 736, firewood: 285, charcoal: 59, grazing: 401).
How well each model performed was worked out by looking at how close the model values were to the real values in each area. We also tested whether more complex models were better and how well human density on its own predicts the use of ecosystem service in that area.
What we found
We found that the existing models provide good predictions for the potential ecosystem services (carbon and water).
When human population density was the only thing considered, the results for charcoal, firewood and grazing ecosystem services were just as good as 45/47 of the new models tested. In the other 2/47 cases, human population density alone was a better predictor than the models. However, the new models for water use were better than human population density on its own.
When we looked at how the complexity of the models affected model performance, we found that more complex models either performed the same or better than less complex models.
What this means
Our study is the first multi-country assessment of multiple ecosystem service models. We have shown that it is possible to assess ecosystem service models in areas where only small amounts of additional data are available.
Our results are particularly relevant to sub-Saharan Africa as we have run and tested models for ecosystem important ecosystem services (carbon, water, firewood, charcoal, grazing and water use) using data from 36 of its countries.
Ecosystem models can now help meet the information demand from policy-makers in sub-Saharan Africa.
Read the original article
Willcock, S., Hooftman, D. A. P., Balbi, S., Blanchard, R., Dawson, T. P., O'Farrell, P. J., Hickler, T., Hudson, M. D., Lindeskog, M., Martinez-Lopez, J., Mulligan, M., Reyers, B., Shackleton, C., Sitas, N., Villa, F., Watts, S. M., Eigenbrod, F., Bullock, J. M. (2019). A continental-scale validation of ecosystem service models. Ecosystems. 22(164): 1-16.