The African Savanna Cyclone Experiment (CyclEx)

CyclEx will test the theory that savannas and dry forests are alternative stable states by examining the long term impacts of Cyclone Idai.

The concept of ecosystems having alternative stable states is crucial to how we manage and predict the future of the biosphere. Savannas are the largest land cover in the tropics and support the livelihoods of 100s of millions of the world's poorest people. They also harbour an ancient, unique, charismatic flora and fauna - the "safari" landscape. However, the regions of the world where most savannas are found are widely considered to be bistable - that is they can support either savannas (a stable state with few trees that burns regularly and can support many grazing animals), or forests (another stable state with a closed tree canopy, high carbon storage, less grass and few fires).

Many scientists believe that these two states can both exist given the same environmental conditions, and that for example dry forests may "flip" to become savannas due to shocks to the system, such as cyclones. These ideas are widely used to guide savanna management and to predict how savannas will respond to climate and other global change.

However, whilst the concept of bistability in the dry tropics is easy to hypothesise, it is very hard to prove and has remained controversial. The evidence for it comes from mathematical models, and observations of static spatial patterns of tree cover and tree height, which show distinct "peaks" in their frequency distributions, which many have interpreted as the different stable states. Recent modelling work suggests however that these peaks are not reliable evidence of bistability - they are consistent with the idea, but other mechanisms associated with non-bistable ecosystems can generate these observations.

Cyclone Idai provides a unique opportunity to test these theories. The cyclone is reported to have toppled or damaged large numbers of trees, in areas where theory suggests alternative stable states are likely. Our field site, Gorongosa National Park, was subjected to some of the highest windspeeds during the cyclone and provides an excellent place to carry out long term observations of the impacts and recovery. We believe that, if alternative stable states do exist, we are very likely to observe them here.

CyclEx will first map the impact of the cyclone using radar remote sensing to examine the structure of the savannas and dry forest pre- and post-cyclone. We will then establish long term plots where we can understand how the vegetation recovers after the cyclone and see if areas where lots of trees were knocked over or damaged get "stuck" in a stable open savanna state, or if forest areas eventually recover their pre-cyclone structure and mix of species. This will tell us if switches between savanna and forests are possible, and whether the theory of alternative stable states is supported.

The results of CyclEx are important both as a test of the key ecological theory of alternative stable states, and on a more practical level. Intense cyclones are predicted to become more common in East Africa due to climate change, and to affect a wider area than at present. We need to know what their impacts will be on an ecosystem that is important for biodiversity and for local livelihoods, and CyclEx will be the first ever analysis of the impacts of cyclones in African savannas.

Our main pathway to impact is to promote high quality, long term monitoring of miombo woodlands. This is important for their management and long term sustainability, and thus, indirectly, the products and income worth $10 billion per year that is derived from these woodlands.

To achieve this, Ryan coordinates the NERC-funded SEOSAW network, and plays an active role in the Miombo Network. These provide, respectively, science coordination and science-policy interactions across the region, with a focus on long term monitoring and close links between managers and researchers. CyclEx will build on these networks and promote high quality standardised techniques for monitoring by providing experiential learning opportunities for two representatives of government agencies, who will actively participate in the research.

One participant from the Zimbabwe Forest Commission (ZFC) will join CyclEx, as ZFC are currently beginning a major programme of permanent sample plots establishment and are keen to collaborate on digital data acquisition and analysis methods. CyclEx will support one member of ZFC staff to be fully trained in the use of tablets for data collection, including form design with the Open Data Kit, and associated data management and analysis.

In addition, a participant from the Mozambican government will participate in CyclEx for one week. Mozambique is also currently establishing a network of permanent plots to help better manage their forest and savanna resource. The support of CyclEx will allow one manager from IIAM (Agricultural Research Institute of Mozambique; the implementing agency) to join the fieldwork for one week. In our experience, time in the field is the most effective means to the two way transfer of information about new methods, approaches and sampling design.

To make the most of the resources available within CyclEx, we will continue these relationships after the 6 month project through close collaboration as part of SEOSAW and Ryan's continuing work with both agencies.

Specific deliverables
1. A field protocol adapted to ZFC's needs that includes an ODK form for data entry, and R code for data analysis
2. A plot design protocol adapted to the Mozambique government's needs, that links with their remote sensing forest monitoring system.

We will also work closely with GNP scientific staff to disseminate the results of CyclEx and translate the research findings into management implications. The initial map of the cyclone impacts will be shared with GNP staff during the project.