Work Package 1:
Savanna Dynamics and Global Change
Savanna Dynamics and Global Change
COORDINATORS Patrick Duncan (UPR 1934 CNRS, Chizé) and Jeremy Midgley (Botany department, University of Cape Town)
Savannas are multi-state systems, and the ecological drivers which cause savannas to change from one state to another (fire, large herbivores wild and domestic etc.) are beginning to be understood. The scientific aims of this WP are to contribute to improving our understanding of impact of global change on processes controlling a few model savanna systems (Hwange and Kruger in particular). As a contribution to the science-policy interface for sustainable development the aim is to use this knowledge to model the systems, and to build scenarios as a basis for management decisions.
Project 1 : The impact of ecosystem engineers, fire and large herbivores, on plant-soil interactions and biogeochemical cycles.
Participants : Jacques Gignoux, Luc Abbadie, Sébastien Barot, Jean-Christophe Lata (CNRS UMR 7618, Paris), Simon Chamaillé-Jammes (CNRS UMR 5175 Montpellier), Hervé Fritz (CNRS UMR 5558 Lyon), William Bond (University of Cape Town).
In savannas the functioning of the nutrient cycles is highly modified by fire, grazers and browsers. Grazing lawns concentrate heavy grazing pressures, making them a useful system for the study of these interactions. A PhD (O. Bonnet) enabled us to show that large grazers (in particular, white rhinos) are able to consume the primary production immediately so no standing crop accumulates. This functioning is interrupted during dry spells and is activated by rainfall, including during the dry season. Other work (L. Abbadie, X. Le Roux) includes measurements of the effects of selective grazing on the microbial community and other soil properties in Hluhluwe-iMfolosi Park where exclosures show major or little effect on the nitrogen and carbon cycle respectively. This GDRI allowed us to submit a research proposal on carbon allocation under various regimes of climate, fire and herbivory to the French Agence National de la Recherche; South African sites play an important role in the project.
The understanding of the impact of large herbivores and fire on grasses (biomass, production, species composition and quality as a food resource), has improved greatly. In particular, the nitrogen cycle in savannas is now relatively well understood; however, the phosphorus cycle is less well known. Recent results show that complex interactions between these cycles are the rule: grasses are limited by phosphorus under acacia trees and by nitrogen in open areas. Mycorrhizae are known to play an important role in the phosphorus cycle, and have rarely been studied in savannas. Our objective for the future will be to better document the nitrogen and phosphorus cycles in savannas, using the gradients of grazing / browsing intensity, fire frequency and intensity, rainfall and soil nutrient retention ability available in parks such as Hluhluwe-iMfolosi and Kruger. This involves (1) characterization of the main microbial functions encountered under various environmental conditions (e.g. nitrification, denitrification, nitrogen fixation), (2) assessment of the phosphorus cycle, including the incidence of mycorhizae, (3) evaluation of plant allocation strategies for carbon, nitrogen and phosphorus (3) modelling the coupled carbon, nitrogen and phosphorus cycles.
Project 2 : Factors controlling large herbivore populations: direct and indirect effects of predation in the context of resource constraints.
Participants: Simon Chamaillé-Jammes (CNRS UMR 5175 Montpellier), Hervé Fritz, Jean-Michel Gaillard, Christophe Bonenfant (CNRS UMR 5558 Lyon), Patrick Duncan (CNRS Chizé), Michel de Garine-Witchatitski, Mathieu Bourgarel (Cirad), Elissa Cameron, André Ganswindt (MRI, Univ. Pretoria), Norman Owen-Smith (Univ. of Wits), David Ward, Adrian Shrader, Rob Slotow (Univ. Kwazulu-Natal).
This project studies the complex interaction between habitat quality, climate, pathogens and predation on herbivore abundance, and the cascading effects on vegetation. We particularly rely on the Hwange Environmental Research for Development programme (HERD) based on an exceptional, long term, cooperation between the Zimbabwe Parks and Wildlife Management Authority, two French organizations the Cirad and the CNRS, the universities of Harare and Bulawayo and two research/conservation programmes (Painted Dog Conservation and Hwange Lion Research, WildCru, Oxford University). This site (Hwange) and the surrounding forestry and communal areas, is currently under evaluation by the CNRS as an “Observatoire Homme-Milieu/Zone Atelier”. This Project currently has a Postdoc in Wits (S. Grange), 7 PhD students; an MPhil student Bulawayo University (NUST).
Interesting results have been obtained on an ‘ecosystem engineer’, the elephant, whose populations are increasing in many protected areas in Southern Africa. In Hwange our work showed that the increase was limited by food resources, whose access is constrained by the availability of surface-water. High elephant densities do not always cause shifts in vegetation communities, and no impact on medium-sized herbivores was detectable.
We have recently focused the program on innovative research on direct and indirect effects of predation. For instance our work on the interaction of resource availability and anti-predator behaviour has shown that the use of habitat of seven species of herbivores is strongly affected by the risk of predation, in the short and medium term. In the period 2011-4, this result will be pursued through the project “Landscape of fear and the use of landscape resource heterogeneity by ungulates of different body sizes. Testing new hypotheses on ungulate species coexistence” (funded by the French Agence National pour la Recherche). A PhD (F. Barnier) and a postdoc (M. Valeix) have been recruited. We are running a design a medium-scale field experiment to investigate behavioural adjustments medium-sized herbivores make when forced to make a trade-off between resource quality and predation risk. We also rely on GPS collars to document the locations of herbivores and predators simultaneously and how both use the various habitats. More than 60 GPS collars are thus currently deployed in Hwange (on 3 species of predators, 5 herbivores). We intend to link with South African research groups working in smaller reserves to replicate this study and assess the generality of our findings. This will create a strong incentive to build a ‘community-of-practice’ research network working on the behaviour and dynamics of large mammals.
Project 3 : Scenario building for sustainable use and conservation of natural resources: modelling the dynamics of complex, spatially organised systems, with special emphasis on interactions between wildlife and people around protected areas
Participants: Hervé Fritz (UMR 5558 CNRS Lyon), Luc Doyen (MNHN Paris), J Manjegwa, B. Mukamuri (CASS, University of Zimbabwe), A. Binot, M De Garine-Wichatitsky, M-N De Wisscher, E. Etter (CIRAD-Agir), P. Mundy, E. Mwenje (NUST, Bulawayo), Shakkie Kativu (TREP, University of Zimbabwe), A. Murwira (Geog. & Env. Sci, University of Zimbabwe), Graeme Cumming (Univ. Cape Town)
In savannas the distribution of vegetation, risk of predation and disturbance are major drivers of animal movements, and the complexity of the systems is even greater when dealing with a mosaic of protected areas and agricultural land: the spatio-temporal organisation of land use types as well as ecological processes may be central in allowing the various components of the system to coexist. The edge of Protected Areas (PAs) is therefore a perfect laboratory to study the determinants and conditions that can allow for the sustainability of an integrated wildlife-human system.
The new framework of this Project will integrate complementary initiatives to link research-policy in order to better manage the interface between PAs and their peripheries (e.g. HERD programme, the AIRD inititiave on research and PAs management, CIRAD's network of research platforms). The main study objects will be socio-ecological systems that rely to a large extent on ecosystem goods and services, in order to produce wealth and improve standard of living of the inhabitants. It thus appears that our system is defined by a set of components that can be most interestingly in interaction, through a pluridisciplinary approach (ecology, political science and economics). The main objectives will be to understand the determinants of coexistence between humans and wildlife at the periphery of PAs, and the sustainability of these integrated systems. As an attempt to produce tools for scenario-building, the results of field studies will be integrated in co-viability models, or used for role-game scenario planning exercises.
During the GDRI 191 we launched two main studies which should provide case studies for the work in 2011-2014. The first model on the Hwange systems shows that the spatial arrangement of the park, safari and communal areas is crucial for the viability of a large elephant population, lucrative trophy hunting and sustainable meat production, and that environmental stochasticity strongly decreases the number of viable scenarios (PhD of C. Guerbois). We then developed a study to model the spatial coexistence of buffaloes and cattle at the periphery of protected two areas (Hwange and Gonarezhou - PhD of E. Miguel).With the objective of developing bio-economical models for integrated and sustainable management of African wildlife including protected areas, we will now integrate the socio-economic component of one of our long-term sites (Hwange) in a model of the complex dynamics of this system.