School of Biodiversity, One Health & Veterinary Medicine

Understanding when human encroachments may push an ecosystem beyond the limits of sustainability is a task made difficult by the complexity and dynamical nature of ecosystems. Establishing baseline conditions and monitoring key ecological attributes are crucial first steps, but they may be insufficient to predict the full consequences of mounting threats. Remote sensing of vegetation patterns is widely used as an assay of landscape condition but the critical links between satellite-derived vegetation data and the trophic interactions that stem from them are still mostly not well understood.

I propose to establish these links in the Afroalpine ecosystem of Bale Mountain national park (BMNP) in Ethiopia through ground-truthing of the relationships between cattle, vegetation and rodents and inputting sequential assessment of vegetation condition, acquired from satellite imagery and ground-truthing, into a spatially explicit trophic model. This step is imperative for identifying the "condition envelope" for vegetation within which a system may be resilient to human-induced landscape change and could be used both to adaptively manage human activities (figure 1) and sustain biodiversity.

The Afroalpine and surrounding areas support an exceptionally high diversity of rare and endemic species, establishing the region as one of 34 biodiversity hotspots in the world (Conservation International). BMNP encompasses the continent’s most extensive pocket of Afroalpine heath and grassland, which is essential habitat for the rarest canid species, the Ethiopian wolf (Canis simensis) (figure 2).This endemic canid feeds upon an n-species guild of rare endemic rodents (figure 3).

Figure 1 & 2: Cattle grazing in BMNP and Ethiopian wolf (© Flavie Vial)

Figure 3: The endemic Giant mole-rat (Tachyoryctes Macrocephalus) and Unstripped grass-rat (Arvicantis abyssinicus) © Juan Manuel Morales & Flavie Vial.

However, over the last 20 years the BMNP has been used to graze increasingly high numbers of livestock that are thought to have impacted on rodent abundance. Investigating the effects of livestock grazing on the functioning of this ecosystem has been identified as the leading research priority in the recently formulated BMNP management plan.

The overall goal of this study is to quantify the indirect interaction between rodent density and demography, and livestock grazing pressure. Critical links between forage conditions, livestock grazing pressures and rodent assemblages will be made through field investigation to parameterise interactions at the lower levels of the food web. Existing data and ongoing studies of the linkages between rodent populations and wolf distribution and demography will be used to parameterise the upper level trophic interactions. The trophic model thus designed will be used to establish the composition and density of rodents required to sustain desired wolf numbers and the amount of vegetation required to maintain sufficient rodent assemblages. This research addresses the highest research priority identified by Bale Mountains National Park.

A fundamental paradox confronting conservation biologists is that the ecosystems they wish to conserve are subject to change (Burns et al. 2003). Changes can arise from endogenously induced fluctuations arising from normal ecosystems dynamic and, very often, accumulating stresses imposed by ever increasing human demands on ecosystem resources (Ludwig et al. 1993). Distinguishing between these agents of change is a major scientific challenge (Green et al. 2005); anticipating when the limits of acceptable change might be exceeded is a conservation imperative. Identifying the "condition envelope" of an ecosystem, within which key ecological attributes can be maintained, is essential for the effective management of anthropogenic change. Thus, laying down baseline data is a critical pre-requisite to identifying ecosystem drift and constitutes a vital line of defence in the management of threats arising from human encroachments.

While essential, simply monitoring key indicators of ecosystem health or integrity is unlikely to provide the fundamental understanding necessary to predict the future consequences of mounting threats (Mace 2005). For example, remote sensing enables broad-scale assessment of certain ecosystem components but understanding what these can tell us about ecosystem function is essential if they are to be useful indicators of future ecosystem conditions (Pettorelli et al. 2005). How ecosystems respond to different stressors will depend on their intrinsic dynamics, this is why anticipating the effects of different ecosystem stresses requires additional knowledge of intra and inter-trophic interactions. The recognition that effective conservation strategies must embrace the dynamical nature of ecosystems lies at the heart of this proposal. I propose to develop an evidence-based system of vegetation condition, using a combination of remote-sensing and ground sampling, to provide a foundation for a model of trophic interactions that will guide an adaptive management process (Salafsky et al. 2002), whereby conservation of biodiversity is ensured and sustainable patterns of livestock grazing are identified.

The Afroalpine in the Bale Mountains has a simple and visible trophic structure with wolves and raptors feeding almost exclusively upon a guild of burrowing rodents, including 10 species endemic to Ethiopia (Sillero-Zuberi & Gotelli 1995; Sillero-Zuberi et al. 1995a). Together, these rodents constitute a collective biomass of 2,500-4,000kg/km2 , rivalling that of large mammals on the Serengeti plains (Sillero-Zuberi et al. 1995a). The rodents, in turn, feed exclusively on forbs and grasses (Clausnitzer & Kityo 2001). Over and above the intrinsic biological interest in the area, major rivers originating in the Bale massif are the only sources of perennial water for agricultural production and people occupying the arid lowlands of East and Southeast Ethiopia (Sillero-Zuberi & Macdonald 1997). As agriculture expands in the lowlands, nomadic pastoralists are forced onto higher ground where persistent overgrazing erodes the stability of the Afroalpine ecosystem and threatens water supply. Thus, the livelihoods and food security of the people in the lowlands areas are highly dependent on good environmental management of the highlands.

The Afroalpine within BMNP is an appropriate setting for this research for several reasons:

First, the ecosystem is not blighted by a history of cultivation, land degradation or recurrent drought. However, negative pressures on natural resources are rapidly growing. The direct threat of canid diseases to the park’s flagship species, the Ethiopian wolf, has received a great deal of attention (Sillero-Zuberi & Macdonald 1997; Haydon et al. 2002; Haydon et al. 2006), but the increasing pressure of livestock grazing constitutes a potentially more profound threat to the integrity of the ecosystem as a whole through its impact on the rodent communities (Keesing 1998; Eccard et al. 2000) and cascading impacts on higher trophic levels (Stephens et al. 2001). Park planners have now recognised livestock grazing as the major stress on the key ecological attributes of the Afroalpine and identified research on stress abatement strategies to be an urgent priority.

Second, the park urgently requires scientific leadership to organise research initiatives able to deliver evidence-based approaches to inform sustainable livestock grazing strategies. Third, recent research conducted by the Ethiopian Wolf Conservation Program (EWCP) on the effect of rodent density on wolf demographic processes (Marino 2003; Ashenafi et al. 2005; Tallents 2006), the functional responses of raptors to rodent density and rodent habitat selection and diet (Sillero-Zuberi et al. 1995b; Tallents 2006) enable the parameterisation of the interactions within and between the upper trophic levels of the food web.

Finally, the proposed research coincides with a time of renewed investment in the research infrastructure of the park through local government initiatives and the support of NGOs, including FARM Africa and the Frankfurt Zoological Society. These NGOs have established respectful, working relationships with local communities and herdsmen; procured funding to develop a GIS facility in the park and acquire regular satellite imagery for monitoring; hired monitoring staff, acquired dedicated research vehicles and field equipment (such as live-traps for rodents); and continue to invest in graduate research programs for students from the University of Addis Ababa for whom scientific monitoring is sorely needed.

The goal of this research is to develop protocols to enable the adaptive management of pastoral grazing pressure so as to ensure the stability of important trophic interactions in a system of high conservation priority- the Afroalpine region of Bale Mountains National Park in Ethiopia. My research strategy will include:

• Construction of a dynamic grazing map of the BMNP in order to understand patterns of pastoral cattle use.
• Establish empirical relationships between grazing pressure, vegetation biomass and rodents density and demography.
• Broad-scale predictive modelling of dynamics of trophic interactions of vegetation/rodent and rodent/wolf/raptors.

This research will deliver a quantitative and technically synthetic framework for managing livestock grazing, thereby contributing to the alleviation of a critical human-wildlife conflict that endangers an area of incalculable biodiversity value. The project will also provide a medium to long-term organising framework for research in the trophic relationships of the Afroalpine. An important additional result of this research will be the knowledge transferred to local communities, park staff and regional governing bodies regarding the issues surrounding livestock grazing in this sensitive ecosystem. The transfer of technical skills to park staff will result in better assessments of ecosystem condition and facilitate iterative evaluation and revision of grazing manipulations. Additionally, opportunities to involve local students in ecological studies will be generated by this research.