Habitat loss and degradation are leading causes of biodiversity loss. When habitat is completely lost (e.g., deforestation), so is most of its native biodiversity. When land is degraded (e.g., overgrazing), biodiversity losses are more variable. There has been piecemeal documentation of this loss in some sites, but rarely has a comprehensive survey sought to examine the effects of different land uses on a wide range of taxa, and the relationships among them. We are proposing to do multi-taxa surveys of multiple properties in Laikipia, Kenya that differ in land use (and soil type). In this semi-arid ecosystem, private land use can be categorized into four basic classes:
- Wildlife reserves with no livestock.
- Ranched with low stocking rates, and that encourage wildlife.
- Ranches with "full" stocking rates, and that discourage wildlife.
- Heavily overgrazed properties.
For each of these land-use classes, we have identified at least three "replicate" properties on each of the two dominant soil types in the ecosystem. We propose to carry out surveys across these properties to document the abundance of a wide range of taxa:
- Herbaceous and woody plants
- Arthropod invertebrates
- Reptiles
- Birds
- Small mammals
- Large mammals
We will also document stocking rates, range condition, fuelwood collection, and human density. In addition to land-use and soil type, these properties differ in mean rainfall, and distance from source populations, two variables that will be used as covariates in our analysis.
In particular we will ask:
- Are some taxa more sensitive than others to land-use change?
- Are biodiversity losses "nested" along a gradient on increasing habitat degradation?
- Are there (keystone) species or guilds whose loss appears to cascade through other taxa?
- How do the different herbivore exclusion treatments of the KLEE experiment simulate the larger-scale (and less controlled) variations in biodiversity across land uses?