The Science Seminar Series: September 27, 2007 4pm

Interaction of Small and Large Mammalian Herbivores in an Experimental System on the Savanna Grasslands of East Africa

Dr. Brad Bergstrom

Department of Biology
Valdosta State University

Powell Hall

Time: 4:00 - 5:00pm

Abstract

Interaction of Small and Large Mammalian Herbivores in an Experimental System on the Savanna Grasslands of East Africa

Two main soil types predominate in the Central Highlands of Kenya: poorly drained black cotton vertisols are dominated by open Acacia drepanolobium savanna, and sandy red soils are more densely bushed with a diversity of Acacia and other woody species. Diversity of rodent species is especially low in the black-cotton grasslands, with Saccostomus mearnsi the only dominant, which often occurs in low densities. In the red soils, one or more species of Tatera, Acomys, and Aethomys are common co-dominants. However, two circumstances can greatly increase abundance and diversity of rodents in both habitats: 1) exclusion of ungulates and elephants, and 2) glades, which represent secondary succession following abandonment of highly enriched pastoralist sites called bomas. Field sampling in 2006 showed the greatest effect occurred with a combination of these two, i.e. in ungrazed glades. The effect of large-herbivore exclosures on abundance and diversity of small mammals was significant on both soil types but more pronounced on red soils. Total small-mammal density on ungrazed red-soil glades (54-112/ha) was about twice that of either ungrazed red-soil nonglades or ungrazed black-cotton glades. Yet densities on grazed red-soil glades (4-12/ha) were similar to grazed black-cotton glades, even though the latter had more grass cover. Also, Arvicanthus (a diurnal rat) was rare to absent except in ungrazed exclosures, primarily on glades (up to 22/ha on black-cotton, 82/ha on red-soil). On red soils, S. mearnsi was rare to absent except within exclosures, both glade and non-glade (up to 32/ha). On both soil types, total small-mammal density was significantly positively regressed (highly so for black-cotton sites) on mean grass height and mean density of dead grass stems (but not live stems). This suggests that abundance and diversity of small mammals is highly responsive to cover, rather than food resources, in this ecosystem dominated by large grazers.