95th ESA Annual Meeting (August 1 -- 6, 2010)

COS 104-8 - Proliferation of a native succulent in semi-arid Kenyan rangelands:  Ecohydrological dynamics and implications for resilience

Thursday, August 5, 2010: 4:00 PM
320, David L Lawrence Convention Center
Elizabeth King, Odum School of Ecology, University of Georgia, Athens, GA and Trenton Franz, Dept. of Hydrology and Water Resources, University of Arizona, Tucson, AZ
Background/Question/Methods In the semi-arid rangelands of north central Kenya, environmental history interviews have documented that the abundance of grasses has steadily declined over the last twenty years.  Concomitantly, the native succulent herb Sansevieria volkensii has greatly proliferated in many areas.  This is creating a shift in ecosystem structure from contiguous herbaceous cover to a landscape with large, dense stands of S. volkensii and interpatches of bare, sealed soil surfaces and small grassy areas. In addition to loss of forage in the landscape, S. volkensii stands also hinder livestock passage, so local pastoralists view this transition as highly undesirable. Understanding the mechanisms and reversibility of the transition is necessary to develop management options.  Since water availability is a key driver in this dryland ecosystem, we sought to assess the ecohydrological and vegetation dynamics that may govern the vegetative spread of S. volkensii stands via rhizomes.  On Koija Group Ranch, a communal landholding in Laikipia District, Kenya, we assessed surface water flow patterns, soil moisture, S. volkensii proliferation, and adjacent vegetation density along the margins of 8 dense patches of S. volkensii, ranging from 15 to 40m in diameter.  Based on visible flow paths on the soil surface, patch margins were divided into segments categorized as run-on (into patches) or run-off/neutral (out of or adjacent to patches).   

Results/Conclusions

Using an electromagnetic induction device that measures average electrical conductivity in 1m3 of soil over which the device is passed, we confirmed that the visually identified run-on zones were significantly correlated with greater soil moisture 0 to 1.5m inside the patch margin.  We found that run-on margin segments had significantly higher S. volkensii ramet density, more new ramets, and greater leaf water content than non-run-on segments.  We also assessed whether adjacent interpatch grass and forb density affected the distribution of new S. volkensii ramets, and found that new ramet density decreased significantly with increasing adjacent herbaceous cover.  From these findings we hypothesize that ecosystem resilience is a function of the disparity between soil water dynamics in S. volkensii patches and interpatch areas.  In areas with sealed soil surfaces and little herbaceous cover, S. volkensii patches appear to create positive feedbacks of water redistribution that favor their proliferation.  Resilience of the desired ecosystem state may be enhanced by manipulations that increase interpatch infiltration and vegetation cover, as this would help disrupt the hydrological feedbacks that are impelling the ongoing transition toward a S. volkensii dominated state.