COS 78-1 - Megaherbivores suppress liana infestation in a semi-arid African savanna

Thursday, August 15, 2019: 8:00 AM
L015/019, Kentucky International Convention Center
Tyler C. Coverdale1, Ryan D. O'Connell2, Matthew C. Hutchinson3, Amanda Savagian2, Tyler R. Kartzinel4, Todd M. Palmer5, Jacob R. Goheen6, David J. Augustine7, Mahesh Sankaran8, Corina E. Tarnita2 and Robert M. Pringle3, (1)Ecology and Evolutionary Biology, Cornell University, Ithaca, NY, (2)Ecology and Evolutionary Biology, Princeton University, Princeton, NJ, (3)Department of Ecology and Evolutionary Biology, Princeton University, Princeton, NJ, (4)Ecology and Evolutionary Biology, Brown University, (5)Biology, University of Florida, Gainesville, FL, (6)Department of Zoology and Physiology, University of Wyoming, Laramie, WY, (7)Rangeland Resources and Systems Research Unit, USDA-ARS, Fort Collins, CO, (8)National Centre for Biological Sciences, Bangalore, India
Background/Question/Methods

Understanding the determinants of savanna tree cover has been a primary goal of ecological research for decades and, historically, this research has focused on the effects of fire, rainfall, and herbivory on tree communities. In contrast to tropical forests, where recent research has revealed important effects of lianas (woody vines) on tree demography, there has been little prior research on lianas in tropical savannas. For example, increases in liana abundance in Neotropical forests over the last forty years have clearly demonstrated the potential of woody vines to reduce tree abundance, diversity, growth, and survival, yet almost nothing is known about the prevalence of lianas in savannas or their impact on savanna trees. In this study, we investigated the direct and indirect interactions between lianas, trees, and large mammalian browsers in a semi-arid Kenyan savanna. Specifically, we used long-term herbivore exclosures, smaller-scale liana manipulations, and DNA metabarcoding to (i) explore the effects of liana infestation on tree growth and fitness and (ii) investigate whether the abundance of lianas in savannas is constrained by large mammalian browsers.

Results/Conclusions

Our results suggest that lianas have a deleterious effect on savanna tree growth and fitness: experimental liana removal increased tree growth by 185% relative to infested controls, and more heavily infested trees produced significantly fewer fruits and flowers than uninfested conspecifics. The severity of liana infestation, in turn, was constrained by large browsing mammals, whose effect varied as a function of body size. Overall, lianas were present in the diets of eight species ranging in size from dik-dik (c. 5kg) to elephants (c. 5,000kg) and were scarce (<5% canopy cover) outside herbivore exclosure plots. Herbivores >120cm tall (e.g., impala, elephant) had the strongest effect on lianas in tree canopies: liana abundance was nearly five-fold greater within plots that excluded these herbivores compared to unfenced controls. Smaller herbivores (e.g., dik-dik), in contrast, had the strongest effect on liana seedling density. Liana abundance increased more than twenty-fold over ten years in the absence of herbivores, resulting in severe liana infestation (liana cover >20%) of c. 50% of trees. Although lianas have received comparatively little attention in savannas, our results suggest that they could become increasingly important determinants of savanna structure and function if ongoing declines of native herbivores continue unchecked.