Get reliable science

Rangeland restoration can improve habitat for threatened species such as the hirola antelope (Beatragus hunteri) that inhabit savannas of eastern Kenya. However, restoration success likely varies across soil types and target restoration species, as well as according to restoration approach. We tested the response of four native grass species (Cenchrus ciliaris, Enteropogon macrostachyus, Eragrostis superba, and Chloris roxbhurgiana) to four different restoration approaches (tilling, manure application + seeding, seeding, no treatment). We also tested the interaction between planted grass and other functional groups using ANOVA. In each of two soil types, we located three 50m x 20m treatment blocks. Within each block, were 16 treatment plots that were randomly assigned to one of 16 species-site preparation combinations (4 species * 4 site preparation treatments). We seeded in May 2017 and assessed species cover in July and August 2017. Preliminary results suggest total grass cover was higher in the seeded treatment than the seeding + manure treatment. Both tilling and no treatment did not result in any significant above ground biomass suggesting that lack of seeds rather than soil capping or water availability might be the key mechanism limiting grass growth. There were no statistically significant interactions between the effects of the treatments and the occurrence of other, non-planted, grass species on planted grass cover. In contrast, the Tukey�s HSD post hoc paired tests showed significant differences among all treatments. Overall, planted grass species performed better in loam soils (median 45% cover) than in high clay (black cotton) soils (median =40% cover). Similarly planted grasses performed better than other grass species and forbs in both loam and black cotton soils. These experiments are aimed at informing landscape level grassland restoration for hirola, where tree encroachment has suppressed their recovery for nearly three decades.