Fire is an integral component of tropical grassland and savanna ecosystems and influences both the flora and fauna of these ecosystems. Fire is a key factor in managing habitat of wild large mammalian herbivores, many of which are threatened by poaching and human encroachment. Previous studies have found that herbivores are attracted to burned areas, presumably because plant nutrient content is often higher following fire (plant nutrient hypothesis). However, herbivores may also preferentially use burned areas to avoid predators (predator avoidance hypothesis) because of increased sight distance and lack of predator camouflage.
The relative benefits of plant nutrition and safety from predators maybe stronger for smaller herbivores that require more nutrient-rich diets and are more vulnerable to predation. We know of no study that has tested these two alternative hypotheses for both whole herbivore assemblages and individual species, so we conducted perhaps the most thorough study of herbivore attraction to burned areas in Serengeti National Park, Tanzania.
Herbivores were counted bi-weekly for a year on 14 paired burned and unburned sites. We measured vegetation height, herbaceous plant biomass and nitrogen and phosphorus content bi-monthly and obtained a long-term database on lion radio-tracking locations to assess predator distribution relative to fire.
Results/Conclusions
We found that overall, herbivores prefer burned areas (P=0.002). Burned areas featured herbaceous forage that was significantly higher than unburned areas in nitrogen and phosphorus during both wet and dry seasons. Lions were sighted much more often than would be expected by chance in unburned areas (P<0.001). Smaller herbivores, such as Thomson’s gazelles, Gazella thomsoni, (P=0.004), impala, Aepyceros melampus, (P=0.04), warthog, Phacochoerus africanus, (P=0.05), and topi, Damaliscus korrigum, (P=0.05) overall preferred burned areas.
Thomson’s gazelle preference for burned areas was strongly evident (P=0.005) during the dry season but disappeared during the wet season (P=0.41), despite more nutrient-rich forage in burned areas. Vegetation height, and presumably risk of predation, was significantly lower in burned areas during the dry season (P=0.001), but during the wet season burned area vegetation height was not significantly lower than in unburned areas (P=0.7).
In contrast, larger herbivores, such as Cape buffalo, Syncerus caffer, (P=0.05) and elephant, Loxodonta africana, (P=0.05), which are less vulnerable to predators showed no preference for burned areas. Our data thus support the predator avoidance hypothesis for smaller herbivores and suggest that predator avoidance maybe a stronger factor than plant nutrition in influencing the response of mammalian herbivores to fire.