Fences and artificial water affect African savannah elephant movement patterns

The IUCN Redlist considers the African savannah elephant (Loxodonta africana) to be "vulnerable" despite it numbering in the 100,000s and having a large geographical range. This seeming paradox stems from how quickly human persecution can eliminate elephants across large areas and how quickly elephant numbers can increase when protected. Much elephant research concentrates on the extent and consequences of elephant persecution. Where elephants thrive, two other human interventions, the provision of artificial water and the construction of fences may have large, and perhaps unintended, impacts on elephant behavior. In general, successful management requires that we understand elephant movements and land-use choices across large areas and long periods. Here, we ask specifically how artificial water and fences might affect these movements. To do this, we first characterize how elephants move in different seasons and landscape types, in different years, and how these patterns change over a region that varies considerably in annual rainfall. We fitted 73 elephants with GPS collars across a large rainfall gradient spanning seven southern African countries over a period of 6 years. We analyzed remotely-sensed environmental data from four satellite borne sensors that measure daily rainfall, weekly temperature, bimonthly greenness, and summarise human infrastructure. Elephants move approximately 6 km/day in dry landscapes, down to approximately 3 km/day in the wettest ones. Strong seasonal differences modulate geographic differences. Elephants move less, cover less area, and are more faithful to landscapes across years in the dry season than the wet. Water availability drives these seasonal patterns. Seasonal differences in the area covered are less pronounced in wet landscapes where permanent water is more dispersed. Within-day movements reveal that elephants are consistently crepuscular but more active at night than midday when temperatures are high. Direction-changes are centered at midnight when elephants are close to water indicating regular nighttime treks to water. By design, our analyses seek to find general patterns of elephant movements - something that one can achieve only across a large range of locations and ecological conditions - in order to understand the impact of human interventions. We show that both interventions reduce seasonal differences in elephant ranging patterns and increase local impacts of elephants on the vegetation. Artificial water sources allow more extensive dry season ranging, allowing elephants to use - and potentially overexploit - vegetation in areas that would have be otherwise inaccessible to them except in the wet season. Fences cause elephants to "bunch-up" against them during the wet season, again locally increasing the pressure elephants put on their resources. (C) 2009 Elsevier Ltd. All rights reserved.