Our giraffe conservation research is discovering where Masai giraffes are doing well, where they are not, and why. Our research is working to protect and connect areas important to Masai giraffe conservation.
The gentle, iconic giraffe indicates the health of African savanna ecosystems, home to the most spectacular displays of wildlife in the world.
But savanna ecosystems are in serious trouble. Habitat loss, illegal hunting, and disease are decimating savanna wildlife. Giraffe numbers have declined drastically to only 97,000. Africa-wide, elephants outnumber giraffes 4 to 1.
Despite the popularity of giraffes, scientists know surprisingly little about them.
Wild Nature Institute scientists are studying wild Masai giraffes in the Tarangire and Serengeti Ecosystems of Tanzania using a computer program that recognizes each animal’s unique fur pattern from photographs. We are monitoring thousands of individual giraffes throughout their lifetimes in an area over 25,000 sq. km. This is the biggest giraffe study, and one of the biggest large-mammal demography studies in history.
In Project GIRAFFE: GIRAffe Facing Fragmentation Effects, we are learning how natural and human factors affect giraffe demography in a landscape where wildlife habitat is increasingly fragmented by humans. Demography is survival, births, and movements, and these processes determine whether a population is growing or shrinking.
Our giraffe research is urgently needed so we can provide effective conservation actions in an ever more fragmented world, and ensure the future of wild giraffes and all creatures of the savanna.
We participated in the global status assessment of giraffes through the IUCN, and our research documented the effectiveness of community conservation and anti-poaching efforts. Wild Nature Institute scientists are affiliated with Pennsylvania State University, University of Zürich, and the Nelson Mandela African Institute for Science and Technology. We also collaborate with scientists from The USA, South Africa, Tanzania, the United Kingdom, and Switzerland.
Precision, accuracy, and costs of survey methods for giraffe Giraffa camelopardalis was one of our first publications from the project. We estimated giraffe density and abundance in the Tarangire Ecosystem in northern Tanzania using two ground survey methods—distance sampling and capture-mark-recapture—and compared our ground-based estimates with those from the most recent aerial survey. We found aerial survey estimates were biased low, while ground-based surveys were more precise and cost less. However, aerial surveys are useful over large regions of Tanzania and thus can provide landscape-scale population estimates. We computed correction factors to improve the accuracy of aerial surveys and suggested ways to further improve aerial survey methods.
In Spatial variation in giraffe demography: a test of 2 paradigms, we examined whether spatial variation in demography of a tropical mega-herbivore (the giraffe) followed the “temporal paradigm” or the “adult survival paradigm” of ungulate population dynamics that were formed from temperate-zone studies. We quantified how giraffe demographic rates of survival and reproduction varied across space at regional (northern Tanzania) and continental (Africa-wide) scales. Spatial variability of demographic rates at the continental scale supported the temporal paradigm of low variability in adult survival and more highly variable reproduction and calf survival. In contrast, at the regional scale, adult female survival had higher spatial variation, which supported the adult survival paradigm. At both scales, variation in adult female survival made the greatest contribution to variation in local population growth rates. We also found human-caused reductions in adult female giraffe survival are the most likely reasons of population declines.
In Giraffe demography and population ecology, we summarized current knowledge of demography and population ecology of giraffes and provided a framework for using population models when developing and evaluating conservation and management efforts for giraffes (or other large herbivore species).
In Migratory herds of wildebeests and zebras indirectly affect calf survival of giraffes, we utilized our data about a large-mammal predator–prey savanna food web to evaluate support for 2 hypotheses relating to the indirect effects of “apparent competition” and “apparent mutualism.” We examined how the presence of migratory herds of wildebeests (Connochaetes taurinus) and zebras (Equus quagga) affected survival of resident giraffe calves, as mediated by their shared predator, African lions (Panthera leo). African lions are generalist predators whose primary, preferred prey are wildebeests and zebras, but lion predation on secondary prey such as giraffes may change according to the relative abundance of the primary prey species. We found that local lion predation pressure on giraffes was reduced by local density of wildebeests and zebras, making giraffe neonatal and calf survival probabilities higher when the migratory herds were present. This supported the apparent mutualism hypothesis. Natural predation had a significant effect on giraffe calf and neonate survival, and could significantly affect giraffe population dynamics, thus if wildebeest and zebra populations in this ecosystem continue to decline as a result of increasingly disrupted migrations and poaching, then giraffe calves will face increased predation pressure as the predator–prey ratio increases. Our results suggest that the widespread population declines observed in many migratory systems are likely to trigger demographic impacts in other species due to indirect effects like those shown here.
We were proud to contribute to the IUCN Red List Assessment for giraffes, which reclassified giraffes as Vulnerable due to an observed population decline of 36–40% over three generations (30 years, 1985–2015). The factors causing this decline (direct killing and habitat loss) have not ceased throughout the species’ range. The best available estimates indicate a total population in 1985 of 151,702–163,452 giraffes (106,191–114,416 mature individuals), and in 2015 a total population of 97,562 giraffes (68,293 mature individuals). Some giraffe populations are stable or increasing, while others are declining, and each population is subject to pressure by threats specific to their local country or region. The populations of giraffes are scattered and fragmented with different growth trajectories and threats, but the species trend reveals an overall large decline in numbers across its range in Africa.
We also documented for the first time that Season of birth affects juvenile survival of giraffe. Variation in timing of reproduction and subsequent juvenile survival often plays an important role in population dynamics of ungulates in temperate and boreal regions. Tropical ungulates often give birth year round, but survival effects of birth season for tropical ungulate species were previously unknown. We found significant differences in juvenile survival according to season of birth, with calves born during the dry season experiencing the highest survival probability. Phenological match (matching birth season with vegetation growth) may explain the juvenile survival advantage to offspring born during the dry season from 1) greater accumulated maternal energy reserves in mothers who conceived in the long rainy season, 2) high-protein browse in the late dry-early short rainy seasons supplementing maternal and calf resources, 3) reduced predation due to decreased stalking cover, or some combination of these. Asynchrony is believed to be the ancestral state of all ungulates, and this investigation illustrated how seasonal variation in vegetation can affect juvenile survival and may have played a role in the evolution of synchronous births.
We also contributed to a lively discussion about How many species of giraffe are there? Giraffes are presently classified as one species, with nine subspecies. A paper in Current Biology presented DNA data and a taxonomy with four species of giraffe. The present consensus of one species divided into nine subspecies had previously been questioned several times over the past few decades. We presented the various taxonomic schemes and offered that the fundamental reason for different taxonomic interpretations is that they are based upon different datasets that adopt different statistical techniques and follow different criteria. These different taxonomies create a basis for future taxonomy discussions and conservation efforts.
Movements and source–sink dynamics of a Masai giraffe metapopulation provided a regional metapopulation analysis of the Tarangire ecosystem to inform conservation management for Masai giraffes in five subpopulations defined by land management designations. We assessed the source–sink structure of the study population, and we created a matrix metapopulation model to examine how variation in demographic components of survival, reproduction, and movement affected metapopulation growth rate. Movement data indicated no subpopulation was completely isolated, but movement probabilities varied among subpopulations. Source–sink statistics and flow of individuals indicated three subpopulations were sources, while two subpopulations were sinks. We found areas with higher wildlife protection efforts and fewer human impacts were sources, and less-protected areas were identified as sinks. Our results highlight the importance of identifying source–sink dynamics among subpopulations for effective conservation planning and emphasize how protected areas can play an important role in sustaining metapopulations.
Together with computer engineers from Microsoft, we published An automated program to find animals and crop photographs for individual recognition. This paper describes a new image processing service using machine learning technology deployed on the Microsoft Azure cloud. Using a computer vision object detection algorithm, the Microsoft team trained a program to recognize giraffe torsos using some existing annotated giraffe photos. The program was iteratively improved using an efficient Active Learning process, where the system identified new images and showed its predicted cropping squares on these images to a human who could quickly verify or correct the results. These new images were then fed back into training algorithm to further update and improve the program. The resulting system identifies the location of giraffe torsos in images with a very high accuracy.
Correlates of home range sizes of giraffes, Giraffa camelopardalisexamined what affects the size of giraffe home ranges. We found that giraffes living closer to towns had larger home ranges than giraffes living far from towns, suggesting a need to range longer distances—and expend more energy—to obtain critical resources in human-impacted areas. No such relationship was evident with bomas, which are homesteads built by indigenous livestock-keeping Maasai people, suggesting that giraffes are tolerant of more traditional, lower-impact land uses.