Botswana Carnivore Monitoring Methodology Development

Wildlife populations are declining across Botswana. To develop conservation actions aimed at reversing this trend, information on Botswana’s wildlife populations is desperately needed. This research project aims to address the knowledge gap by identifying and implementing a sustainable method for monitoring carnivore communities. Camera trap and spoor studies will be employed to identify a monitoring programme. Communities and government agencies will be offered field training workshops to implement the programme. In addition, the project has started an outreach programme, Wild Joys, to help instil a conservation ethic in future generations in Botswana

Region: Northern Botswana
Organization: Department of Fish and Wildlife Conservation, Virginia Tech
Qualifications:
MSc. University of Montana, 2010
BSc. Colorado State University, 2005

Northern Botswana has an abundance of wildlife and one of the most diverse carnivore guilds in Africa. Similar to the rest of the continent, however, these carnivores are in decline and threatened by illegal hunting and large-scale changes in land management policies. The imminent risk of declining biodiversity is of concern to Botswana as wildlife has both cultural and economic value.

In response to reports of declining wildlife, Botswana’s Ministry of Environment, Wildlife, and Tourism banned wildlife hunting and announced its intentions to implement additional conservation actions. To develop effective actions, however, information on the distribution, density, and ecology of Botswana’s wildlife populations is needed and this is what this research project aims to achieve.

Furthermore, the long-term conservation of wildlife ultimately depends on today’s youth. In Botswana, children from rural communities rarely have the opportunity to view the diversity of wildlife and consequently, retain negative perceptions of wildlife passed on by older generations. To address this, the project started Wild Joys, an outreach programme that provides children with positive, educational wildlife encounters.

• Estimate the density of up to 13 indigenous carnivores using camera trap surveys, in combination with advanced statistical models.
• Estimate overall carnivore richness and identify how ecological variables influence the distribution of carnivores across a multi-use landscape.
• Evaluate the efficacy of spoor surveys, as compared to camera trap surveys, for monitoring the distribution and population trends of large carnivores.

In addition to the research objectives, two outreach objectives have been identified:

• Build local capacity in regards to monitoring wildlife populations.
  Develop Wild Joys.

During the pilot study, 3 250 square km of focal study sites were identified, offering differing levels of human impact and falling along a gradient of resource availability and quality. The study sites will be divided into 50.5 square km sampling units.

For camera trap surveys, two camera stations will be deployed within each sampling unit – one in the centre and another at a randomly selected location. Separating camera stations by a range of distances will enable the photographing of individual animals at multiple camera stations so that information about their movement distances, a requirement for density estimators, can be obtained.

For each study area, half of the units will be sampled for 25 trap nights, and then the camera stations be will rotated and the remaining half will be sampled for an additional 25 trap nights. A simulation from the pilot study, based on estimated occupancy and detection probabilities found this sampling design balanced precision of occupancy estimates with survey efforts. When camera traps are deployed, spoor surveys will also be conducted. Every seven days, 20 km sand road routes will be surveyed for large carnivore spoor, with the help of local wildlife guides.

Following the collection of the field data, statistical analyses will be carried out. Spatially explicit capture-recapture models will be used to quantify the densities of carnivores with uniquely identifiable fur patterns (e.g. leopard). For more uniformly coloured carnivores (e.g. lion) newly developed spatial mark-resight models will be used.

To estimate carnivore richness and explore species and community-level effects of ecological factors, hierarchical multispecies occupancy models will be used. For each camera station, ecological factors hypothesised to influence carnivore occupancy include vegetative cover, human impact, prey counts, distance to water, and micro-habitat features will be estimated. Lastly, the occupancy of large carnivores will be estimated using spoor surveys in combination with spatial-autocorrelation occupancy models. Density and occupancy estimates will be compared based on camera trap and spoor surveys, respectively.

To build local capacity with regards to the monitoring of wildlife, local wildlife guides will be trained. Additionally, field training works and presentations will be offered to communities, government agencies and research institutes.

Finally, in collaboration with a wildlife guide from Sankuyo, a weekend programme geared towards children aged 8-15 has been developed, called Wild Joys. Each weekend, between seven and 10 children are taken into wildlife-protected areas and they are taught about natural history, animal behaviour and ecotourism in their local language, Setswana.

Annual Report June 2016
The research project explores the utility of camera trap surveys and advanced modelling techniques to inform biodiversity management, by applying a multi-species approach to a community of mammals in the Okavango Delta, Botswana, with a focus on carnivores.

Camera traps were deployed at 221 locations across a 550 km2 study area between February and August 2015. The study area fell across a gradient of human use, including a game reserve, wildlife management areas, and a livestock grazing area.

During 2015 sampling, 10 766 detections of 54 species were recorded during 6 607 trap nights, including 44 mammal species of which 22 were carnivores. Brown hyaena Hyaena brunnea (n = 3) and cheetah Acinonyx jubatus (n = 3) were photographed least often while elephant Loxodonta africana (n = 1665) and impala Aepyceros melampus (n = 900) were photographed most often.

With the help of colleagues at the Botswana Predator Conservation Trust (BPCT), photographed carnivore species (e.g., leopard Panthera pardus, wild dog Lycaon pictus, and spotted hyaena Crocuta crocuta) were identified to the individual level using pelage patterns; for example, 67 individual leopards and 209 spotted hyaena were identified.

A children’s outreach programme was carried out in 2015 with the help of Dikatholo Kedikilwe, a long-time research assistant at BPCT. Each of the 90 children in grades 3 – 7 in the rural community of Sankuyo, Botswana, participated in the programme twice; ~10 children participated per day as well as a teacher. The programme took children on day trips to the bush to see wildlife in its natural environment and learn about field techniques, natural history, animal behaviour, ecosystems and conservation.

July 2015
Between February and June 2015, two study areas were sampled, each divided into two sections. Study Area 1, in Wildlife Management Areas NG33/34 and Moremi Game Reserve, is used by self-drive tourist and commercial safari companies. Study Area 2 is in Wildlife Management Area NG34 and has minimal tourism activity. Cameras are currently deployed in Study Area 3 located on the livestock side of the veterinary fence. This area will be sampled until 19 August 2015.

In 2015, 23 and 20 species of carnivore were photographed in study areas 1 and 2 respectively. As happened in 2014, cheetah were rarely photographed despite being visually observed on numerous occasions. Spotted hyaena were the most frequently photographed as well as the most ubiquitous carnivore in both study areas. Leopard were the second most pervasive carnivore, being photographed in 91% and 74% of the sampled grid cells in study areas 1 and 2 respectively. African civet, honey badger, lion and serval had higher rates of detection and were more pervasive in Study Area 1, while bat-eared fox, Cape fox and wild dogs had higher rates of detection and were more pervasive in Study Area 2. Among the meso and large carnivores, side-striped jackals and aardwolves were photographed the least often.

Together with the Botswana Predator Conservation Trust (BPCT), leopard, wild dog, spotted hyaena, serval, civet and lion (when possible) were identified to the individual level using pelage patterns. A total of 44 additional species of wildlife were photographed. Baboon, impala and warthog had higher rates of detection in Study Area 1 where the habitat is dominated by acacia scrub, woodlands and grassland savannah. Other species, such as Cape hare, common duiker, steenbok and roan antelope had higher rates of detection in the mopane-dominated Study Area 2. Eland and sable antelope were only photographed in Study Area 2, while most of the photographed bird species were detected more often in Study Area 1.

The children’s outreach programme was successfully completed with the assistance of Dikatholo Kedikilwe, a research assistant at the BPCT. Between February and June 2015, educational day trips for schoolchildren as well as their teachers were hosted in the bush. All 90 children in Sankuyo community participated in the programme twice. For many children this was the first time they had seen live carnivores in their natural habitat.

April 2015
The project’s final field season commenced on 3 February 2015. Cameras were deployed in Study Area 1 in Wildlife Management Areas NG33/34 and Moremi Game Reserve – which was also sampled during the dry season in 2014. A total of 42 camera stations were deployed in the southern part of the study area and sampled from 6 February to 9 March 2015. During this time, 19 species of carnivore were photographed.

The camera trap survey photographed 15 carnivore species less often in the 2015 wet season than in the 2014 dry season. This reduction was most noticeable for lion, serval, wildcat, white-tailed mongoose, slender mongoose and aardwolf. Several ecological explanations for these reduced photographic capture rates will be explored during the analysis of the survey. Similar to 2014, spotted hyaena and African civet were among the species most photographed, while zorilla and the side-striped jackal were among the least.

All leopard and spotted hyaena photographs from the southern section were identified to the individual level, using pelage patterns. In total, 21 and 19 leopards and 58 and 50 spotted hyaena were identified in 2014 and 2015 respectively. Of the leopard, 13 that were photographed in 2014 were photographed again in 2015. The project has commenced work on identifying all wild dog, serval and civet to the individual level.

In addition, 39 species of wildlife were photographed in the southern section of Study Area 1 in 2014. African buffalo, African elephant, hippo and warthog showed a large increase in their rate of capture when compared to 2014. Contrasted with this, several bird species showed a large decrease when compared to 2014, including doves, francolins, red and yellow-billed hornbill and guineafowl.

The children’s outreach programme has recommenced, with efforts focused on the community of Sankuyo. In February, a meeting was held with the chief, primary school teachers and parents to explain the programme and obtain permission. Following this, each weekend was devoted to a specific school grade level and a total of 76 children have participated in the programme and learnt about different field techniques, natural history and animal behaviour.

September 2013
I completed my pilot study in the summer of 2013. Upon arriving in Botswana, I worked with researchers from the Botswana Predator Conservation Trust, my local collaborator, to identify three focal sites.

A total of 25 camera stations where deployed in each study site for five weeks. More than 500 000 photographs were collected over 74 days of sampling. The cameras photographed all meso- and large carnivores known to occupy this region including aardwolf, African civet, bat-eared fox, black-backed jackal, caracal, cheetah, honey badger, leopard, lion, serval, spotted hyaena, wild cat, and wild dog. The animal data is being entered into an Excel spreadsheet for further analysis.

Eight spoor surveys were carried out within each study site. Over 200 km of transect was covered and spoor of all large carnivores was detected. Hyaena spoor was detected most often, while cheetah was detected least often. The data is being prepared for analysis using spatial autocorrelation occupancy models.

Eight distance sampling surveys for prey species were conducted. The GPS coordinates of the sighting, angle to the animal, distance to the animal, species, cluster size and habitat type was recorded. This date is being prepared for analysis in the programme DISTANCE. Following the analysis, prey density estimates from DISTANCE will be compared to camera trapping rates of prey. If they are strongly and positively correlated, this would suggest that camera-trapping rates of prey could be used as an index of prey densities.

To implement a long-term, sustainable monitoring programme it is imperative to involve and train individuals from local organisations. I met with several members of the Sankuyo Community Trust to discuss my research. Two wildlife guides from Sankuyo received intensive, hands-on training on how to implement my field techniques. When I return to Botswana I plan to have several field workshops for the Sankuyo Trust and the Department of Wildlife and National Parks to build capacity in regards to monitoring wildlife populations.

I also piloted a conservation programme for children from local communities. I worked with a wildlife guide from Sankuyo, named Monthusi “Sixteen” Sinvula to develop a programme that takes kids into the bush to provide them with positive wildlife encounters and expose them to natural history, animal behaviour, and field techniques. In August 2013, we took 42 children from the village into the field over the course of six days. This was the first time the majority of the children had ever seen lion, wild dog and leopard. We hope by exposing children to wildlife in their natural environment that we will inspire the next generation of Botswana to care about conservation.

May 2016
The first publication resulting from Lindsey’s dissertation research in Botswana was featured in the Research Highlights section of Nature, May 2016.
The full article can be found here: http://www.nature.com/nature/journal/v533/n7601/full/533011d.html