An Initial Checklist of Mammals of the A Rocha Dakatcha Nature Reserve, Kenya, 2022

Study Extent

All mammals were surveyed within and adjacent to the A Rocha Dakatcha Nature Reserve in an area c.7,000 ha in size.

Sampling

Sampling was carried out between March 2019 and December 2021. Camera traps were deployed for two months at a single location and then moved immediately to another for a further two months etc. Ten Audio-moth recorders were deployed on a very ad hoc basis for a total of just 10 nights, primarily to record nocturnal bird vocalisations within which mammal vocalisations were also captured and identified where possible.

Quality Control

The identification to species level was done by the authors using the most recent literature available (Stokes et al. 2015, Kingdom et al. 2016, Musila et al. 2019) and in case of doubt, external experts were consulted. Most species recorded in the checklist were photographed, allowing for external verification and reproducibility. The camera trap data have been published separately (Nussbaumer et al. 2019) and the photos can be accessed at http://n2t.net/ark:/63614/w12001317. The checklist was edited to align with Darwin Core (DwC) standards and the taxon identifiers were taken from GBIF.Open-refine (https://openrefine.org/) was used in cleaning the dataset before publication.

Method steps

1. The study area was divided into 140 survey grids each measuring 30 seconds by 30 seconds (approximately 900 x 900 m) thus aligning with lines of latitude and longitude. The aim was to survey each of the 140 grid cells in the classical ‘atlas’ method, and record all species encountered. The grid can then be queried for which squares any given species has been recorded and thus its distribution across the whole area described. Initially ten motion censored infra-red camera traps (Reconyx Hyperfire; https://www.reconyx.com/) were deployed to record mammal presence. Two cameras stopped functioning about a year into the sampling so the remaining eight were used for the remainder of the period. Up to ten acoustic monitoring devices (AudioMoth 1.1.0; https://www.openacousticdevices.info/audiomoth) were also used to record sounds from animals and complement the camera traps, particularly for nocturnal animals. For the camera traps, a general area was identified in which to deploy the cameras using the grids as a guideline and selecting grids which had not already been surveyed. Once a general location was selected, the camera would be taken to that area and a search made for a site along a trail which showed indications of mammals using it (footprints, droppings etc). A strong tree was selected beside the trail in a site where it had a good field of view on the path and where mammals clearly used it. The camera was then attached to the tree c.50cm above the ground facing what was considered the optimum direction to capture maximum mammal movement. The exact geolocation was recorded for each camera position using a GPS-enabled mobile device and the deployment details recorded using Kobo Collect. The camera traps were set to trigger with infrared motion sensors to take multiple pictures at each trigger. After every month of deployment, camera traps were checked to ensure they functioned properly, batteries replaced and adjusted if needed, and left for another month. After the second month, images captured by the camera traps were retrieved and analysed. The camera traps were then re-deployed to other sites using the atlas grids as a guide. The acoustic monitoring devices were deployed in 25 locations to collect mammal sounds, especially from the nocturnal canopy-dwelling and illusive species. These were located in patches of forest where the Sokoke Scops Owl (Otus ireneae) was expected to occur and thus mammal sounds captured opportunistically. The devices were left for two nights before retrieval. Captured images and sounds were identified and confirmed by mammologists. It was then left to run for two months with a single intermediary service check to replace batteries. After two months, images were retrieved, labelled, and carefully stored on a hard drive as well as uploaded to Wildlife Insights (https://www.wildlifeinsights.org/) for assistance in image management, identification of subjects and mapping.

All records are in the Class Mammalia. 25 were identified to species level and one record identified up to subspecies level.

1. Bdeogale omnivora
   rank: species
2. Caracal caracal
   rank: species
3. Cephalophus harveyi
   rank: species
4. Cercopithecus mitis subsp. albogularis
   rank: subspecies
5. Chlorocebus pygerythrus
   rank: species
6. Civettictis civetta
   rank: species
7. Cricetomys ansorgei
   rank: species
8. Crocuta crocuta
   rank: species
9. Genetta maculata
   rank: species
10. Helogale parvula
    rank: species
11. Hystrix cristata
    rank: species
12. Ichneumia albicauda
    rank: species
13. Leptailurus serval
    rank: species
14. Madoqua kirkii
    rank: species
15. Mellivora capensis
    rank: species
16. Neotragus moschatus
    rank: species
17. Otolemur garnettii
    rank: species
18. Panthera pardus
    rank: species
19. Papio cynocephalus
    rank: species
20. Paraxerus ochraceus
    rank: species
21. Paraxerus palliatus
    rank: species
22. Petrodromus tetradactylus
    rank: species
23. Potamochoerus larvatus
    rank: species
24. Rhynchocyon chrysopygus
    rank: species
25. Sylvicapra grimmia
    rank: species
26. Tragelaphus scriptus
    rank: species

The data were collected in and around the A Rocha Dakatcha Nature Reserve (ARDNR) which is a small block (c.7,000 ha) of the larger Dakatcha Key Biodiversity Area (KBA) which is around 180,000 ha in size. The central point of the Dakatcha KBA is located c.45km to the northwest of Malindi town (S-2.884° E39.833°) which is more-or-less where the ARDNR is also located within the KBA – i.e. centrally.