Evaluation de l'impact de l'exploitation forestière sélective sur la biodiversité mammalienne à l'aide de pièges photographiques et de capteurs acoustiques dans le Sud-Est du Cameroun

Considering the importance of mammals in regulating ecological processes in rainforests and the constant expansion of logging in Central Africa, it is essential to study the influence of logging on mammalian communities. This study aims to (i) assess the impact of selective logging on mammalian biodiversity and (ii) compare two methods of surveying elephant populations, one using camera traps and the other using acoustic sensors. The study took place at the SEFAC logging company; located in the South-East of Cameroon, north of the Lobéké National Park. Wildlife inventories were conducted in two sites with contrasting logging histories, the first having been logged until January 2021 and the second 17 to 45 years ago. In each of these, 12 camera traps and 6 acoustic sensors were set up on a 2 km2 sampling grid. Both devices were left in the field for 7 weeks. A total of 36 mammalian taxa were recorded in both study areas. The vast majority of species had comparable mean detection rates between the two sites (p > 0.05). Similarly, species richness, specific composition, and activity patterns of the most detected species differed little between the two areas. Regarding the second objective of this study, passive acoustic monitoring was characterized by a higher detection rate of elephants than camera trap surveys. The false positives produced by the acoustic data processing justify that the detection events of the two surveys do not follow the same trends from a temporal point of view. In conclusion, this study confirms the high conservation potential of sustainably managed forest concessions. The one studied has a species richness comparable to the one estimated in an adjacent protected area and higher relative abundances of wild bongos, chimpanzees and elephants. This work also shows the quick recovery of mammalian communities after logging as long as hunting activities are contained. By investing in wildlife management and anti-poaching, logging concessions can act as ecological corridors at a landscape scale. Moreover, this study demonstrates the usefulness of acoustic and camera trap surveys in rainforests, it also explains their advantages and disadvantages. Further studies would help identify which type of survey is most appropriate for monitoring elephant populations. Finally, this study highlights the need to develop standardized protocols and efficient automatic detection algorithms to optimize these two survey methods.