Empowering indigenous wisdom: co-creating forest inventory through citizen science in Royal Belum State Park, Malaysia

Democratising science and bridging scientific data gaps with the reduction of emissions from deforestation and degradation (REDD+) framework to strengthen the conservation capacity of forest communities requires empowering the indigenous people with the necessary technical competencies. Considering the value of indigenous knowledge, involving local communities in scientific data collection might benefit science and society. Nevertheless, the output quality has raised scepticism. This study aims to empower indigenous wisdom in citizen science by co-creating forest tree inventories in the Royal Belum State Park in Perak, Malaysia. The objectives are to identify the potential of involving indigenous communities in remote locations in scientific tree data collection, a locally-based tree census for carbon assessment, and to demonstrate the feasibility of location-based mobile smartphones in forest tree inventories. An established tropical forest protocol was adopted to estimate the above-ground biomass (AGB). A pantropic allometric model was employed, with tree diameter at breast height (DBH) and height. Unsupervised tree naming and tree positioning data collection was also performed with a smartphone. The information provided by the indigenous communities and professionals was compared. Trees that were taller and recorded bigger DBH demonstrated the lowest root mean square error (RMSE) of 0.965 cm and 4.616 m, respectively. Although the DBH measurements were of sufficient quality, the tree heights and names were subpar due to the different strategies employed. The mean absolute error (MAE) of the absolute positions recorded with the smartphone in relation to traverse surveying was 4.907 m (northing) and 5.817 m (easting). The percentage of trees that fall within the 20-m and 30-m quadrants is 67% and 79%, respectively. The results revealed that location-based smartphones might not be the best equipment for precise tagging individual tree positions. The equipment required established quadrants to confirm the tree locations. Nevertheless, the findings could contribute to Malaysian indigenous forest communities' ability to meet the REDD+ forest conservation targets, which is of international interest. The data procured in this study could also address the challenges of implementing locally-based carbon stock assessments in remote forests.