From Chaos to Clarity: Making Sense of Citizen Science in Oil Palm Landscapes

By Emily Meijaard, Communications Officer, Borneo Futures

When we think of palm oil, images of displaced wildlife and destroyed habitats often come to mind. But what if agriculture and wildlife didn’t have to be at odds? With the right approach, they can co-exist and even thrive.

Research shows that in well-managed agricultural landscapes, wildlife populations can be sustained or even increase. That’s why understanding how animals move, where they live, and how their populations are changing is critical to developing adaptive management strategies that mitigate the negative impacts of palm oil plantations on biodiversity.

One example of this in practice comes from Indonesia, where plantation workers are playing a key role in tracking and protecting wildlife, including the critically endangered Orangutan. This latest study explores a citizen science initiative under the PENDAKI (Care for Biodiversity) programme, launched in 2019 by Indonesian palm oil company Austindo Nusantara Jaya (ANJ). The programme empowers workers to record sightings of wild animals and plants during their routine tasks, creating a rich and growing database of local biodiversity. Of central importance to this initiative is the monitoring and management of Orangutans, whose chances of survival are significantly impacted by oil palm expansion.

The palm oil industry has largely responded to this threat with increasing willingness to reduce biodiversity loss. Measures such as High Conservation Value (HCV) set-asides and broad-scale monitoring initiatives like PENDAKI have already demonstrated the benefits of maintaining intact forest patches within plantation areas, particularly for forest-dependent species like the Orangutan.

Although citizen science is an increasingly popular data collection method, one common critique is that data collected informally, through casual, unstructured observations, can be hard to analyse. Without standardised survey methods, it is often unclear whether a species wasn’t seen because it was absent or simply missed.

To tackle this, researchers applied a statistical tool called occupancy modelling. These models estimate the proportion of areas occupied by a species and account for the possibility that a species was present but undetected. This makes them especially useful when detailed survey data is lacking.

Surprisingly, the study found that even with the "messiness" of unstructured data, occupancy estimates remained reliable. The models revealed important patterns. Species like Orangutans (Pongo pygmaeus) and Gibbons (Hylobates albibarbis) had much higher occupancy in forested areas and near large forest patches. In contrast, the Water Monitor (Varanus salvator) was found to occupy nearly the entire study area, with a striking 97 percent mean occupancy.

Still, there were challenges. For many species, the precision of estimates was low, largely because most observations included only one or two species. These "short species lists" make it harder to assess how often a species is missed. The study also highlighted how observer differences, such as varying levels of skill or experience, affected species detection.

Improving precision means gathering more complete data. To that end, a subset of observers is now using fixed species lists and a smartphone app that logs sightings along with location data, boosting both data quality and insights into survey effort.

While the current models don’t yet account for every spatial or temporal variable, such as changes in observer skill or fine-scale habitat features, they represent a promising step toward more effective biodiversity monitoring in tropical plantations.

Ultimately, the goal is to integrate citizen science data and occupancy modelling into day-to-day biodiversity management. With more data and better tools, scientists can sharpen occupancy estimates for current study species, expand the species list, and zoom in on finer spatial scales, helping conservation and agriculture work hand in hand. Link habitat fragmentation study