blog

Kevin Darras, University of Goettingen, talks about his fieldwork in Berbak National Park, Indonesia, recording the dawn chorus to survey the bird communities in tropical peat swamp as reported in a Research Article published on F1000Research.

In 2012, I was carrying out research in the Jambi province of Indonesia as part of my PhD in the Ecological and Socioeconomic Functions of Tropical Lowland Rainforest Transformation Systems Sumatra (EFForTS CRC 990) project in agroforestry plantations like rubber, palm oil, and other forest sites. We soon realized that our forest sites were secondary, degraded forests, which may not represent a pristine habitat to host diverse bird communities.

We were interested in finding out what primary forest sites and other land use systems looked like; whether there were more birds and what species made up the communities.

The Berbak National Park is not far from Jambi city, but it is difficult to access, so it is still relatively well preserved. What’s more, the ornithological studies in that forest date back to 1987, so we wanted to update them.

Luckily, some NGOs were doing research in this National Park, so we decided to collaborate with the Zoological Society of London (ZSL) to avoid wandering on our own into new territory without proper support from experienced people.

Wading through black water

We were usually out in the field for one week at at time for each recording survey. It involved a lot of preparation from buying food and supplies, to selecting potential areas where we would find our target habitats, and organising transport by car and speedboat.

Sometimes we would enter the park through the Air Hitam river on a large motorboat with a cabin and deck, and for the surveys near to Seponjen village we would use four-person canoes. Finding the right sites involved getting wet all the time and wading through black water while trying not to get the expensive sound recorders damaged.

It rained every day, and at night we either slept in a field house inside the National Park, on the motor boat, in a stilt hut or in a bivouac (improvised campsite), depending on the location and the conditions – the large motorboat broke down during the study period. While the recorders were on, we accompanied the ZSL crew for their own surveys about tree phenology or primates.

The dawn chorus

Sound recorders have many advantages. We needed simultaneous recordings to disentangle the effect of the location from the effect of the survey time, which influences which birds will be vocalizing, and we needed dawn recordings. Our sites would be too difficult or too dangerous to reach in time for point counts. We also wanted to collect as many data as possible within a short time period as this was only a side study.

The audio clip is from a recording in secondary forest habitat. The bird calls featured include the following species: Mixornis gularis, Prinia flaviventris, Anthracoceros albirostris, Pericrocotus flammeus

With normal human point counts to do simultaneous surveys, we would have needed three ornithologists, as many boats, and many more crew members to make it happen, and there would still have been a sampling bias from the different people. Generally, visual observations  are very rare in dense forest sites, such as these, so we expected minimal loss of data with our sampling method.

SoundEFForTS

After recording the ambient sound from all sites, I extracted the dawn chorus from each one, and we hired ornithologist Irfan Fitriawan to identify the birds within. We use our own internet platform SoundEFForTS to archive the recordings so that they are available to anyone from almost anywhere.

The audio clip is from a recording in shrub swamp. The bird calls featured include the following species: Rhipidura javanica, Copsychus saularis, Pycnonotus plumosus, Prinia flaviventris, Centropus sinensis

The recordings can be displayed as spectrograms and played back online, and individual bird calls can be tagged directly with species names. At the end of the day, I extracted the data in the form of a data table and got on with the statistical analysis.

Old World Babblers

We were surprised to find out that all three habitats (primary swamp forest, secondary swamp forest, and shrub swamp) harboured just as many individuals and species. The communities were also functionally very similar based on the bird feeding guilds, so the only difference was in species identity. The forest had more Timaliidae (Old World Babblers) like Stachyris and Malacopteron species, also blue-rumped parrots and long-tailed parakeets (all Near Threatened in the IUCN Red List of Threatened Species). Rhinoceros hornbills (Near Threatened) were also recorded in the bush swamp, but this habitat was populated by more common species in terms of distribution area, like the Lesser Coucal and the Crested Serpent Eagle (both Least Cconcern).

Ornithologists wanted

So far, we have  not detected many threatened species, probably because of the relatively low sampling effort of 40 minutes per site. However, we have only processed 0.7% of the available data. With much more data yet to be analysed we hope to attract potential ornithologists to process more of our sound recordings.

This will help to create a more complete picture of the bird communities living in the national park. Actually, this is one of the reasons we chose F1000Research, because if we get more data, we will be able to publish a second, improved version of the article.