Digging Deep: How animal soundscapes reveal levels of Amazon forest degradation

A study published in PNAS this month finds how biodiversity is being impacted as a result of the degradation of the Brazilian Amazon rainforest. Naturally, any on-ground biodiversity survey brings with it its own challenges and constraints. To augment this problem, Rappaport et al. (2022) monitored sounds coming out of the forest in different pockets and examined how well that dovetailed with the acoustic niche hypothesis.
The acoustic niche hypothesis (ANH) says that in a well-established ecosystem, different species tap into different bandwidths in order to avoid competition. So, each species sings at a unique pitch/frequency at unique times. However, in durbed ill-formed ecosystems, a particular bandwidth is tapped into more than one species while other bandwidths frequencies may have been left completely un-utilised.
‘The more degraded an ecosystem is, we can expect to see “holes” in the soundscape that represent local species extinctions, and more overlapping signals due to less time for adaptive/evolutionary processes for differentiation into dinct niches,’ says Danielle Rappaport, the lead author of the study, in an email with . The ANH argues that there is a positive, linear relationship between how intact a habitat is (intactness being measured aboveground biomass ie the trees etc) and the acoustic space occupancy the faunal community.
‘An increase in ASO means that more “acoustic channels” are utilized coexing species in a habitat. In other words, species are emitting sounds in a greater range of hours and frequencies during the 24-hour cycle,’ Rappaport added.
This is similar to the concept of resource partitioning or niche differentiation that is often observed in terms of food. Two or more species that devour a certain fruit will eat different parts of the fruit, or search for food at different times of the day.The team of researchers from University of Maryland measured the number of bandwidths tapped-into after three different stages of degradation: once burnt, frequently burnt, and logged. The acoustic measurements were made for each of the three sites at 1 minute and 1 hour timesteps – and, quite naturally, these soundscapes are affected degradation brought about logging and fire.

However, the patterns of change do not follow the predictions laid out acoustic niche hypothesis consently, ‘instead, acoustic analyses revealed contrasting impact of fire and logging on acoustic community structure,’ the study reports. First of all, daily ASO increased with aboveground biomass after fire, but not after logging.
The dinction in ASO in the burnt setting was largely driven insects, for it was during the insect-dominated durations of the day that the change in percentage ASO was most significantly observed. For logged areas, the only time intervals that did show a change in percentage ASO ended up showing a negative relationship i.e. the more the time that elapsed since logging, the more the decline in ASO.
The time periods when insects dominate – mid-morning, noon and night – are rarely sampled in most surveys, as they largely focus on birds. In fact, this particular study did not notice much difference in ASO during dawn and dusk, the time periods dominated birds.
In forests that were burnt recurrently, ASO was the lowest, and occupied the smallest amount of frequency space across all time periods. Compared to logging, ‘with recurrent fire, we see a few animal signals dominating the 24-hour soundscape, as opposed to a more differentiated soundscape that represents a more diverse animal assemblage,’ says Rappaport.

Furthermore, the paper states that in once-burnt and logged forests, one still observes frequencies below 3.7 kHz and above 5.2 kHz. But frequencies outside the midrange (3.7-5.2 kHz) are less prevalent. Apart from quantifying the total ASO, the team also evaluated the interactions and patterning of individual sound signals. Rappaport adds that ‘In the case of recurrently burned forests, the animal communication networks over the 24-hour cycle are more homogenous from day to night (fewer peaks and dinct choruses as you might expect when you think of a tropical rainforest soundscape).’
What could possibly explain this contradiction between Rappaport et al.’s study and the acoustic niche hypothesis? Why is it that, ‘contrary to the hypothesis’s predictions, less intact forests did not contain more soundscape gaps (i.e. empty acoustic niches)’?
One could be that acoustic niche partitioning is a product of long term evolution; and when there is a durbance, such as fire or logging, acoustic niche differentiation loses its relevance. The other reason could be that the study did pay heed to the time periods when insects were active, and they tend to have more nuanced responses to durbances. For instance, cricket species are very good at partitioning frequencies, whereas cicada species tend to have more frequency overlaps. Therefore, cicadas have evolved to live and transmit sounds at different heights, making them more vulnerable to temperature changes, such as those stemming from fire. The resultant soundscape in burnt forests is, therefore, a consequence of an intricate interaction between forest structure, microclimate and animal habitats.
Overall, the study highlights that the potential eco-acoustic studies have to monitor biodiversity change in the event of degradation, for ‘competition for acoustic space is the strongest.’ For example, this study allowed researchers to realise links between habitat conditions and animal communication other than that of just birds. In future, studies measuring acoustic differentiation can be clubbed with satellite measurements and ground-based assessments to shed light on biodiversity responses especially in the hyper diverse tropics (see, for e.g., Aide et al. 2017 or Planque et al. 2008), the team argues.
The author is a research fellow at the Indian Institute of Science (IISc), Bengaluru, and a freelance science communicator. He tweets at @critvik