Amazonia, the largest extent of remaining tropical forest on earth, covers 12% of the planet's land surface, yet it hosts about two-thirds of all terrestrial animals. It encompasses the Amazon River basin and the Guiana Shield in South America, and is home to more animals than any other terrestrial landscape on the planet.
It's always difficult to spot wildlife in these dark and tense forests dotted with insects and spiny palms. This is because of the very nature of biodiversity in Amazonia, where there are a small number of abundant animals and a greater number of rare species that are difficult to observe adequately.
For ecology and conservation, understanding what species are present and how they relate to their environment is of great importance, since it provides us with essential information on human-made disturbances such as climate change, logging, or wood-burning. In turn, this can enable us to adopt sustainable human activities, such as selective logging, which involves cutting one or two trees and keeping the rest intact.
To overcome these obstacles and deepen our understanding of Amazonian wildlife, we are deploying a wide variety of technological solutions as part of BNPs Bioclimate project. These devices excel at allowing animals to go about their business unharmed.
The trees have their eyes set among the trees.
Camera traps are small devices that are activated by disturbances in their vicinity, such as animal movements. They have been essential to our fieldwork in the Tapajos National Forest in Para, North West Brazil, allowing us to investigate whether environmental events such as climate change have influenced the presence and behavior of animals, which are, in turn, necessary to natural processes.
Animals' dispersal of seeds, which allows for forest regeneration, is one of these processes. They will most likely excrete or drop the seeds elsewhere by eating fruits or carrying nuts.
We also know that many of these animals are heavily impacted by disturbance. To better understand the consequences of losing these seed-dispersing species, we need to know which ones spread which plants and how far.
We have tried to investigate this by installing cameras at the foot of fruit-bearing trees on our study site, revealing which species was eating which fruits and thus spreading seeds across the forest.
Over 30,000 hours of footage were captured, and 5,459 videos contained animals. There were 152 species of birds and mammals recorded, including rare records of endangered species such as the vulturine parrot (Pyrilia vulturina).
The videos included amazing glimpses into animal behavior, including an ocelot (Leopardus pardalis) after a common opossum (Didelphis marsupialis), a giant anteater (Myrmecophaga tridactyla) carrying an infant on its back, and a curious female tufted capuchin monkey (Sapajus apella) who discovered a camera and ended up throwing it onto the floor.
Importantly, we also recorded 48 species eating fruit, including species considered essential seed dispersers, such as the South American tapir (Tapirus terrestris), which is able to scatter large seeds over longer distances due to its size.
Our study demonstrated that bird species such as the white-crested guan (Penelope pileata) and mammals such as the Amazonian brown brocket deer (Mazama nemorivaga) are frequent consumers of fruits. Many of these species are overhunted in the study area, which may have significant effects on forest regeneration.
Forests are pulsing.
On the other hand, audio recorders are crucial to preparing in-depth records of the species-rich bird community. Despite their absence in dense forests, their vocalizations reveal their existence.
Because it is often logistically difficult to return to individual locations, ornithologists are limited in how often they can conduct counts. Traditional surveys are often of a quite long duration between 5 and 15minutes, with only a few repeat counts at each surveyed location. This means that only a small portion of the time period when birds are most active the two hours after sunrise, commonly known as the dawn chorus, can be surveyed.
Despite the fact that birds do not all sing at the same time, a few species prefer to sing very early in the morning, most wait until it is slightly warmer and the sun is fully out, and a few more rise late. Furthermore, surveys that are only conducted on a few days can completely alter which species are detected.
We found that by allowing autonomous acoustic recorders to record 240 very brief 15-second surveys totaling one hour of surveying, we could record 50% more species at each location that we surveyed, in comparison to four fifteen-minute surveys that replicated the duration of human surveys. Most importantly, we could record half of the dawn chorus as a result of the extra surveys.
With a total of just one hour of surveying at each location, we were able to detect 224 species of bird in 29 locations.
The diversity of species present in intact and disturbed forests has also confirmed our previous research that revealed that undisturbed, primary forests possess unique bird communities that are lost when forests are damaged by selective logging or wildfires.
With over 10,000 hours recorded so far, Acoustic recorders have allowed us to collect data over extended periods of time.
Collecting data on this scale also means that it is not feasible for a scientist to listen to all of the recordings. Instead, the new field of ecoacoustics has developed statistical techniques to characterize whole soundscapes. These allow scientists to process large volumes of acoustic data in an efficient manner.
We have used acoustic indexes to demonstrate that undisturbed primary forests have distinctive soundscapes that may be identified using machine-learning techniques. Such data allows us to compare soundscapes that have been disturbed by phenomena such as fires or logging and identify the species groups that have suffered the most.
Camera traps and acoustic recorders allow us to see and hear animals even when our researchers are not there. As technology develops, we will continue to utilize the latest tools to improve animal behavior and ecology.
We are particularly interested in utilizing deep-learning algorithms to identify species and, in some instances, to distinguish between individuals of the same species. Automated recorders are opening up new ways of assessing animal abundance and behavior, providing new insights into the mysterious world of tropical forest fauna.
As part of the Climate and Biodiversity Initiative program, the BNP Paribas Foundation sponsored Bioclimate, managed by the Rede Amazonia Sustentavel (RAS).
Oliver Metcalf, a postdocoral researcher at Manchester Metropolitan University, and Liana Chesini Rossi, an invited user at the University of Sao Paulo State.
This article is licensed under a Creative Commons license from The Conversation. Read the original article.