Sounds of the Pantanal - The Pantanal Automated Acoustic Biodiversity Monitoring Program
This study is part of the Instituto Nacional de Ciência e Tecnologia em Áreas Úmidas (INAU / CNPq / UFMT ), Cuiaba, Mato Grosso, research program on the Pantanal wetlands, coordinated by Prof. Dr. Wolfgang Junk, Max-Planck Institute, Plön, Germany, and Prof. Dr. Paulo Teixeira, Federal University of Mato Grosso, Cuiaba, Brazil.
The main goal of this project is the study of sound-emitting animals as bioindicators for habitat quality and conservation status. The study will establish a monitoring approach in different Pantanal ecosystems to improve our knowledge on local biodiversity which is essential for the sustainable use of natural resources and for the implementation of appropriate conservation measures. For this purpose an automated remote monitoring system will be established for long-term surveying of grasshoppers, amphibians, birds, and mammals. Specifically for birds, migratory species will be evaluated as potential bioindicators for aquatic and terrestrial habitats. Besides, a baseline data set will be gathered on the aforementioned taxonomic groups, using conventional survey methods and automated recording units in order to determine appropriate indicators. Likewise, animal communities affected by water level cycles (cyclic inundation) will be analysed.
Major research questions
(1) How well suited are migratory bird species to serve as bioindicators for aquatic and terrestrial habitats in comparison to resident species?
(2) How do avian long-distance and short-distance migratory species ecologically fit into the avian resident assemblages?
(3) Which species, species groups or biodiversity integrity indices are the best indicators for certain conditions and changes in vegetation cover, land use, and climate?
(4) How accurate and precise are the data collected by the automated systems in comparison to those obtained by conventional survey techniques (e.g., mist netting and audiovisual methods)?
(5) Which is the current composition of the aforementioned vertebrate and invertebrate communities in natural (aquatic and semi-aquatic) Pantanal ecosystems?
Throughout our 3-year study which started in October, 2012 we will perform inventories of grasshoppers, anurans, birds, and mammals. Besides conventional avian survey techniques and mist netting to obtain measurements of morphometric and body mass characteristics, we will build up a digital sound archive, containing the vocalizations of the aforementioned vertebrates and insects. The CO.BRA team will use commercial available recordings, online sound collections, and new recordings made during the surveys and by the automated sound recorders. The sound archive is necessary to develop and evaluate the automated sound-identification software for the remote biodiversity monitoring system that we will develop in the scope of the project.
Figure 1. Field record station with night flight weatherproof microphone (Wildlife Acoustics, USA), right top.
Figure 2. Recording device (Song Meter SM2+, Wildlife Acoustics, USA).
In the case of migratory birds we will take blood samples from selected individual species. This is necessary to identify the genetic pool, i.e., the geographical origin (Neartic vs Neotropic) of these taxa by comparing their DNA sequences with existing reference data (gene banks). This service will be provided by the UFMT, Laboratory of Genetics, and the University of Primorska, Department of Molecular Sciences, Slovenia.
In order to characterize the community structure of avian migrants, such as their taxonomic composition, body mass categories, and guild structure (time / space-activity patterns, territorial and foraging behavior, habitat, and functional types), traditional survey methods are mandatory.
The distribution pattern and abundance of non-resident birds in different Pantanal ecosystems will provide information on the influence of the habitat status and connectivity, land-use strategies, and possible climatic impacts on migratory birds.
The methods used in this study are expected to allow for a comparative analysis of the efficiency of survey techniques (audiovisual surveys vs. automated systems). For the metadata analysis we will employ a multivariate and regression statistics.
The technology development effort in the Sounds of the Pantanal Biomonitoring Project aims to design, deploy, and evaluate an integrated communication framework for the automated acquisition, transmission, storage, and processing of audio and image data. The resultant data collection and processing system will allow for long-term monitoring of biodiversity indicators in the Pantanal region. The main purpose of this technology is to facilitate a cost-efficient method to assess biodiversity in various habitats types of the Pantanal.
In the first two years of the project we conduct:
(1) Research and development of software for automated detection of avian, mammalian, anuran, and insect species from their sound emissions.
(2) Design, construction, deployment, and evaluation of autonomous multi-sensor recording stations (hardware and software) for remote monitoring of various ecosystems. These remote monitoring stations will be self-powered by solar cells, and will record audio, environmental parameters (temperature, humidity, illumination, wind speed, and direction), and images. Subsequently, this information is compressed and transmitted to a base station at UFMT for automatic statistical analysis.
(3) The development and implementation of the 'Pantanal Biodiversity Database', which will integrate multiple data resources, such as audio, environmental, and image data. Database management tools will be develop in support of the transfer of field information to the database.
(4) Design of an appropriate interface for data access and management.
(5) Extension of the data repository by integrating additional metadata on mammalian, anuran, and insect species. The long-term goal is the creation of a representative database, which would allow for long-term statistical analysis of biodiversity composition and population trends.
(6) Research and development of software for large-scale automatic recognition of avian, mammalian, anuran, and insect taxa by their sound emissions.
Figure 3. Process features of sound analysis procedures.
The technology developed in the Pantanal Biomonitoring project (2012-2015) will provide the basis for future advances of automated biodiversity monitoring in the Pantanal wetlands. In particular, incremental research and technology development activities would increase the functionality of the remote biodiversity monitoring system with:
(i) Population density estimation for terrestrial species (birds, bats, anurans, and insects).
(ii) Underwater data recording stations for automated acquisition of multichannel audio and aquatic environmental data.
(iii) Automated detection of sound emitting fish species.
(iv) Statistical analysis of the correlation between species activity and environmental data recorded by the remote monitoring stations. With the increasing accumulation of data and human expertise, this software will evolve to an expert system for natural resource managers and conservation practioners in the Pantanal region. Thus, the remote biodiversity monitoring system developed during our study could be easily adapted for the use in other areas.
(A) Determine which avian species groups are best suited to be monitored with remote acoustic monitoring techniques (e.g., migratory vs resident species).
(B) Contribute to INAU's 'Pantanal habitat classification system' by identifying indicator species groups associated with specific wetland habitat types.
(C) Establish bird-based indicators for different levels of environmental change, such as fluctuations in water levels, habitat alteration, land-use strategies, and climate change.
(D) Establish a sophisticated automated monitoring system for avian species, which will facilitate the long-term observation of changes in the conservation status of remote tropical ecosystems, using appropriate bio-indicators.
(E) Comparative data obtained with conventional survey methods, will allow us to calibrate and optimize the automated approach and demonstrate its strengths and limitations.
(F) Identify the geographical origin of the avian migratory species by performing genetic studies on blood samples taken of individual birds using existing gene bank data as a reference. This service will be provided by the UFMT, Laboratory of Genetics, and the University of Primorska, Department of Molecular Sciences, Slovenia.
(G) The study will provide insights into the temporal and spatial distribution of avian migrants from the Nearctic and Austral Regions within the Pantanal wetlands. These birds will be identified acoustically, visually, and genetically.
(H) The project will contribute to our understanding on how the avian migratory community composition is correlated with certain habitat characteristics, land use strategies, and climate settings, topics that will be addressed in coordination with other INAU projects.
(I) Assessment of edge effects and habitat connectivity (e.g., biological corridors, gradual ecotones).
(J) Transfer of technology and knowledge by involving our Brazilian partners and their students during the establishment and maintenance of the automated monitoring units and in the re-collection of comparative biodiversity data with conventional survey methods.
(K) International graduate courses on the theory and field practice of these methods will be held for students from other Brazilian and European universities.
(L) Providing guidelines for ecotourism with special concern of sensitive wildlife areas.
The establishment of automated remote monitoring stations is a considerable technical challenge. Thus, the optimization of the system will continue beyond the timeframe of the project, particularly regarding the identification abilities of sound-recognition software and the wireless real-time transfer of large amounts of sound and video data.
It is the aim of this study to include additional Pantanal ecosystems and to cover those already included with a higher number of automated monitoring stations in order to assist habitat conservation and restoration efforts and to observe the impacts of climate change on avian migratory species and resident bird communities. We will provide bioindicator-based data for the development of conservation and management plans and improved predictions for the impacts of wetland changes (natural and anthropogenic) on avian migratory species and on other important groups of wetland-dwelling organisms. This study could foster scientific impacts and interdisciplinary cooperation with UFMT Institutes, the Pantanal Research Center (CPP), and the National Wetland Institute of Science and Technology (INAU), and may even result in the development and foundation of patents and spin-off companies.
In the proposed project we will use the most sophisticated methods for audiovisual biodiversity monitoring, including automated remote recording units (sound loggers, wildlife cameras), to determine and survey appropriate vertebrate-based bioindicators for the conservation value of the studied ecosystems, facilitating the assessment of impacts of land-use strategies and climate change. In addition, game birds (e.g., guans and curassows) and larger mammals (e.g., monkeys, peccaries, and tapirs) are good indicators for hunting pressure in natural landscapes. Although, less well studied as a group, arthropods and anurans are known to be very sensitive to environmental disturbances. Not surprisingly, nearly one-third (32%) of the world's amphibian species are already threatened with extinction, making the medium-term inclusion of amphibians in our automated acoustic monitoring scheme a top priority.
Figure 4. General application overview of the animal sound detection and recording by automated recorded devices.
In this project the training of undergraduate, M.Sc., and Ph.D. students from Brazil and abroad in sophisticated bio-monitoring methods, including advanced analytical software, is one of the major goals of educational scientific transfer during the period of joined research with other INAU Pantanal field projects. Furthermore the participants of this project will be engaged in seminars and workshops to be held at the UFMT in Cuiabá (e.g., annual international tropical ecology courses with topics covering edge effects, life histories, biodiversity, and biogeography of Pantanal ecotones). The first course with such a topic was held in 2011, 2012, and will be continued in 2013, involving university professors from European Union institutions, particularly members of the International Consortium for Biodiversity Research (IBISCOS) of which the UFMT is an associate.
Through joint supervision of undergraduate, M.Sc., and Ph.D. studies our Brazilian partners will be involved in all phases of the project, particularly in the data collection and the database repository development, the research and technology development, deployment, and evaluation related to technical methods, and in the research and software development for automated acoustic detection of species. This project will strengthen the long-term scientific collaboration between Brazilian and European institutions.
The scientific collaborators will participate in the teaching of Undergraduate, Master, and Ph.D. students through courses in the fields of computational bioacoustics, automated sound detection and sound analysis, statistical design of monitoring studies, and the analyses of faunistic data for conservation purposes.