Spinomantis aglavei (Photo: Robin Hoskyns)
Surveying amphibians in tropical environments traditionally involves undertaking visual / acoustic transect surveys. This involves surveyors walking a set out transect and recording any individuals they see or hear (depending on the methodology).
Whilst this method is tried and tested, detecting cryptic species, such as those that live underground or in the forest canopy is challenging. This is especially true for poorly known species as their activity patterns are unknown and it may be that surveys are undertaken when they are not active.
One piece of technology that can help with detecting such species is remote bio-acoustic recorders. Detecting species via remote recording of calls is not new and has been used for frequently for birds, bats and marine mammals. Its use for amphibians has long been promoted but, with some exceptions, not been widely implemented in diverse tropical environments.
Madagascar is one such environment and home to over 300 species of amphibians, of which 45% are threatened with extinction. Despite a number of researchers working on amphibians in the country there is a lack of long-term monitoring of amphibian populations. Furthering the use of static acoustic recorders for monitoring amphibians is a key action in the New Sahonagasy Action Plan 2017-2020.
Importantly this project will also allow the feasibility of using remote acoustic recorders to monitor amphibian populations in local NGO context such as Madagascar, as whilst it may seem straightforward there are a lot of factors that need to be considered. Can the recorders be charged sufficiently for use in a rural setting? Can the recorders handle the humidity and other environmental factors of a tropical forest? Can you store the huge quantities of data on basic IT setups? Can the data be analysed at a local level?
These last two points are particularly important as acoustic recorders can – depending on the schedule – record hundreds of Gigabytes and hours of worth data. Also whilst analysis programmes will check recordings for species of interest, the models to enable have to be built first which involves a human listening to the recordings. This takes time especially when there at least 13 species likely to be present!
With expert help from ex-Mitisnjo amphibian conservation Director Devin Edmonds we are in the slow process of building these models. Even at this preliminary stage ten species have been identified from the calls including one surprising call. It is still early stages but as it continues the process will hopefully speed up and can bring you more updates.
This work is being undertaken through a grant from the Riverbanks Zoo and Garden Conservation Support Fund. We are using Arbimon Acoustics recorders and analysis platform from Sieve Analytics (www.sieve-analytics.com).