Learning to Listen: Understanding the Population Health of Threatened Amphibians
Last year, I participated in the Southern African Red List process for the first time. As an ecologist, the experience was both eye-opening and unsettling. The methods used to determine conservation status are meticulously researched, rigorously defined, and defensible. Yet, they still left me uneasy, specifically in how the Red List status of a species is assigned based on its geographical occurrence using two key measures: Extent of Occurrence (EOO) and Area of Occupancy (AOO).
Extent of Occurrence (EOO) refers to the total area within which a species is known to occur, forming a broad outer boundary around all known populations. It provides a measure of the potential geographical spread of the species but does not consider habitat availability or fragmentation within that range. Area of Occupancy (AOO), on the other hand, is a finer-scale metric that quantifies the actual area occupied by the species within the EOO. This is determined by overlaying a standardised 2km by 2km grid over known localities and counting the occupied grid squares. The issue with AOO is that it requires confidence that all surveys have been exhaustive, leaving little room for undiscovered populations. EOO, while useful for mapping a species’ broad distribution, can be misleading when used to assess species that are highly localised and dependent on specialised habitats.
Take, for example, a hypothetical amphibian with five known sub-populations. To determine its AOO, a grid is placed over these localities, and the occupied squares are counted. However, one must be absolutely certain that surveys in the broader landscape have been exhaustive and that no additional populations remain undiscovered. In practice, no ecologist will claim absolute certainty in this regard. The preferred Red List methodology therefore tends to rely on EOO, drawing a boundary around known populations and using this to assess conservation status.
This approach nearly led to the downlisting of the Long-toed Tree Frog (Leptopelis xenodactylus) from Endangered to Least Concern. Over recent years, extensive research has led to the discovery of "new" (i.e previously unknown to science) populations, which, when mapped as an EOO, suggested that the species was widespread enough to no longer warrant concern. However, when assessed using AOO, the species still met the criteria for an Endangered listing. Fortunately, the Long-toed Tree Frog had the benefit of detailed surveys conducted by outstanding researchers, and it was defensible to use AOO in its assessment. But this case highlighted a troubling question for the Anura Africa team: What happens when a species is both range-restricted and a habitat specialist, living in fragmented landscapes increasingly altered by agriculture and forestry?
The challenge with EOO-based assessments is that they can create the illusion of security, especially for species that persist only in isolated, fragmented pockets of habitat. While the mapped extent of occurrence may appear broad, the reality for any habitat specialist is very different. A species might only survive in select microhabitats within this range, meaning its actual viability is far more precarious than the EOO suggests. This discrepancy between EOO and AOO raises serious concerns about the accuracy of Red List classifications for such species.
Species that are range-restricted exist in small, finite areas. This is not necessarily because their habitat type is rare but because they require very specific microhabitats and microclimates. A habitat generalist, like the Clicking Stream Frog (Strongylopus grayii) for example, can thrive in a variety of environments, from pristine mountain streams to farm dams and even sewerage pipes in urban areas. Habitat specialists, however, are very different. Their survival is tied to highly specific ecological conditions, and any disruption—however small—can have devastating consequences.
When landscapes are transformed by human activities, habitat specialists suffer the most. A wetland may appear intact at a landscape scale, but if its hydrology is altered, its function may be irreversibly changed. This is where the concept of extinction debt comes into play. A population may persist for several years after a habitat has been degraded, giving the illusion of stability, but without the right conditions, it will eventually decline and disappear.
Historically, species have been able to respond to environmental change by dispersing to new areas. But in a fragmented landscape—where towns, cities, roads, plantations, sugarcane fields, and pastures dominate—the remaining patches of natural habitat become islands in a human sea. This is where island biogeography becomes relevant. When a species is isolated in small pockets of habitat, even minor changes can push it toward local extinction. The problem is, we may not detect these declines until it's too late.
The lesson we took from the Red List process was clear: we need to incorporate population studies into all of our conservation projects. Something that has not been adequately done for most amphibian studies, because collecting such data is often challenging. Before declaring a species "safe," (really nothing is truly ‘Least Concern), we must first determine whether its populations are stable. Conservation decisions must be based on actual population trends, not just mapped distributions. And the best way to do that? Ask the frogs themselves.
That’s how Learning to Listen was born.
We adopted an advanced acoustic spatially explicit capture-recapture (aSCR) methodology to study the Long-toed Tree Frog. This approach involves setting up six microphones around a calling population and recording six synchronised audio channels. By precisely measuring the distances between microphones, we can use computational models to differentiate between individual frogs and estimate population size.
On paper, it sounds simple. In reality, it’s anything but. Conducting fieldwork in wetlands—often in remote, difficult terrain—requires immense effort. Data processing is equally gruelling. Audio recordings must be meticulously analysed to filter out false positives before being imported into statistical software, which can turn even the most powerful computer into an overworked brick if not optimized for processing. But the reward is immense: we can generate accurate population estimates and track trends over time.
Bioacoustics doesn’t just tell us how many frogs are calling—it provides crucial insights into wetland health. By monitoring amphibian populations over time, we can assess the impact of conservation interventions. If population numbers decline, we can investigate causes and implement corrective actions such as adjusting burning and grazing cycles, restoring wetland hydrology by closing drainage ditches, and controlling invasive alien vegetation. Then, we sit back and listen again.
Over time, this data will help us establish science-based management guidelines that can be adopted by willing landowners, farmers, and plantation managers. These stewards are critical allies in conservation, and by providing them with practical, evidence-based recommendations, we can ensure that threatened amphibians continue to thrive on their properties.
During Red List discussions, I sat at dinners where conversations revolved around whether we were failing the species most at risk—the range-restricted, the habitat specialists, the ones now trapped on ecological islands. We took that message to heart. At Anura Africa, we have committed to advocating for these species by listening to them.
By learning to listen, we are not just monitoring populations—we are giving amphibians a voice in their own conservation. And by doing so, we are ensuring that their calls will continue to echo through the wetlands of South Africa for generations to come.
Comentarios