Amazonian bird species more vulnerable to climate change than previously thought, new study finds
Spatial resolution of available data can drastically alter estimates of species vulnerability to climate change
Species are facing an uncertain and rapidly changing future where climate change threatens to push their habitats past the tolerable limits, leaving them with no choice but to move, adapt, or die. The ecological and conservation implications are tremendous, and accordingly, predicting the vulnerability of different species to climate change is a hot topic among researchers.
But, an underappreciated variable in these studies – with a potentially large impact on the veracity of the results – is the spatial resolution of the data used.
A new paper by BGC Center research associate and former PhD student Muyang Lu and Center Director Walter Jetz, published last month in Global Change Biology, reveals that the choice of spatial data resolution, or grain size, in climate change vulnerability studies matters quite a lot: using different resolutions within the same analysis leads to substantially differing estimates of species vulnerability, meaning recent studies using coarse-grain data may be misidentifying up to half of the most vulnerable species.
“Coarse-grain data – commonly used in global climate vulnerability assessments – can seriously misrepresent species’ extinction risk, especially for tropical species,” said Lu, now Assistant Professor at Sichuan University. The consequences of these biased analyses could affect not just our ecological understanding, but also the delegation of conservation resources to protect the vulnerable species. “Conservation resources are limited, so accurate vulnerability assessments let us prioritize protection for species most imperiled by climate change, ensuring efficient resource use.”
Put simply, a basic climate change vulnerability analysis involves two key aspects: the species’ environmental niche, or the set of temperature, precipitation, and other conditions tolerable for the species, which is revealed by observing the average climatic conditions a species has experienced in its habitat up to the present and determines its sensitivity to climate change; and estimates of what environmental conditions that species will experience under future climate change projections, assuming the species stays in the same place. The difference between the average current conditions and the projected future conditions reveal the predicted exposure of that species to climate change. Assuming species share similar sensitivities to change, the larger the gap between the two sets of environmental conditions, the more vulnerable the species. “And, greater exposure to changed climatic conditions usually means greater extinction risk,” said Lu.
In contrast to other commonly used vulnerability metrics that consider only the magnitude of exposure, Lu and Jetz’s novel metric also assesses the position and direction of climate change exposure – considering how close or far from the ideal conditions a given population is to begin with, and whether the projected climatic change pushed a population closer or farther away from the ideal conditions relative to the full range of its niche.
Many previous metrics, as they explain in the paper, do not consider the divergent effects different subpopulations of the same species might face. The environmental niche is determined for the whole species population, so certain subpopulations may already be experiencing conditions on the edge of the tolerable limits – for example, a subpopulation already living at the higher limit of tolerable temperatures are likely more sensitive to further change in climatic conditions, whereas a subpopulation currently experiencing the ideal conditions (the “center” of the niche) may be able to endure a greater magnitude of change before being pushed into intolerable conditions.
Lu and Jetz’s study analyzed the climate change vulnerability of 1,804 bird species in North and South America. Using a range of data resolutions from 1x1 km2 to 128x128 km2 for past and projected environmental data, they revealed that vulnerability estimates diverge significantly as grain size increases. Between the 1km2 and 128km2 analyses, more than 500 total species saw a 20% difference in their estimated vulnerability to climate change between now and the periods between 2040 to 2070. More than 140 species saw a difference in vulnerability as high as 70%.
Overwhelmingly, species judged more vulnerable to climate change at the finer grain size – 1x1 km – tended to be tropical, including Amazonian species like the white-eared jacamar (Galbalcyrhynchus leucotis) and Spix’s guan (Penelope jacquacu), while species judged less vulnerable at coarser grain size tended to be temperate, including species in the Atlantic Forest region of South America, like the chestnut-bellied euphonia (Euphonia pectoralis) and the black-throated grosbeak (Saltator fuliginosus).
These discrepancies occur because important environmental information may be lost, or biological information incorrectly assumed, when using a coarser resolution with this data. Within a heterogeneous landscape, a species’ suitable habitat may occur in small patches which are accurately detected at 1km2 but lost at coarser grains that average out the landscape conditions, leading to a higher detected vulnerability at fine grains but low vulnerability at coarse grains. The researchers describe this phenomenon as a “mismatch” between the coarse resolutions that environmental data is commonly available at versus the fine resolutions that better capture biologically relevant information. Thus, the species whose assessments are more sensitive to grain size tend to be those with very narrow ranges and specific habitat conditions and those whose habitats occur in heterogeneous landscapes.
“Our study shows that the poor spatial resolution of available data has misled recent evaluations of biodiversity vulnerability to climate change - erroneously highlighting some and missing other hotspots of change”, says Walter Jetz, who guided and supervised the work. “In contrast to climatic and environmental data - increasingly available globally at fine resolution, this limitation to biodiversity evidence persists for most parts and species groups and is hindering effective management and decision-making. Our work offers to correct for this imbalance and highlights the importance of local data collection.”
The study was made possible by the data infrastructure of Map of Life and funding from the EO Wilson Biodiversity Foundation in furtherance of the Half-Earth Project.
