The Past as a Lens for Biodiversity Conservation on a Dynamically Changing Planet: Changes in the functional diversity of modern bird species over the last million years

Despite evidence of declining biosphere integrity, we currently lack understanding of how the functional diversity associated with changes in abundance among ecological communities has varied over time and before widespread human disturbances. We combine morphological, ecological, and life-history trait data for >260 extant bird species with genomic-based estimates of changing effective population size (N_e) to quantify demographic-based shifts in avian functional diversity over the past million years and under pre-anthropogenic climate warming. We show that functional diversity was relatively stable over this period, but underwent significant changes in some key areas of trait space due to changing species abundances. Our results suggest that patterns of population decline over the Pleistocene have been concentrated in particular regions of trait space associated with extreme reproductive strategies and low dispersal ability, consistent with an overall erosion of functional diversity. Further, species most sensitive to climate warming occupied a relatively narrow region of functional space, indicating that the largest potential population increases and decreases under climate change will occur among species with relatively similar trait sets. Overall, our results identify fluctuations in functional space of extant species over evolutionary timescales and represent the demographic-based vulnerability of different regions of functional space among these taxa. The integration of paleodemographic dynamics with functional trait data enhances our ability to quantify losses of biosphere integrity before anthropogenic disturbances and attribute contemporary biodiversity loss to different drivers over time.

The safeguarding of biological diversity is one of the society’s most pressing challenges (1), but our understanding of its decline and consequences is still limited and often constrained by the spatial and temporal scales at which natural history studies are conducted. Despite such constraints, novel opportunities exist for using information from the Quaternary period (~2.6 Mya–present) to inform conservation practices and policies in the context of climate change, including providing natural baselines of biotic responses and quantifying the vulnerability of the natural world to varying rates and magnitudes of climate change (2). Exploring biodiversity responses during pre-human periods allows us to more clearly infer the impact of climate change without the additional effects of other anthropogenic pressures and thus tackle some of the challenges posed by the difficulty in accurately attributing observed changes to different drivers of biodiversity change (3–5). This “attribution conundrum” exists in part because collinearity, scale dependencies, or interactions among the different potential drivers of biodiversity change (e.g., ongoing climate change, habitat loss, over-exploitation) may obscure their individual effects on species declines across taxa.In addition to their overall effects on species declines across taxa, drivers of biodiversity change can also cause changes in the relative abundance of different species, shifting representation of key functional traits (i.e., the morphological, behavioral, or physiological features of organisms that can affect their fitness) in communities and leading to overall erosion of functional diversity (i.e., the variation among such features) over time (6–8). The loss of functional diversity associated with both species extinctions and overall declines in abundance can lead to declines in the functioning and stability of ecosystems over time, particularly during periods of broad-scale environmental variation such as climate change (9–11). An overrepresentation of species exhibiting similar traits can buffer these effects of declining abundances, allowing key ecosystem functions to be maintained during periods of biodiversity loss. Indeed, over the past century, changes in species abundances are more prevalent than global extinctions (12–15), emphasizing the need to incorporate estimates of species abundances in analyses of global functional diversity change (16). By identifying changes in representation across avian functional space (i.e., multi-dimensional space where species are grouped based on their functional traits) during the Pleistocene epoch (~2.6 Mya – 11.7 thousand years ago [kya]), we aim to provide baseline estimates of functional diversity within avian communities before the widespread effects of humans on global wildlife. Doing so will identify the regions of functional space that are more resilient or more sensitive to climate warming, thus helping provide more plausible scenarios for future changes in avian functional diversity.Current understanding of global and regional dynamics in functional diversity is limited due to missing information about the past abundances of species (2, 17). This missing information hampers our ability to estimate baselines before human impacts and to quantify the magnitude and rates of ongoing losses of functional diversity and associated ecosystem functioning during periods of rapid environmental change. The modern genomic revolution offers a unique opportunity to investigate how demographic changes may have affected functional diversity over the Earth’s history. In particular, species-specific estimates of changing effective population size (N_e) over thousands-to-millions of years, inferred from whole-genome sequence data (18), can provide key estimates of fluctuations in abundance over evolutionary time and across biogeographic realms. Coupling such information with morphological, life-history, and ecological trait data for a broad selection of species permits reconstruction of the representation (in terms of abundance) of different functional traits over different periods of the distant past, revealing trait sets associated with functional diversity across a dynamically changing global environment. For instance, the last million years on Earth have been punctuated by periods of intense climate warming and cooling (19, 20). A recent analysis of genomic data in birds revealed that while overall bird abundance has steadily declined over the past million years, certain sets of morphological/life-history traits were associated with either species-specific population increases or decreases during discrete periods of climate warming and cooling (21). These observations suggest that the functional diversity of avian communities also fluctuates over time and during periods of climate variation. Identifying regions of functional space that are less occupied in more recent time periods due to population declines of certain species can provide baseline information on the diversity and composition of long-term ecological strategies leading up to the modern Holocene epoch (~11 kya–present). Such information will also illuminate regions of the global functional trait space which may be less resilient to future environmental challenges, beyond those documented under contemporary studies of population responses to current climate change.Here, we combine morphological, ecological, and life-history trait data for >260 extant bird species with estimates of changing N_e to quantify shifts in avian functional diversity over the past million years and explore patterns of change over time and space. We reconstruct the abundance-weighted functional trait space of these bird species in three time periods to quantify how functional diversity has changed 1) over time, 2) across the Earth’s major zoogeographic realms, and 3) how abundances have shifted across the avian functional space to potentially influence these patterns. We further quantify changes in functional diversity during the most recent period of abrupt climate warming over the past million years to identify the areas of the functional space most sensitive to broad-scale climate warming.