Francesco Maria Sabatini
Alma Mater Studiorum - University of Bologna, Department of Biological Geological and Environmental Sciences

Winning article: Global patterns of vascular plant alpha diversity (Nature Communications, 2022)

“Our work helps prioritize efforts to protect and restore plant biodiversity globally, which is crucial to keep us and the planet healthy”

Understanding the distribution of vegetation on Earth is critical for assessing and preserving the many essential services which plants provide, both directly and indirectly, to human and planetary well-being. But such knowledge is also a moral imperative in times of global change and biodiversity loss.

Over the past two centuries we have witnessed remarkable advances in our knowledge about the global distribution of plant biodiversity. We are now able to pinpoint regional hotspots, i.e., large areas with an exceptional diversity of plant species but threatened by human activities, and are able to answer questions related to, for instance, the country hosting the highest number of plant species (Brazil), or the number of species occurring in Madagascar (more than 11,000). But focusing only on the ‘regional scale’ neglects one fundamental fact of plant life: plants do not live alone. Rather they co-exist and self-organize within communities. Much more than a group of species or individuals, a community is a tightly knit network of interactions, some mutually beneficial, others resulting in fierce competition. If we want to understand how plant biodiversity will be shaped by the ongoing global change, we must consider these entanglements.

Communities can be more or less rich in species. And we know surprisingly little about the global distribution of biodiversity at the community level. For instance, we do not know where the most species rich communities occur, nor whether one hectare of forest in Congo contains more or less species than one hectare in New Guinea. This is a problem, because when thinking about the challenge of conservation, an old adage applies: “Think globally, act locally”. Species-rich and species-poor communities require different protection approaches and restoration strategies on the ground.

In our work, we strive to understand the spatial patterns and environmental and historical drivers of plant community diversity over continental-to-global extents. We define this approach as Global Vegetation Science. As a starting point, we generated a set of global maps of local species richness across spatial resolutions that are most meaningful for plant communities, from 10 m2 to 1 ha. Compared to classic maps of biodiversity hotspots, mapping species richness at multiple resolutions represent a paradigm shift in the understanding of the distribution of plant biodiversity, which is, ultimately, a matter of scale. The rationale is that plant communities, not regions, are the functional units that hold the key to a mechanistic understanding of our evolving planet. Different communities react differently to stressors related to key planetary boundaries, be them climate change, Nitrogen deposition or loss of genetic diversity. Understanding these differential impacts is crucial to predict how the Earth system will look like under possible future trajectories. Not on average, but from place to place.

Studying communities at a global scale is extremely challenging, however. Even in well-sampled regions, vegetation data are scattered and inconsistent, because they have been collected using sampling units of various shapes and sizes, and with varying data standards and protocols. The number of species occurring in a habitat patch increases as a power of its area. Yet the exact scaling function changes across regions, habitat types and sampling designs, which turns the comparison of these heterogeneous data across studies into a wicked problem. It is no surprise that global biodiversity studies have mostly focused on regions, rather than communities.

To overcome these barriers, we developed a global vegetation science approach rooted in the realm of big data, as well as advanced modeling techniques. We tap into sPlot, a global database of vegetation plot data, which is the result of a collective and collaborative effort of thousands of vegetation scientists worldwide. In less than ten years, sPlot has aggregated and harmonized more than 2.5 million vegetation plots from 308 research groups and 144 countries. We used sPlot for a wealth of applications, for instance to model global trait–environment relationships of plant communities, understand to what extent plant rooting depth and xylem vulnerability are related to climatic or topographical features across ecosystems and biogeographical regions, and to make inferences on the relationship between niche breadth and range size of different species.

Figure 1. Vegetation shows striking differences across and within continents. The sPlot vegetation-plot database aims at understanding the variation across plant communities at global scale and create a synthetic understanding of how communities respond to environmental drivers and global stressors.

But biodiversity big data alone is not enough if we lack effective ways to extract ecologically meaningful information. To this end, we employed a plethora of methods, but machine learning proved particularly useful when modelling the distribution of specific characteristics of plant community, such as species richness or functional diversity, as a function of their relationship with environmental drivers. But we are also increasingly using parametric methods which focus on the dissimilarity between communities, i.e., how communities vary in space or time, when considering the identities, characteristics or evolutionary histories of plant species. While computationally very intensive, these tools hold great promise for uncovering ecological patterns and processes over large spatial scales, and have already been successfully used to address a range of conservation-related questions in a spatially (and sometimes temporally) explicit fashion.

The solutions which we are pioneering have repercussions beyond the scientific realm. By providing a baseline description of global plant community distribution and a holistic understanding of the underlying environmental drivers, global vegetation science can help achieve a better quantification of the impacts of land-use and climate change, and formulate better predictions of future vegetation, which is critical in an era where ecosystems may undergo transformations with no analogues to those humanity has experienced so far. This knowledge, when applied, can guide sustainable land-use planning, inform conservation decisions, and drive restoration efforts, thus mitigating the impact of human activities on ecosystems.

In the context of planetary boundaries science, global vegetation science can also refine and bridge the gap between two key boundaries: biosphere integrity and land system change. Biosphere integrity is currently measured based on the proportion of Net Primary Productivity (NPP), a proxy measuring the flow of photosynthetic energy and materials into the biosphere, that humans use for their scopes. By proving a better understanding of the current distribution and future trajectories of plant communities, global vegetation science can improve current estimates and projections of future NPP, therefore improving Earth System Models and refining our understanding of climate-biosphere interactions. Similarly, vegetation also relates to a second key boundary, land system change. At the moment, this boundary is simply assessed as the proportion of forest cover remaining compared to the potential area of forest in pre-industrial conditions. Global vegetation science holds potential to refine this critical indicator, by contributing spatially-explicit information on the compositional or functional difference between forests before and after climate change, besides highlighting areas of maximum change and possible areas that could buffer the negative effects of climate change.

In the ever-evolving landscape of biodiversity research, global vegetation science is witnessing promising trends that can be leveraged to advance the science of planetary boundaries, for instance informing new management and conservation solutions for sustaining ecosystems amidst evolving environmental pressures. Integration across multiple sources of biodiversity data, coupled with advancements in remote sensing for near-real time monitoring of vegetation, and novel modelling methodologies is propelling global vegetation science towards an era that will transcend the idiosyncrasies of local plant ecology and the simplifications of global approaches based on plant functional types. A full integration of global vegetation science with Earth system science holds promise to provide new insights into ecosystem dynamics, global change impacts on biodiversity and climate-biosphere feedbacks and tipping points on a planetary scale.

The Quest for a Global Vegetation Science

The Recovery of European Freshwater Biodiversity Has Come to a Halt

Bringing the Planet Back to the Safe Operating Space by Controlling Nutrient Pollution to Lakes