Ranking threats to biodiversity and why it doesn’t matter

Conservation scientists have proposed several rankings of the relative importance of global threats to biodiversity. Here, we argue that relative ranking of biodiversity threats depends on local context and metrics used, and so has little application for conservation.

latitudes and altitudes because species are constrained by the physical environment (geographic barriers and mountain tops) to follow their optimal isotherms.
The relative importance of threats also depends on the taxon considered. At the global scale, vertebrates are primarily threatened by habitat loss, overexploitation, and then biological invasions. But even within the vertebrates rankings differbirds and mammals are mainly affected by overexploitation, while amphibians have a higher probability of succumbing to habitat loss 6 . Because of species-specific traits and adaptations, some species are likely to respond differently to global threats even within a clade. Large-bodied vertebrates are more likely to be threatened by overexploitation, whereas small-bodied vertebrates are more prone to habitat loss or pollution (Fig. 1). Threat ranking also depends on the habitat under consideration. Marine mammals are more threatened by overexploitation and pollution than terrestrial mammals for which habitat loss is the primary threat (Fig. 1). On islands, habitat loss is secondary to the pressures of biological invasions in freshwater systems, but the former is more important for terrestrial vertebrates and plants 3 . Another source of uncertainty is that most studies examining threats are based on wellstudied taxa such as terrestrial vertebrates, which only represent a small subset of the tree of life. For instance, only 0.2% of fungi, 1.7% of invertebrates, and 10% of described plants are assessed in the IUCN update of 2019 7 , potentially underestimating the intensity of some threats and biasing conservation priorities for these groups. Similarly, there is a bias of research effort towards regions with high-income countries, while research from low or middle-income countries is generally underrepresented 8 . This may give the false impression of absence of threats in some regions of the world.
Likewise, period-specific global threat ranks are subject to the vagaries of temporal dynamics (Fig. 1). However, distinguishing past, current, and future threats is essential for current or future conservation interventions. Historically, overexploitation caused most of the Pleistocene megafauna extinctions, likely exacerbated by climate change. As agricultural practices intensified, habitat loss played a major role in extinctions. As humans later colonized islands, biological invasions caused the extinction of hundreds of species worldwide 3 . In contrast, climate change is only predicted to become major in the near future 9 . In fact, the effects of recent threats might be masked by delayed species' responses, especially in under-studied regions, resulting in a large extinction debt. For instance, the severity of biological invasions often causes native species to decline rapidly to local extinction, while other threats such as habitat loss might affect species more slowly. In both cases, the eventual extinctions are ultimately if similar magnitude.
Threat rank depends on the metric The inconsistencies among different rankings also arise from the different methodologies applied and metrics used to assess biodiversity losses. The IUCN Classification Threat, one of the most used tools in ecology and conservation, is based on expert assessment and a list of quantitative criteria (i.e., population sizes and area of occupancy) to provide a species' extinction probability.    Table 1 with their associated references.
Adopted by the Convention of Biological Diversity (CBD), the WWF Living Planet Index assesses threats to biodiversity based on population time series of over 4000 vertebrate species. Recently, a new metric based on species' extinction risk status, anthropogenic pressure, and country-scale mitigation measures has provided an interesting, complementary approach of congruence among species' extinction risks and their exposure to threats 10 . However, most of the aforementioned metrics rely on species-level data that are often unavailable. Yet, because biodiversity is a multidimensional concept, including genetic, functional, and taxonomic diversities, it is important for threat assessments to consider all dimensions. Attempting to be more integrative, the IPBES proposed a new ranking based on eleven different metrics, from species populations (e.g., Living Planet Index, local species richness) and traits (e.g., mean body length) to ecosystem structure (e.g., mangrove forest, percentage of live coral cover) 5 . Those multiple metrics result in different rankings according to the scale considered. For instance, habitat loss and degradation appear to be important for explaining changes at the population level (i.e., Living Planet and Red List indices), while pollution is among the most important threats of changes of ecosystem structure (e.g., % live coral cover) (Fig. 1). Some of these metrics are less appropriate to emphasize certain threats. For example, species traits like mean body length of fish is mostly driven by overexploitation 11 , while biological invasion had little effect on this trait, or on the percentage of live coral cover. Moreover, each threat is in fact multifaceted. For instance, climate change includes sea-level rise, altered precipitation, increase of extreme events, and some of which may even have opposite effects on certain species. For example, some plants may benefit from CO2 increases while suffering from changed in precipitation regime. As a consequence, the definition of threat itself may lead to various rankings.

Conservation implications of ranking threats
The tendency to rank biodiversity threats is not entirely due to the tendency of scientists to ordinate factors or to their relevance to justify studies of the greatest threats. It is also driven by practical considerations. Rankings are in fact commonly applied to establish priorities for conservation interventions, especially by policy makers 12 . For instance, comparative studies have successfully prioritized the conservation of some populations, species, or sites at local scales 12 . Global analyses and rankings can also guide governments and other actors, and offer leverage to implement measures with limited funding. By ranking individual threats independently, scientists implicitly prioritize the top threats and classify others as lower concern. However, ranking global threats by considering each threat independently is likely to underestimate the full consequences to biodiversity worldwide. A recent study showed that the nature of the threat is almost irrelevant when species are exposed to multiple threats simultaneously, as multiple threats exacerbate the loss of resilience in vertebrate populations 13 . Threatened insular species are currently exposed to an average of 2.6 different threats 3 , and according to the IUCN 80% of species are exposed to more than one threat threatened species 4 . In this context, we expect cumulative effects among global change drivers, which may result in unexpected interactions such as synergies or antagonistic interactions. This further weakening the usefulness of independent ranking.
Perspectives and potential opportunities By over-synthesizing the relative prevalence of threats across juxtaposed contexts, taxa, times, and metrics, these macroecological analyses confuse messages to the public, the media, and policy makers in terms of where and when biodiversity losses occur, and which factors are responsible 14 . Instead, global assessments in conservation should offer synthetic views on how threats operate in combination to threaten biodiversity. New developments in databases and tools could provide advances in conservation evaluation 15 . For instance, network analyses 3 or cumulative-impact approaches 16 can consider multiple threats simultaneously. But exposure to threats alone provides only partial information on conservation needs. New approaches assessing species' vulnerability to global changes by simultaneously considering exposure to the intensity of the threats, variable sensitivity, and different adaptive capacity are also a valuable change of paradigm 17 .
Because the ultimate objective is to lessen the severity of the biodiversity crisis driven by combination of major threats, we might even need to reconsider the actual usefulness of global threat ranks. Ranking the potential benefits of conservation actions would certainly have more impact that only ranking threats. Providing policy makers with a range of conservation scenarios with their predicted biodiversity outcomes, which consider the multiple sources of stress, different metrics, and variation across taxa, could provide a more realistic description of the state of biodiversity and maximize conservation outcomes. While international biodiversity-conservation instruments derived by the CBD, IPBES, IUCN, and others are exceptional tools to monitor and assess biodiversity, the disparities among rankings create a reductive and oversimplified perspective of the multiple threats underlying the current biodiversity crisis. This could also precipitate unsubstantiated priorities in conservation. Although useful to highlight areas of concern, we argue instead for considering the specificities, complexities, and interactions among threats to tackle this global crisis and for communicating more strategically with policymakers about the limits of global rankings.