Putting disagreement into context
Highlighting these three, unequivocal principles is important because ecologists and conservation biologists have long discussed how best to manage native habitat to sustain biodiversity (5, 6) . Earlier discussions revolved around the SLOSS question (should conservation prioritize ‘a single large or several small’ habitat patches?). Through time, SLOSS matured into a conversation around the effects of habitat fragmentation relative to effects of habitat amount. And more recently, the discussion among scientists has focused on whether habitat fragmentation may even have a positive effect on species. The problem is, while disagreement is healthy in an academic setting, it confuses management and policy making.
Still, the extensive body of literature addressing these topics has not been sufficient to reach consensus on them. Some scientists have concluded that landscapes containing many small patches of native habitat can sustain rare and/or habitat specialist species (e.g., due to beta diversity patterns or spatio-temporal dynamics across patches)(5, 9) , whereas others have suggested that the detrimental effects of reduced patch size inevitably depauperate biodiversity even at high cumulative habitat areas (e.g., due to extinction debt that might be paid in landscapes containing many small patches) (6, 8) . Underlying different perspectives are several factors determining our understanding of patterns in species occurrence and biodiversity. Some of these factors are contextual to different ecosystems, including biogeographical differences or intraspecific variation, and some theoretical, including issues of spatial scaling or the balance between deterministic and stochastic forces across landscapes.
Despite the different schools of thought in this field, there is no debate about the need to conserve habitat: the effects for biodiversity of increasing native habitat area are overwhelmingly positive. Furthermore, embracing the three principles we outline (Fig. 1) helps to put disagreements in the conservation literature into perspective. For instance, it is true that large areas of nature are important and must be protected (2) , as much as it is true that ensuring the conservation or restoration of multiple small habitat patches is fundamental for global biodiversity conservation, particularly in regions with high land use (9) . These are neither incompatible nor competing strategies; instead, they are complementary approaches to protect biodiversity across all regions. Healthy disagreement can be wrongly translated into a false dichotomy between the protection of large or small patches, a mistake that must be avoided at all costs for the sake of biodiversity conservation because both are important.
Habitat existing as small patches is often deemed less valuable than large swaths of habitat in less modified regions (8) , which is inadvertently leading to widespread cumulative losses of habitat from millions of small patches across the globe. For instance, a recent global analysis found that smaller (< 1000 ha) forest patches are more likely to suffer a given amount of habitat loss than larger (> 10,000 ha) patches (4) . Policies that protect only patches larger than a minimum size hinder biodiversity conservation as they fail to protect biodiversity in highly-modified regions that clearly need protection. In Southern Ontario, Canada, where only about 15% of native habitat remains, about 95% of recent wetland habitat loss has occurred below the 2-ha minimum area criterion used to identify “Provincially Significant Wetlands” (10) . Failing to maintain small habitat patches also reduces landscape connectivity among larger patches, due to the loss of “stepping stones” (Fig. 1.3), where small patches distributed through a landscape can facilitate movement between larger patches.
At the same time, very large tracts of native habitat are now limited to a small number of regions (2) , and their continued conversion to human land uses is placing biodiversity at risk. For instance, continued deforestation in the Amazon could trigger an ecosystem state-shift, because the persistence of this biome depends on feedbacks between vegetation and climate (11) . Losing even a fifth of the Amazonian forest could shift the remaining area of forest into savanna, a death-knell for the forest-dependent species of the Amazon (11).Similarly, while the few remaining extensive grasslands worldwide sequester large amounts of carbon and host unique species, they remain poorly protected and continue to shrink (12). Beyond biodiversity per se, loss of these extensive, natural habitats would bring significant losses to the economic, cultural, and ecological identity of large regions (11, 12).
Ultimately, initiatives that set targets for habitat protection in the coming decades require protection of both small and large areas of native habitat. In ecoregions where vast areas of native habitat still exist, ongoing habitat transformation needs to stop to both protect biodiversity and reduce the pace of climate change. In ecoregions with high land use, urgent protection and restoration of large numbers of small patches is needed to sustain not only biodiversity but also crucial ecosystem services including soil retention, water security, pest control, pollination, and human well-being and health.