Back

Areas of high risk for mammalian biodiversity and Nature's Contributions to People under global warming

Cimatti, M.; Sacchi, A.; Di Marco, M.

2024-08-30 ecology
10.1101/2024.08.30.610332 bioRxiv
Show abstract

Climate change has reached unprecedented levels, causing frequent extreme events like droughts and fires. Combined with land-use change, this crisis has impacted biodiversity, increasing species extinction rates, and Natures Contributions to People (NCP), degrading ecosystem functions. We developed a comparative extinction risk model for mammals sensitive to fire, drought, and extreme temperatures, utilizing a Random Forest algorithm to predict future extinction probabilities under different climatic scenarios. We then identified high-risk areas for both mammals and NCP under global warming, aiming to find synergies between biodiversity conservation and NCP preservation. Our results show that 288 out of 454 species (63%) face an increased extinction risk (mean increase 0.28), while 166 species (37%) show a predicted decrease (mean decrease 0.20) under the extreme "Fossil-fueled development" scenario. The highest risk increase was observed in Malaysia, Western Indonesia, Madagascar, Eastern Australia, and South Africa, under both pessimistic and optimistic ("Sustainability") scenarios. These regions also represent high-risk areas for several NCP: freshwater regulation, air quality, mitigation of extreme events. Preserving these high-risk regions is crucial for reducing habitat loss and human-induced extinctions. Safeguarding these ecosystems, which provide vital contributions like carbon storage, clean water, and extreme fire mitigation, should be a high priority. These regions warrant targeted policy and management interventions, including sustainable land-use practices and climate adaptation actions, to benefit both biodiversity and human well-being.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.