Predicting extinctions with species distribution models

Incorporating Genetic Diversity to Optimize the Plant Conservation Network in the Third Pole

Climate change poses a significant threat to the survival of many species. Although protected areas can slow down biodiversity loss, they often lack systematic planning and do not integrate genetic diversity. Genetic diversity is a key prerequisite for species survival and the ability to tolerate new conditions. Using population genetic and distribution data from 96 plant species in the Third Pole (encompassing the Tibetan Plateau and adjacent mountains), we mapped patterns of genetic diversity, projected climate-driven range dynamics and future genetic erosion, and designed an optimal conservation framework for the region. We identified several patches of high haplotype diversity (HD), with a relatively high number of haplotypes in southeastern Third Pole. Regression models revealed that climate and topography have interacted to shape patterns of genetic diversity, with latitude and precipitation being the best predictors for HD of cpDNA and nrDNA, respectively. Ecological niche modeling predicted an approximate 43 km northwestward and 86 m upward shift in suitable habitats under future climate scenarios, likely leading to a significant loss of up to 13.19% and 15.49% of cpDNA and nrDNA genetic diversity, respectively. Alarmingly, 71.20% of the newly identified conservation priority areas fall outside of the existing protected areas and planned National Park Clusters. Therefore, we recommend expanding the network by 2.02 × 105 km2 (5.91%) in the Third Pole, increasing the total conserved area to 1.36 × 106 km2 (39.93%) to effectively preserve the evolutionary potential of plants. This study represents an innovative attempt to incorporate genetic diversity into biodiversity conservation efforts.

The ecology of plant extinctions

Extinctions occur when enough individual plants die without replacement to extirpate a population, and all populations are extirpated. While the ultimate drivers of plant extinctions are known, the proximate mechanisms at individual and population level are not. The fossil record supports climate change as the major driver until recently, with land-use change dominating in recent millennia. Climate change may regain its leading role later this century. Documented recent extinctions have been few and concentrated among narrow-range species, but population extirpations are frequent. Predictions for future extinctions often use flawed methods, but more than half of all plants could be threatened by the end of this century. We need targeted interventions tailored to the needs of each threatened species.

Marine species and assemblage change foreshadowed by their thermal bias over Early Jurassic warming

A mismatch of species' thermal preferences to their environment may indicate how they will respond to future climate change. Averaging this mismatch across species may forewarn that some assemblages will undergo greater reorganization, extirpation, and possibly extinction, than others. Here, we examine how regional warming determines species occupancy and assemblage composition of marine bivalves, brachiopods, and gastropods over one-million-year time steps during the Early Jurassic. Thermal bias, the difference between modelled regional temperatures and species' long-term thermal optima, predicts a gradient of species occupancy response to warming. Species that become extirpated or extinct tend to have cooler temperature preferences than immigrating species, while regionally persisting species fell midway. Larger regional changes in summer seawater temperatures (up to +10 °C) strengthen the relationship between species thermal bias and the response gradient, which is also stronger for brachiopods than for bivalves, while the relationship collapses during severe seawater deoxygenation. At +3 °C regional seawater warming, around 5 % of pre-existing benthic species in a regional assemblage are extirpated, and immigrating species comprise around one-fourth of the new assemblage. Our results validate thermal bias as an indicator of immigration, persistence, extirpation, and extinction of marine benthic species and assemblages under modern-like magnitudes of climate change.

With super SDMs (machine learning, open access big data, and the cloud) towards more holistic global squirrel hotspots and coldspots

Species-habitat associations are correlative, can be quantified, and used for powerful inference. Nowadays, Species Distribution Models (SDMs) play a big role, e.g. using Machine Learning and AI algorithms, but their best-available technical opportunities remain still not used for their potential e.g. in the policy sector. Here we present Super SDMs that invoke ML, OA Big Data, and the Cloud with a workflow for the best-possible inference for the 300 + global squirrel species. Such global Big Data models are especially important for the many marginalized squirrel species and the high number of endangered and data-deficient species in the world, specifically in tropical regions. While our work shows common issues with SDMs and the maxent algorithm ('Shallow Learning'), here we present a multi-species Big Data SDM template for subsequent ensemble models and generic progress to tackle global species hotspot and coldspot assessments for a more inclusive and holistic inference.

How many species will Earth lose to climate change?

Climate change may be an important threat to global biodiversity, potentially leading to the extinction of numerous species. But how many? There have been various attempts to answer this question, sometimes yielding strikingly different estimates. Here, we review these estimates, assess their disagreements and methodology, and explore how we might reach better estimates. Large-scale studies have estimated the extinction of ~1% of sampled species up to ~70%, even when using the same approach (species distribution models; SDMs). Nevertheless, worst-case estimates often converge near 20%-30% species loss, and many differences shrink when using similar assumptions. We perform a new review of recent SDM studies, which show ~17% loss of species to climate change under worst-case scenarios. However, this review shows that many SDM studies are biased by excluding the most vulnerable species (those known from few localities), which may lead to underestimating global species loss. Conversely, our analyses of recent climate change responses show that a fundamental assumption of SDM studies, that species' climatic niches do not change over time, may be frequently violated. For example, we find mean rates of positive thermal niche change across species of ~0.02°C/year. Yet, these rates may still be slower than projected climate change by ~3-4 fold. Finally, we explore how global extinction levels can be estimated by combining group-specific estimates of species loss with recent group-specific projections of global species richness (including cryptic insect species). These preliminary estimates tentatively forecast climate-related extinction of 14%-32% of macroscopic species in the next ~50 years, potentially including 3-6 million (or more) animal and plant species, even under intermediate climate change scenarios.