Species distribution models (SDMs) have become an essential tool in ecology, biogeography, evolution and, more recently, in conservation biology. How to generalize species distributions in large undersampled areas, especially with few samples, is a fundamental issue of SDMs. In order to explore this issue, we used the best available presence records for the Hooded Crane (Grus monacha, n = 33), White-naped Crane (Grus vipio, n = 40), and Black-necked Crane (Grus nigricollis, n = 75) in China as three case studies, employing four powerful and commonly used machine learning algorithms to map the breeding distributions of the three species: TreeNet (Stochastic Gradient Boosting, Boosted Regression Tree Model), Random Forest, CART (Classification and Regression Tree) and Maxent (Maximum Entropy Models). In addition, we developed an ensemble forecast by averaging predicted probability of the above four models results. Commonly used model performance metrics (Area under ROC (AUC) and true skill statistic (TSS)) were employed to evaluate model accuracy. The latest satellite tracking data and compiled literature data were used as two independent testing datasets to confront model predictions. We found Random Forest demonstrated the best performance for the most assessment method, provided a better model fit to the testing data, and achieved better species range maps for each crane species in undersampled areas. Random Forest has been generally available for more than 20 years and has been known to perform extremely well in ecological predictions. However, while increasingly on the rise, its potential is still widely underused in conservation, (spatial) ecological applications and for inference. Our results show that it informs ecological and biogeographical theories as well as being suitable for conservation applications, specifically when the study area is undersampled. This method helps to save model-selection time and effort, and allows robust and rapid assessments and decisions for efficient conservation.

Tropical rainforests play a central role in the Earth system by regulating climate, maintaining biodiversity, and sequestering carbon. They are under threat by direct anthropogenic impacts like deforestation and the indirect anthropogenic impacts of climate change. A synthesis of the factors that determine the net ecosystem exchange of carbon dioxide (NEE) at the site scale across different forests in the tropical rainforest biome has not been undertaken to date. Here, we study NEE and its components, gross ecosystem productivity (GEP) and ecosystem respiration (RE), across thirteen natural and managed forests within the tropical rainforest biome with 63 total site-years of eddy covariance data. Our results reveal that the five ecosystems with the largest annual gross carbon uptake by photosynthesis (i.e. GEP > 3000 g C m -2 y -1 ) have the lowest net carbon uptakeor even carbon lossesversus other study ecosystems because RE is of a similar magnitude. Sites that provided subcanopy CO2 storage observations had higher average magnitudes of GEP and RE and lower average magnitudes of NEE, highlighting the importance of measurement methodology for understanding carbon dynamics in ecosystems with characteristically tall and dense vegetation. A path analysis revealed that vapor pressure deficit (VPD) played a greater role than soil moisture or air temperature in constraining GEP under light saturated conditions across most study sites, but to differing degrees from -0.31 to -0.87 μmol CO2 m -2 s -1 hPa -1 . Climate projections from 13 general circulation models (CMIP5) under the representative concentration pathway that generates 8.5 W m -2 of radiative forcing suggest that many current tropical rainforest sites on the lower end of the current temperature range are likely to reach a climate space similar to present-day warmer sites by the year 2050, warmer sites will reach a climate not currently experienced, and all forests are likely to experience higher VPD. Results demonstrate the need to quantify if and how mature tropical trees acclimate to heat and water stress, and to further develop flux-partitioning and gap-filling algorithms for defensible estimates of carbon exchange in tropical rainforests.

The purging of deleterious alleles has been hypothesized to mitigate inbreeding depression, but its effectiveness in endangered species remains debatable. To understand how deleterious alleles are purged during population contractions, we analyzed genomes of the endangered Chinese crocodile lizard (Shinisaurus crocodilurus), which is the only surviving species of its family and currently isolated into small populations. Population genomic analyses revealed four genetically distinct conservation units and sharp declines in both effective population size and genetic diversity. By comparing the relative genetic load across populations and conducting genomic simulations, we discovered that seriously deleterious alleles were effectively purged during population contractions in this relict species, although inbreeding generally enhanced the genetic burden. However, despite with the initial purging, our simulations also predicted that seriously deleterious alleles will gradually accumulate under prolonged bottlenecking. Therefore, we emphasize the importance of maintaining a minimum population capacity and increasing the functional genetic diversity in conservation efforts to preserve populations of the crocodile lizard and other endangered species.