Eight freshwater asexual strains collected from China and Thailand were subjected to morpho-molecular analyses. BLAST search results showed that the new strains belonged to Pleurotheciales, Sordariomycetes. Phylogenetic analyses based on a combined LSU, SSU, ITS and RPB2 sequence dataset are used to clarify the placement of the new strains within Pleurotheciales. Pleurothecium guttulatum and Pleurotheciella sympodia are introduced as two new species from freshwater habitats. New geographical record of Pleurotheciella nilotica from Asia (Thailand and China) is reported. New habitat and geographical records of Phaeoisaria annesophieae (from freshwater habitats in Asia) are reported. Descriptions and illustrations of the new species and records, as well as known species Rhexoacrodictys erecta and Phaeoisaria annesophieae, are provided. A key to Pleurotheciella species is provided.

Adversarial attacks deceive deep neural network models by adding imperceptibly small but well-designed attack data to the model input. Those attacks cause serious problems. Various defense methods have been provided to defend against those attacks by: (1) providing adversarial training according to specific attacks; (2) denoising the input data; (3) preprocessing the input data; and (4) adding noise to various layers of models. Here we provide a simple but effective Noise-Fusion Method (NFM) to defend adversarial attacks against DNN image classification models. Without knowing any details about attacks or models, NFM not only adds noise to the model input at run time, but also to the training data at training time. Two l_∞-attacks, the Fast Gradient Signed Method (FGSM) and the Projected Gradient Descent (PGD), and one l₁-attack, the Sparse L1 Descent (SLD), are applied to evaluate defense effects of the NFM on various deep neural network models which used MNIST and CIFAR-10 datasets. Various amplitude noises with different statistical distribution are applied to show the defense effects of the NFM in different noise. The NFM also compares with an adversarial training method on MNIST and CIFAR-10 datasets. Results show that adding noise to the input images and the training images not only defends against all three adversarial attacks but also improves robustness of corresponding models. The results indicate possibly generalized defense effects of the NFM which can extend to other adversarial attacks. It also shows potential application of the NFM to models not only with image input but also with voice or audio input.

Understanding the spatial distribution and habitat preference for rare and endangered species are essential for effective conservation practice. We examined the spatial distribution and habitat preference of four Diploderma species (Diploderma drukdaypo, D. laeviventre, D. batangense, and D. vela), which are endemic to the Qinghai-Tibet Plateau and are currently under state protection. We used the ensembles of small models (ESM) approach and predicted potential distribution ranges of the species in current and two future climate scenarios (SSP126 and SSP585). The degree of overlap between the predicted distribution ranges and existing natural reserves was further analyzed. Habitat preference was examined using a paired quadrat method. Our results predicted that D. drukdaypo has a current distribution range of 600 km2, which would decrease to 50 km2 and 55 km2 under the SSP126 and SSP585 respectively. For D. laeviventre, the current distribution range is 817 km2, with minimum changes in the two future climate scenarios (774 km2 and 902 km2). For D. batangense, the current distribution range is 875 km2, which would expand to 1522 km2 and 3340 km2 in the two future climate scenarios. Similarly for D. vela, the current distribution range is 1369 km2, which would change to 1825 km2 and 2043 km2 respectively under the two future climate scenarios. The effect of protection of current nature reserves are likely low for those species; we found no overlap (D. drukdaypo, D. laeviventre) or little overlap (D. batangense 2.04%-3.56%, D. vela 15.52%-16.87%) between the currently designated protection area and distribution range under current and future climate scenarios. For habitat preference, stones appear to be the critical habitat element for those species although different species had different stone requirement. Taken together, we provided critical information on potential distribution ranges and habitat preference for four endangered Diploderma species, and confirmed the inadequacy of current nature reserves. The establishment of new or expansion of existing nature reserves is urgent for the conservation of those species.