Land-Cover Classification Using MaxEnt: Can We Trust in Model Quality Metrics for Estimating Classification Accuracy?

Multitemporal Land Use and Cover Analysis Coupled with Climatic Change Scenarios to Protect the Endangered Taxon Asphodelus bento-rainhae subsp. bento-rainhae

Climate change and land use and land cover (LULC) change are impacting the species' geographic distribution, causing range shifts and reducing suitable habitats. Asphodelus bento-rainhae subsp. bento-rainhae (AbR) is an endangered endemic plant restricted to Serra da Gardunha (Portugal), and knowledge of those changes will help to design conservation measures. MaxEnt was used to model AbR's current distribution and project it into the future, 2050, using the Shared Socioeconomic Pathway SSP3-7. The Portuguese LULC maps from 1951-1980, 1995, 2007, and 2018 were used to assess and quantify LULC changes over time. The results showed that the AbR current predicted distribution matches its actual known distribution, which will not be affected by future predicted climate change. The significant LULC changes were observed during the study periods 1951-1980 to 2018, particularly between 1951-1980 and 1995. Scrubland and Agriculture decreased by 5% and 2.5%, respectively, and Forests increased by 4% in the study area. In the occurrence area, Agriculture increased, and Forests decreased between 1980 and 2018, due to Orchard expansion (34%) and declines in Chestnut (16.9%) and Pine (11%) areas, respectively. The use of species distribution models and the LULC change analysis contributed to understanding current and future species distribution. The LULC changes will have a significant impact on future species distribution. To prevent the extinction of this endemic species in the future, it is crucial to implement conservation measures, namely species monitoring, replantation, and germplasm conservation, in addition to guidelines for habitat conservation.

Effects of climate change on the geographical distribution and potential distribution areas of 35 Millettia Species in China

Climate change has an extremely important impact on the geographic distribution of plants. The genus Millettia is an important plant resource in China and is widely used in medicine and ornamental industries. Due to the continuous changes of climate and the development and utilization of plant resources of the genus Millettia, it is of great significance to systematically investigate the geographic distribution of plants of the Millettia and their potential distribution under climate change. DIVA-GIS software was used to analyze 3492 plant specimens of 35 species of genus Millettia in the herbarium, and the ecological geographic distribution and richness of Millettia were analyzed, and the MaxEnt model was used to analyze the current and potential distribution in the future. The results show that the genus Millettia is distributed in 30 provinces in China, among which Yunnan and Guangdong provinces are the most distributed. Our model determines that precipitation in the driest month and annual temperature range are the most important bioclimatic variables. Future climate changes will increase the suitable habitat area of M. congestiflora by 16.75%, but other cliff beans Suitable habitats for vines will decrease significantly: M. cinereal by 47.66%, M. oosperma by 39.16%, M. pulchra by 36.04%, M. oraria by - 29.32%, M. nitida by 22.88%, M. dielsiana by 22.72%, M. sericosema by 19.53%, M. championii by 7.77%, M. pachycarpa by 7.72%, M. speciose by 2.05%, M. reticulata by 1.32%. Therefore, targeted measures should be taken to protect and develop these precious plant resources.

Geographical distribution and predict potential distribution of Cerasus serrulata

Climate change is closely related to the distribution of plant resources. Cerasus serrulata is an important plant resource in China. The study on the influence of environmental factors on the distribution of suitable areas of C. serrulata is conducive to the protection and development of C. serrulata. In this paper, the distribution information of 238 Chinese Cerasus serrulata plants was processed by DIVA-GIS. The MaxEnt model was used to simulate the current and future distribution, and the ecological distribution and richness of Cerasus were analyzed. The results showed that the Cerasus serrulata was widely distributed in Hebei, Heilongjiang, Shandong, Jiangsu, Zhejiang, Anhui, Jiangxi, Hunan, and Guizhou provinces, mainly in the low and middle elevation areas of 10 to 1200 m. Based on this model, the precipitation of the warmest quarter, the precipitation of the driest month, and the mean temperature of the coldest were the most significant bioclimatic variables affecting the distribution of C. serrulata. In the future, climate change may lead to a slight increase of 2.31% in the area of suitable habitat for Cerasus serrulata, while the optimal habitat will decrease from 20.81 to 14.55%. Therefore, conservation measures should be taken to protect these precious resources.

An Improvised Machine Learning Model Based on Mutual Information Feature Selection Approach for Microbes Classification

The accurate classification of microbes is critical in today's context for monitoring the ecological balance of a habitat. Hence, in this research work, a novel method to automate the process of identifying microorganisms has been implemented. To extract the bodies of microorganisms accurately, a generalized segmentation mechanism which consists of a combination of convolution filter (Kirsch) and a variance-based pixel clustering algorithm (Otsu) is proposed. With exhaustive corroboration, a set of twenty-five features were identified to map the characteristics and morphology for all kinds of microbes. Multiple techniques for feature selection were tested and it was found that mutual information (MI)-based models gave the best performance. Exhaustive hyperparameter tuning of multilayer layer perceptron (MLP), k-nearest neighbors (KNN), quadratic discriminant analysis (QDA), logistic regression (LR), and support vector machine (SVM) was done. It was found that SVM radial required further improvisation to attain a maximum possible level of accuracy. Comparative analysis between SVM and improvised SVM (ISVM) through a 10-fold cross validation method ultimately showed that ISVM resulted in a 2% higher performance in terms of accuracy (98.2%), precision (98.2%), recall (98.1%), and F1 score (98.1%).