Search

The emergence of metacommunity theory has significantly contributed to our understanding of the drivers of community assembly and biome variation. The Network Location Hypothesis (NPH) posits that tributary communities situated at the source regions of a river are disproportionately susceptible to environmental filtering due to their remote location and consequent reduced connectivity to downstream reaches of the river system. However, downstream communities located in central parts of the river network exhibit increased connectivity to other communities, thereby making them more susceptible to spatial effects. Nonetheless, empirical studies testing this theory have been relatively scarce to date. Additionally, it is widely acknowledged that integrating multiple dimensions of beta diversity can enhance our understanding of the mechanisms driving community assembly. Based on the above, we collected macroinvertebrate samples from a boreal river in China to verify these views. Specifically, we examined the significance of network location on metacommunity assembly (NPH hypothesis) by utilizing a distance-decay relationship and simultaneously assessing multiple dimensions of ecological drivers of beta diversity. Our results revealed that the predictions of the NPH hypothesis were not supported in the study area, with the impact of environmental filtering on community assembly being prevalent regardless of network location. Taxonomic beta diversity consists almost entirely of turnover, with turnover contributing more to functional beta diversity than nestedness, while phylogenetic beta diversity consists of a combination of turnover and nestedness. We observed that a uniform species composition across sites led to higher taxonomic beta diversity in the study area. However, functional redundancy and the presence of closely related species across sites resulted in lower functional and phylogenetic beta diversity compared to taxonomic beta diversity. Although we found some correlation between phylogenetic and functional beta diversity, their mechanisms of variation were not consistent, with phylogenetic beta diversity showing greater variability than functional beta diversity. This suggests that differences in functional traits may be primarily driven by more distantly related species. Therefore, our findings do not fully support the use of phylogenetic distance as a surrogate for functional distance. The present study emphasizes the significance of incorporating multiple dimensions of beta diversity in metacommunity research, as they offer unique insights into beta diversity. Specifically, we found that environmental factors play a crucial role in shaping macroinvertebrate community composition and functional traits, which is associated with the higher environmental heterogeneity within the study area. In contrast, spatial processes, such as dispersal limitations, lead to variations in the evolutionary history of organisms across different locations, which is associated with the larger geographical extent of the study area.

Training adversarially robust discriminative (i.e., softmax) classifier has been the dominant approach to robust classification. Building on recent work on adversarial training (AT)-based generative models, we investigate using AT to learn unnormalized class-conditional density models and then performing generative robust classification. Our result shows that, under the condition of similar model capacities, the generative robust classifier achieves comparable performance to a baseline softmax robust classifier when the test data is clean or when the test perturbation is of limited size, and much better performance when the test perturbation size exceeds the training perturbation size. The generative classifier is also able to generate samples or counterfactuals that more closely resemble the training data, suggesting that the generative classifier can better capture the class-conditional distributions. In contrast to standard discriminative adversarial training where advanced data augmentation techniques are only effective when combined with weight averaging, we find it straightforward to apply advanced data augmentation to achieve better robustness in our approach. Our result suggests that the generative classifier is a competitive alternative to robust classification, especially for problems with limited number of classes., Comment: Report

We present a new method for face recognition from digital images acquired under varying illumination conditions. The method is based on mathematical modeling of local gradient distributions using the Radon Cumulative Distribution Transform (R-CDT). We demonstrate that lighting variations cause certain types of deformations of local image gradient distributions which, when expressed in R-CDT domain, can be modeled as a subspace. Face recognition is then performed using a nearest subspace in R-CDT domain of local gradient distributions. Experiment results demonstrate the proposed method outperforms other alternatives in several face recognition tasks with challenging illumination conditions. Python code implementing the proposed method is available, which is integrated as a part of the software package PyTransKit., Comment: 14 pages, 9 figures

Exposure of prey to predation risk usually leads to increased metabolism and decreased nutrient deposition as physiological stress responses (PSRs) of prey to their predators. Many studies using freshwater animals as models failed to find evidences as predicted from PSRs. One possible reason is that energetic costs during development of inducible defenses may lower the quantity of resources, which can be allocated to production of costly PSRs. Results obtained from the classic Brachionus-Asplanchna model indicated that inducible defenses did not alleviate PSRs in B. calyciflorus, which were manifested as increased respiration and enhanced body molar ratios of carbon (C) to nitrogen (N) and C to phosphorus (P). Moreover, activities of antioxidant systems also increased in defended B. calyciflorus. There were no alterations in somatic growth and swimming activity in B. calyciflorus when they were exposed to Asplanchna kairomones; however, we observed predator-induced –reduction of food intake in Brachionus. Our results suggest that enhanced metabolisms of defended prey in support of fight-or-flight responses are likely to be the physiological adaptation of prey to diminish hydrodynamic costs accompanied with morphological alterations (e.g., decreased swimming speed) and maintain their fitness under predation risk.

In freshwater ecosystems, hatching strategy of diapausing eggs (DEs) under predation risk has important ecological implication for zooplankters. Although kairomones released by predators can induce phenotypic responses of prey, hatching patterns of DEs in response to kairomones have received contradictory conclusions in zooplankters. Maternal environment may also affect hatching strategy of DEs during predator���prey interactions. We used classical Brachionus calyciflorus���Asplanchna models to determine the timing and proportion of diapausing-egg hatching in association with parental and embryonic exposure to kairomones. Results obtained from two Brachionus clones supported the hypothesis that DEs could detect Asplanchna kairomones and adjust hatching patterns. DEs showed early and synchronous hatching patterns in the environment with kairomones. Data also supported the prediction that DEs could gain information about predators from maternal environments and adjusted their hatching pattern in response to the presence of kairomones. Compared with DEs from Brachionus mothers not exposed to kairomones, DEs produced by mothers experienced kairomones attained a higher hatching rate when both of them hatched in the environment either with kairomones or not. Our results suggest that DEs of B. calyciflorus possess dormant plasticity to defend against predation from Asplanchna, which may be regulated by maternal environmental effects during sexual life cycles.

Dwarfism in males is a common phenomenon in planktonic monogonont rotifers. Considering the small body size of males, some studies suggested that the development time of male rotifers is faster than that of females. However, empirical studies associated with such a speculation are scarce. In this study, we used two common rotifer species, Brachionus calyciflorus and Brachionus plicatilis (each with two geographical populations), to test the hypothesis that the development time of dwarf males is significantly shorter than that of females. Results showed that male eggs (embryos) were deposited much faster than female eggs (embryos) in both Brachionus species. However, the embryonic development time of these small male eggs (embryos) was longer than that of large female eggs (embryos). As a result, males needed significantly longer total development time than females. The total development time of males was 2–3 h longer than that of females in both Brachionus species. Male-producing and female-producing females in the two Brachionus species did not show any difference in development time. Because sexual reproduction begins at high population densities in both species, postponement of development in males will be advantageous by decreasing mating costs in male rotifers.

Biological community structures have long been used to assess the impact of human activity on running waters.Compared to macroinvertebrates and fish community,periphyton community was infrequently used as indicator of stream conditions because of the large number of species and the need for taxonomic expertise.However,periphyton is valuable indicator of stream health because they are relative simple to collect and response rapidly and predictably to changes in stream chemistry and habitat quality.In the present work,we investigated the community structure of periphyton and characteristics of environmental factors among 69 sample sites in Taizi River Basin of Liaoning Province,Northeast China.Using the periphyton assemblages,we assessed stream conditions of Taizi River Basin based on Qualitative Habitat Evaluation Index(QHEI),Diatom Bioassessment Index(DBI) and Periphyton Index of Biological Integrity(P-IBI). The results indicated that the dominant species of periphyton in Taizi River Basin were Melosira varians,Diatoma vulgare,Cymbella tumida,C.hustedtii,Navicula cryptocephala,Gomphonema olivaceum,Achnanthes minutissima,Synedra acus,S.ulna,S.ulna oxyrhynchus,Nitzschia palea,Cocconeis placentula,Scenedesmus quadricauda,Oscillatoria tenuis and Phormidium tenus.There was great difference in the periphyton community structure in Taizi River Basin.Based on principal component analysis,conductivity,suspended solid,velocity,total dissolved solid,hardness,alkalinity,ammonia nitrogen,nitric acid nitrogen and total nitrogen were the main environmental factors in Taizi River Basin.Canonical correspondence analysis showed that conductivity,total dissolved solids and total nitrogen significantly influenced the community structure of periphyton in Taizi River Basin. The metrics of DBI were species richness,diversity index,tolerance index,sensitive species and similarity of community,and the metrics of QHEI were epifaunal substrate,embeddedness,velocity,sediment deposition,channel flow status,channel alteration,frequency of riffles,bank stability,vegetative protection and riparian vegetative zone width.The merits of P-IBI were filtered with box-plots analysis and correlation analysis,and the ultimate index merits of P-IBI were total genera of green algae,relative species richness of sensitive diatoms,relative richness of diatom taxons,relative richness of blue-green algae taxons,percent of Nitzschia,percent of Oscillatoria,percent of motile diatoms,percent of diatoms with handle,Chlorophyll a content of periphyton and cell density of periphyton. The results of stream condition assessment based on DBI,P-IBI and QHEI were in consensus in Taizi River Basin,although assessment results of some tributary were not in compatible.The results of stream condition assessment showed that upstream area of Taizi River Basin were in excellent ecological integrity,while,the downstream area of Taizi River Basin were in severe impairment.The results of correlation analysis showed that there were significant correlations among the scores of DBI,P-IBI and QHEI in sample sites of Taizi River Basin,and moreover,significant correlations were also detected between scores of DBI,P-IBI,QHEI and conductivity,total dissolved solid and total nitrogen.In the paper,we also discussed the relationship between periphyton community structure and environmental factors,and moreover,we evaluated the merit and deficiency of DBI,P-IBI and QHEI in the assessment of stream conditions of Taizi River Basin.

In the present work, we evaluated the feeding selectivity of starved Daphnia magna on two freshwater green algae Chlamydomonas sajao and Chlorella pyrenoidosa. Compared to C. pyrenoidosa, food quality of C. sajao are better in food palatability (cell size and digestibility), but poor in nutritional content (total carbon content). D. magna was starved for 0 and 8 d, and then was allowed to graze on a mixture of C. sajao and C. pyrenoidosa with following proportion: 5 × 104: 35 × 104 cells ml−1, 20 × 104: 20 × 104 cells ml−1 and 35 × 104: 5 × 104 cells ml−1. The results indicated that the ingestion rate and filtration rate of starved D. magna, comparing with satiated groups, on C. pyrenoidosa increased significantly, while, inverse trends was observed in C. sajao. Base on selectivity coefficient of D. magna, we observed that when D. magna was in satiation C. sajao will be preferred, while, C. pyrenoidosa will be selected when D. magna was in starvation, and moreover, these foraging behaviors were not influenced by the relative food abundance of each green alga. Therefore, a tradeoff between food palatability (physical makeup) and food nutritional content (chemical composition) can be hypothesized in the foraging behavior of D. magna, which is modified by the starvation of feeder. High valuable food is always selected by D. magna as predicted by optimal foraging theory. However, when D. magna is in satiation food diets with adequate size and easy digestibility will be preferred, while, those foods with relatively higher lipid or total carbon content will be selected when D. magna is in starvation.