Severe Bottleneck Research Articles

The Quaternary evolutionary history of Bristol rock cress (Arabis scabra, Brassicaceae), a Mediterranean element with an outpost in the north-western Atlantic region.

Bristol rock cress is among the few plant species in the British Isles considered to have a Mediterranean-montane element. Spatiotemporal patterns of colonization of the British Isles since the last interglacial and after the Last Glacial Maximum (LGM) from mainland Europe are underexplored and have not yet included such floristic elements. Here we shed light on the evolutionary history of a relic and outpost metapopulation of Bristol rock cress in the south-western UK. Amplified fragment length polymorphisms (AFLPs) were used to identify distinct gene pools. Plastome assembly and respective phylogenetic analysis revealed the temporal context. Herbarium material was largely used to exemplify the value of collections to obtain a representative sampling covering the entire distribution range. The AFLPs recognized two distinct gene pools, with the Iberian Peninsula as the primary centre of genetic diversity and the origin of lineages expanding before and after the LGM towards mountain areas in France and Switzerland. No present-day lineages are older than 51 ky, which is in sharp contrast to the species stem group age of nearly 2 My, indicating severe extinction and bottlenecks throughout the Pleistocene. The British Isles were colonized after the LGM and feature high genetic diversity. The short-lived perennial herb Arabis scabra, which is restricted to limestone, has expanded its distribution range after the LGM, following corridors within an open landscape, and may have reached the British Isles via the desiccated Celtic Sea at about 16 kya. This study may shed light on the origin of other rare and peculiar species co-occurring in limestone regions in the south-western British Isles.

The genetic legacy of extreme exploitation in a polar vertebrate

Understanding the effects of human exploitation on the genetic composition of wild populations is important for predicting species persistence and adaptive potential. We therefore investigated the genetic legacy of large-scale commercial harvesting by reconstructing, on a global scale, the recent demographic history of the Antarctic fur seal (Arctocephalus gazella), a species that was hunted to the brink of extinction by 18th and 19th century sealers. Molecular genetic data from over 2,000 individuals sampled from all eight major breeding locations across the species’ circumpolar geographic distribution, show that at least four relict populations around Antarctica survived commercial hunting. Coalescent simulations suggest that all of these populations experienced severe bottlenecks down to effective population sizes of around 150–200. Nevertheless, comparably high levels of neutral genetic variability were retained as these declines are unlikely to have been strong enough to deplete allelic richness by more than around 15%. These findings suggest that even dramatic short-term declines need not necessarily result in major losses of diversity, and explain the apparent contradiction between the high genetic diversity of this species and its extreme exploitation history.

A single polyploidization event at the origin of the tetraploid genome of Coffea arabica is responsible for the extremely low genetic variation in wild and cultivated germplasm

The genome of the allotetraploid species Coffea arabica L. was sequenced to assemble independently the two component subgenomes (putatively deriving from C. canephora and C. eugenioides) and to perform a genome-wide analysis of the genetic diversity in cultivated coffee germplasm and in wild populations growing in the center of origin of the species. We assembled a total length of 1.536 Gbp, 444 Mb and 527 Mb of which were assigned to the canephora and eugenioides subgenomes, respectively, and predicted 46,562 gene models, 21,254 and 22,888 of which were assigned to the canephora and to the eugeniodes subgenome, respectively. Through a genome-wide SNP genotyping of 736 C. arabica accessions, we analyzed the genetic diversity in the species and its relationship with geographic distribution and historical records. We observed a weak population structure due to low-frequency derived alleles and highly negative values of Taijma’s D, suggesting a recent and severe bottleneck, most likely resulting from a single event of polyploidization, not only for the cultivated germplasm but also for the entire species. This conclusion is strongly supported by forward simulations of mutation accumulation. However, PCA revealed a cline of genetic diversity reflecting a west-to-east geographical distribution from the center of origin in East Africa to the Arabian Peninsula. The extremely low levels of variation observed in the species, as a consequence of the polyploidization event, make the exploitation of diversity within the species for breeding purposes less interesting than in most crop species and stress the need for introgression of new variability from the diploid progenitors.

Historical Population Declines Prompted Significant Genomic Erosion​ in the Northern and Southern White Rhinoceros (​ <i>Ceratotherium simum</i>)

Large vertebrates are extremely sensitive to anthropogenic pressures, and their populations are declining fast. The white rhinoceros (​Ceratotherium simum) is a paradigmatic case: this African megaherbivore suffered a remarkable population reduction in the last 150 years due to human activities. The two white rhino subspecies, the Northern (NWR) and the Southern white rhino (SWR), however, underwent opposite fates: the NWR vanished quickly after the onset of the decline, while the SWR recovered after a severe bottleneck. Such demographic events are predicted to have an erosive effect at the genomic level, linked to increased genetic drift and inbreeding. However, there is little empirical data available with which to reconstruct the subtleties of such processes. We therefore generated a whole-genome, temporal dataset consisting of 52 re-sequenced white rhino genomes, that represents both subspecies at two time windows throughout the past ~ 170 years. Our data not only reveals previously unknown population substructure within both subspecies, but allowed us to quantify the genomic erosion they underwent, with post-bottleneck white rhinos showing significantly lower levels of heterozygosity and higherinbreeding coefficients than pre-bottleneck individuals. Moreover, the effective population size (N​e​) has suffered a decrease of two and three orders of magnitude in the NWR and SWR, respectively. Our data provides empirical support for theoretical predictions about the genomic consequences ofshrinking populations, which is relevant for understanding the process of population extinction. Furthermore, our findings have the potential to inform tailored management approaches for the conservation of the remaining white rhinos.

Recent Development of Mg Ion Solid Electrolyte.

Although the successful deployment of lithium-ion batteries (LIBs) in various fields such as consumer electronics, electric vehicles and electric grid, the efforts are still ongoing to pursue the next-generation battery systems with higher energy densities. Interest has been increasing in the batteries relying on the multivalent-ions such as Mg2+, Zn2+, and Al3+, because of the higher volumetric energy densities than those of monovalent-ion batteries including LIBs and Na-ion batteries. Among them, magnesium batteries have attracted much attention due to the promising characteristics of Mg anode: a low redox potential (−2.356 V vs. SHE), a high volumetric energy density (3,833 mAh cm−3), atmospheric stability and the earth-abundance. However, the development of Mg batteries has progressed little since the first Mg-ion rechargeable battery was reported in 2000. A severe technological bottleneck concerns the organic electrolytes, which have limited compatibility with Mg anode and form an Mg-ion insulating passivation layer on the anode surface. Consequently, beneficial to the good chemical and mechanical stability, Mg-ion solid electrolyte should be a promising alternative to the liquid electrolyte. Herein, a mini review is presented to focus on the recent development of Mg-ion solid conductor. The performances and the limitations were also discussed in the review. We hope that the mini review could provide a quick grasp of the challenges in the area and inspire researchers to develop applicable solid electrolyte candidates for Mg batteries.

Purging of highly deleterious mutations through severe bottlenecks in Alpine ibex

Human activity has caused dramatic population declines in many wild species. The resulting bottlenecks have a profound impact on the genetic makeup of a species with unknown consequences for health. A key genetic factor for species survival is the evolution of deleterious mutation load, but how bottleneck strength and mutation load interact lacks empirical evidence. We analyze 60 complete genomes of six ibex species and the domestic goat. We show that historic bottlenecks rather than the current conservation status predict levels of genome-wide variation. By analyzing the exceptionally well-characterized population bottlenecks of the once nearly extinct Alpine ibex, we find genomic evidence of concurrent purging of highly deleterious mutations but accumulation of mildly deleterious mutations. This suggests that recolonization bottlenecks induced both relaxed selection and purging, thus reshaping the landscape of deleterious mutation load. Our findings highlight that even populations of ~1000 individuals can accumulate mildly deleterious mutations. Conservation efforts should focus on preventing population declines below such levels to ensure long-term survival of species.

Historic and recent status of the Kosrae flying fox (Pteropus ualanus) (Chiroptera: Pteropidae) on Kosrae, Micronesia

Seven species of flying foxes of the genus Pteropus have been described from the Caroline Islands and Mariana Islands of Micronesia. Among the five extant species, P. ualanus of Kosrae is the least studied. We review the history of research on P. ualanus and summarize data from recent surveys. Early historical accounts suggest that P. ualanus was abundant until a severe population bottleneck, attributed to an infectious disease, allegedly reduced the population to two individuals during 1926–1927. During 1982 and 1983, we found three colonies and a minimum of 600–700 bats in the mountains. Subsequently two new colonies became established in coastal localities by 1994 and 2002 and currently represent the largest colonies. During 2013 and 2014, Hayes visited four confirmed and two possible colonies (three in the mountains, three along the coast) ranging in size from two to 1038 individuals and counted a minimum of 1765 individuals. Long-term changes occurred in the size of colonies. Because not all individuals were counted and at least one other colony likely exists, the population probably exceeds 2000 and may exceed 3000 individuals. The population may not have reached its carrying capacity and may still be increasing. Potential existential threats include infectious disease, typhoons, overhunting, and habitat loss.

Understanding Computational Costs of Cellular-Level Brain Tissue Simulations Through Analytical Performance Models

Computational modeling and simulation have become essential tools in the quest to better understand the brain’s makeup and to decipher the causal interrelations of its components. The breadth of biochemical and biophysical processes and structures in the brain has led to the development of a large variety of model abstractions and specialized tools, often times requiring high performance computing resources for their timely execution. What has been missing so far was an in-depth analysis of the complexity of the computational kernels, hindering a systematic approach to identifying bottlenecks of algorithms and hardware. If whole brain models are to be achieved on emerging computer generations, models and simulation engines will have to be carefully co-designed for the intrinsic hardware tradeoffs. For the first time, we present a systematic exploration based on analytic performance modeling. We base our analysis on three in silico models, chosen as representative examples of the most widely employed modeling abstractions: current-based point neurons, conductance-based point neurons and conductance-based detailed neurons. We identify that the synaptic modeling formalism, i.e. current or conductance-based representation, and not the level of morphological detail, is the most significant factor in determining the properties of memory bandwidth saturation and shared-memory scaling of in silico models. Even though general purpose computing has, until now, largely been able to deliver high performance, we find that for all types of abstractions, network latency and memory bandwidth will become severe bottlenecks as the number of neurons to be simulated grows. By adapting and extending a performance modeling approach, we deliver a first characterization of the performance landscape of brain tissue simulations, allowing us to pinpoint current bottlenecks for state-of-the-art in silico models, and make projections for future hardware and software requirements.

Heterogeneity in migration strategies of Whooping Cranes

AbstractMigratory birds use numerous strategies to successfully complete twice-annual movements between breeding and wintering sites. Context for conservation and management can be provided by characterizing these strategies. Variations in strategy among and within individuals support population persistence in response to changes in land use and climate. We used location data from 58 marked Whooping Cranes (Grus americana) from 2010 to 2016 to characterize migration strategies in the U.S. Great Plains and Canadian Prairies and southern boreal region, and to explore sources of heterogeneity in their migration strategy, including space use, timing, and performance. Whooping Cranes completed ~3,900-km migrations that averaged 29 days during spring and 45 days during autumn, while making 11–12 nighttime stops. At the scale of our analysis, individual Whooping Cranes showed little consistency in stopover sites used among migration seasons (i.e. low site fidelity). In contrast, individuals expressed a measure of consistency in timing, especially migration initiation dates. Whooping Cranes migrated at different times based on age and reproductive status, where adults with young initiated autumn migration after other birds, and adults with and without young initiated spring migration before subadult birds. Time spent at stopover sites was positively associated with migration bout length and negatively associated with time spent at previous stopover sites, indicating Whooping Cranes acquired energy resources at some stopover sites that they used to fuel migration. Whooping Cranes were faithful to a defined migration corridor but showed less fidelity in their selection of nighttime stopover sites; hence, spatial targeting of conservation actions may be better informed by associations with landscape and habitat features rather than documented past use at specific locations. The preservation of variation in migration strategies existing within this species that experienced a severe population bottleneck suggests that Whooping Cranes have maintained a capacity to adjust strategies when confronted with future changes in land use and climate.

In-FPGA Instrumentation Framework for OpenCL-Based Designs

The productivity achieved when developing applications on high-performance reconfigurable heterogeneous computing (HPRHC) systems is increased by using the Open Computing Language (OpenCL). However, the hardware produced by OpenCL compilers in field-programmable gate arrays (FPGAs) can result in severe performance bottlenecks that are challenging to solve. The problem is compounded by the fact that the generated netlist details are disorganized, making them mostly unreadable and only partially visible to designers. This paper proposes an in-FPGA instrumentation method and a new framework for extracting the FPGA-cycle-accurate timing performances of OpenCL-based designs. The results clearly show that the chosen execution model for OpenCL-based designs strongly affects the timing performance when it is not properly implemented. Our framework is implemented on an HPRHC platform that contains a CPU and two Arria10 FPGAs, and it is evaluated with a wide variety of benchmarks with different complexities. After testing on the reported benchmarks, the average logic overhead for one inserted instrument is 0.2 % of the total amount of adaptive look-up tables (ALUTs) and 0.1 % of the total registers in an FPGA. This resource utilization is between 1.5 and six times lower than those reported in the best previously published works. The scalability of the framework is also evaluated by inserting up to 50 instruments. The experimental results show that the average logic utilization per instrument is 0.19 % of the ALUTs and 0.17 % of the registers in the FPGA when 50 instruments are inserted.

Real Scene Pickup Method of Elemental Image Array Based on Convergent Camera Array

Limited available methods for the generation of an elemental image array (EIA) may be the most severe bottleneck for the promotion and application of integral imaging. To obtain rich information of three dimensional (3D) real scene at a relatively low cost, and solve the pseudoscopic problem in real display mode, we propose an EIA generation method based on convergent camera array through establishing the optical mapping model between the proposed method and the classical computer virtual generation method. We expound the basic principle of solving the pseudoscopic problem by changing the relative position of object and center of the convergent camera array in this paper. The experimental results indicate that the generated EIA can be displayed without the pseudoscopic problem in real display mode, relative position relationships of the objects in 3D real scene can also be accurately restored. Both preliminary optical experiments and the optical analysis are conducted to prove the feasibility and validity of the proposed method.

Prototype system for real-time IM/DD MDM transmission based on multiple-ring-core FMF and degenerate-mode-selective reception.

Multiple-input-multiple-output digital signal processing (DSP) has become a severe bottleneck for mode division multiplexing (MDM) because of its huge computational complexity. In this paper, we propose a novel scheme for real-time DSP-free intensity-modulation/direct-detection (IM/DD) MDM transmission, in which the transmission few-mode fiber (FMF) is characterized by multiple-ring-core structure to suppress modal crosstalk among each LP mode, while each pair of non-circularly-symmetric degenerate modes is simultaneously demultiplexed by a degenerate-mode-selective fiber coupler for DSP-free reception. Based on a 10 km ultralow-modal-crosstalk double-ring-core FMF and a pair of all-fiber 4-LP-mode MUX/DEMUX, we demonstrate the first IM/DD MDM prototype system using commercial single-mode (SM) 10 Gbps SFP + modules and 4K video transceivers without any hardware modifications. The temperature and wavelength dependence are evaluated. The stable Q2-factor performance proves that it can be a smooth evolution scheme from conventional SM IM/DD systems. Moreover, the scheme can be further extended to support more modes with improved FMF design adopting more ring areas.

A Multiorgan Trafficking Circuit Provides Purifying Selection of Listeria monocytogenes Virulence Genes.

Listeria monocytogenes can cause a life-threatening illness when the foodborne pathogen spreads beyond the intestinal tract to distant organs. Many aspects of the intestinal phase of L. monocytogenes pathogenesis remain unknown. Here, we present a foodborne infection model using C57BL/6 mice that have been pretreated with streptomycin. In this model, as few as 100 L. monocytogenes CFU were required to cause self-limiting enterocolitis, and systemic dissemination followed previously reported routes. Using this model, we report that listeriolysin O (LLO) and actin assembly-inducing protein (ActA), two critical virulence determinants, were necessary for intestinal pathology and systemic spread but were dispensable for intestinal growth. Sequence tag-based analysis of microbial populations (STAMP) was used to investigate the within-host population dynamics of wild-type and LLO-deficient strains. The wild-type bacterial population experienced severe bottlenecks over the course of infection, and by 5 days, the intestinal population was highly enriched for bacteria originating from the gallbladder. In contrast, LLO-deficient strains did not efficiently disseminate and gain access to the gallbladder, and the intestinal population remained diverse. These findings suggest that systemic spread and establishment of a bacterial reservoir in the gallbladder imparts an intraspecies advantage in intestinal occupancy. Since intestinal L. monocytogenes is ultimately released into the environment, within-host population bottlenecks may provide purifying selection of virulence genes.IMPORTANCEListeria monocytogenes maintains capabilities for free-living growth in the environment and for intracellular replication in a wide range of hosts, including livestock and humans. Here, we characterized an enterocolitis model of foodborne L. monocytogenes infection. This work highlights a multiorgan trafficking circuit and reveals a fitness advantage for bacteria that successfully complete this cycle. Because virulence factors play critical roles in systemic dissemination and multiple bottlenecks occur as the bacterial population colonizes different tissue sites, this multiorgan trafficking circuit likely provides purifying selection of virulence genes. This study also serves as a foundation for future work using the L. monocytogenes-induced enterocolitis model to investigate the biology of L. monocytogenes in the intestinal environment.

Amorphous Sulfur Decorated Gold Nanowires as Efficient Electrocatalysts toward Ambient Ammonia Synthesis

The electrocatalytic nitrogen to ammonia has attracted intensive attention but suffers from severe bottlenecks of low Faradaic efficiency and NH3 yield. Herein, we report a two-step process to prepare amorphous sulfur multilayer decorated gold nanowires (S/Au NWs) as superior catalysts for electrochemical NH3 production. The chemisorbed sulfur multilayers are favorable for selective hydrogenation of nitrogen to ammonia and provide sufficient exposed active sites for nitrogen reduction reaction (NRR). The S/Au NWs show a high NH3 yield of 21.04 μg h–1 mg–1cat. and Faradaic efficiency of 15.34%, as well as excellent durability. The superior NRR performance of Au catalysts with S chemisorption would stimulate interest toward NRR catalyst design of sulfur bonding to metal surfaces.

The conflict within: origin, proliferation and persistence of a spontaneously arising selfish mitochondrial genome.

Mitochondrial genomes can sustain mutations that are simultaneously detrimental to individual fitness and yet, can proliferate within individuals owing to a replicative advantage. We analysed the fitness effects and population dynamics of a mitochondrial genome containing a novel 499 bp deletion in the cytochrome b(1) (ctb-1) gene (Δctb-1) encoding the cytochrome b of complex III in Caenorhabditis elegans. Δctb-1 reached a high heteroplasmic frequency of 96% in one experimental line during a mutation accumulation experiment and was linked to additional spontaneous mutations in nd5 and tRNA-Asn. The Δctb-1 mutant mitotype imposed a significant fitness cost including a 65% and 52% reduction in productivity and competitive fitness, respectively, relative to individuals bearing wild-type (WT) mitochondria. Deletion-bearing worms were rapidly purged within a few generations when competed against WT mitochondrial DNA (mtDNA) bearing worms in experimental populations. By contrast, the Δctb-1 mitotype was able to persist in large populations comprising heteroplasmic individuals only, although the average intracellular frequency of Δctb-1 exhibited a slow decline owing to competition among individuals bearing different frequencies of the heteroplasmy. Within experimental lines subjected to severe population bottlenecks (n = 1), the relative intracellular frequency of Δctb-1 increased, which is a hallmark of selfish drive. A positive correlation between Δctb-1 and WT mtDNA copy-number suggests a mechanism that increases total mtDNA per se, and does not discern the Δctb-1 mitotype from the WT mtDNA. This study demonstrates the selfish nature of the Δctb-1 mitotype, given its transmission advantage and substantial fitness load for the host, and highlights the importance of population size for the population dynamics of selfish mtDNA. This article is part of the theme issue 'Linking the mitochondrial genotype to phenotype: a complex endeavour'.

Chromosome level assembly and comparative genome analysis confirm lager-brewing yeasts originated from a single hybridization

BackgroundThe lager brewing yeast, S. pastorianus, is a hybrid between S. cerevisiae and S. eubayanus with extensive chromosome aneuploidy. S. pastorianus is subdivided into Group 1 and Group 2 strains, where Group 2 strains have higher copy number and a larger degree of heterozygosity for S. cerevisiae chromosomes. As a result, Group 2 strains were hypothesized to have emerged from a hybridization event distinct from Group 1 strains. Current genome assemblies of S. pastorianus strains are incomplete and highly fragmented, limiting our ability to investigate their evolutionary history.ResultsTo fill this gap, we generated a chromosome-level genome assembly of the S. pastorianus strain CBS 1483 from Oxford Nanopore MinION DNA sequencing data and analysed the newly assembled subtelomeric regions and chromosome heterozygosity. To analyse the evolutionary history of S. pastorianus strains, we developed Alpaca: a method to compute sequence similarity between genomes without assuming linear evolution. Alpaca revealed high similarities between the S. cerevisiae subgenomes of Group 1 and 2 strains, and marked differences from sequenced S. cerevisiae strains.ConclusionsOur findings suggest that Group 1 and Group 2 strains originated from a single hybridization involving a heterozygous S. cerevisiae strain, followed by different evolutionary trajectories. The clear differences between both groups may originate from a severe population bottleneck caused by the isolation of the first pure cultures. Alpaca provides a computationally inexpensive method to analyse evolutionary relationships while considering non-linear evolution such as horizontal gene transfer and sexual reproduction, providing a complementary viewpoint beyond traditional phylogenetic approaches.

Bottleneck and selection in the germline and maternal age influence transmission of mitochondrial DNA in human pedigrees

Heteroplasmy-the presence of multiple mitochondrial DNA (mtDNA) haplotypes in an individual-can lead to numerous mitochondrial diseases. The presentation of such diseases depends on the frequency of the heteroplasmic variant in tissues, which, in turn, depends on the dynamics of mtDNA transmissions during germline and somatic development. Thus, understanding and predicting these dynamics between generations and within individuals is medically relevant. Here, we study patterns of heteroplasmy in 2 tissues from each of 345 humans in 96 multigenerational families, each with, at least, 2 siblings (a total of 249 mother-child transmissions). This experimental design has allowed us to estimate the timing of mtDNA mutations, drift, and selection with unprecedented precision. Our results are remarkably concordant between 2 complementary population-genetic approaches. We find evidence for a severe germline bottleneck (7-10 mtDNA segregating units) that occurs independently in different oocyte lineages from the same mother, while somatic bottlenecks are less severe. We demonstrate that divergence between mother and offspring increases with the mother's age at childbirth, likely due to continued drift of heteroplasmy frequencies in oocytes under meiotic arrest. We show that this period is also accompanied by mutation accumulation leading to more de novo mutations in children born to older mothers. We show that heteroplasmic variants at intermediate frequencies can segregate for many generations in the human population, despite the strong germline bottleneck. We show that selection acts during germline development to keep the frequency of putatively deleterious variants from rising. Our findings have important applications for clinical genetics and genetic counseling.

The Marketing System of Agricultural Products in Bangladesh: A Case Study from Sylhet District

The monotonous demand for rice in Bangladesh is shifting rapidly. Positive economic growth andurbanization are the key factors behind this change. The agricultural marketing system is traditional andnot properly functional. A group of intermediaries controls the market and supply chain of agriculturalproducts. A proper functional market is vital for food producers, processors, traders, and consumers.Bangladesh has become self-sufficient in cereal crop production, but the production of horticulturalcrops is not sufficient yet. This situation has led to severe bottlenecks in the processes of marketing. Thestudy aimed at assessing the present marketing system of agricultural products in Sylhet and atdetermining the marketing constraints and potentialities of farm products in this area. It was found thatthe net marketing margin was the highest in the case of the retailer and the lowest in case of Aratdar. ButReturn on operating capital was the highest for the Aratdar because they did not need to purchase theproduct they handle. The farmers identified the high price of seed as the topmost constraint in productionand marketing systems. Re-structuring of Market Management System and updating regulatory andinstitutional set up are highly essential for the proper functioning of the market system.

Regularization strategies for deep-learning-based salt model building

Salt model building has long been considered a severe bottleneck for large-scale 3D seismic imaging projects. It is one of the most time-consuming, labor-intensive, and difficult-to-automate processes in the entire depth imaging workflow requiring significant intervention by domain experts to manually interpret the salt bodies on noisy, low-frequency, and low-resolution seismic images at each iteration of the salt model building process. The difficulty and need for automating this task is well-recognized by the imaging community and has propelled the use of deep-learning-based convolutional neural network (CNN) architectures to carry out this task. However, significant challenges remain for reliable production-scale deployment of CNN-based methods for salt model building. This is mainly due to the poor generalization capabilities of these networks. When used on new surveys, never seen by the CNN models during the training stage, the interpretation accuracy of these models drops significantly. To remediate this key problem, we have introduced a U-shaped encoder-decoder type CNN architecture trained using a specialized regularization strategy aimed at reducing the generalization error of the network. Our regularization scheme perturbs the ground truth labels in the training set. Two different perturbations are discussed: one that randomly changes the labels of the training set, flipping salt labels to sediments and vice versa and the second that smooths the labels. We have determined that such perturbations act as a strong regularizer preventing the network from making highly confident predictions on the training set and thus reducing overfitting. An ensemble strategy is also used for test time augmentation that is shown to further improve the accuracy. The robustness of our CNN models, in terms of reduced generalization error and improved interpretation accuracy is demonstrated with real data examples from the Gulf of Mexico.

A 50K SNP array reveals genetic structure for bald eagles (Haliaeetus leucocephalus)

Bald eagles (Haliaeetus leucocephalus) underwent a severe population bottleneck in the mid-1900s due to Dichlorodiphenyltrichloroethane (DDT) use as an insecticide. After its ban in 1972, the population began to recover with the increase being attributed to banning DDT and reintroduction/translocation programs. Although bald eagles have increased and may be in a phase of exponential growth, many populations continue to experience anthropogenic stressors. Assessing levels of standing genetic variation and the partitioning of genetic variation within and among populations is critical for the development of informed conservation management plans. To begin addressing these concerns, we developed a custom 50K Affymetrix Axiom myDesign single nucleotide polymorphism (SNP) array and performed preliminary population genomic analyses on geographically disparate populations of bald eagles to test the utility of this SNP array for assessing levels of standing genetic variation in the gene pool and determining the partitioning of genetic variation within and among populations. To develop the array, a combination of RAD-tag sequencing and genome sequencing was used with the final chip consisting of 50,789 SNPs in both genic and intergenic regions of the genome. After genotyping 169 hatchlings, 45,952 SNPs from the array were found to be of quality and were used in Structure, Admixture, and a discriminant analysis of principal components (DAPC) analyses. Results from all of these analyses indicate that despite the significant population bottleneck, sufficient genetic variation is detectable within the bald eagle gene pool. Moreover, based on our disparate geographic sampling of bald eagles, our preliminary analyses indicate statistically significant partitioning of the genetic variation among broad sampling areas.