Review

The relationship of sea level changes to climatic change in northeast Asia and northern North America during the last 75 ka B.P.

  • Received: 09 November 2018 Accepted: 02 February 2018 Published: 26 February 2019
  • The Arctic air mass is the cold, dry body of air slowly moving eastwards around the North Pole in the northern hemisphere. Its southern boundary consists of four planetary waves known as the Rossby waves that mark the interface with subtropical air bringing heat polewards. The Arctic air mass is constantly being modified by the addition of heat and moisture over the oceans, as well as by winter cooling over the land masses due to limited incoming solar radiation and constant reradiation of heat into the atmosphere. The coldest air in winter is located over northeastern Siberia and moves east, cooling Canada. Warm ocean currents add large quantities of heat to the air mass moving over them, but without this addition of heat, the Arctic air mass becomes significantly colder. Research in Tibet and Northeast Asia on depression of sea level shows that during the Late Wisconsin cold event (65–10 ka B.P.), vast quantities of sea water were sequestered on land primarily as ice sheets, exposing the sea bed in the Bering Strait from 50–10 ka B.P. together with the bottom of the South China Sea between 30–20 ka B.P.. The East China Monsoon failed to reach Tibet and much of Northeast China, resulting in severe cooling of northeast Siberia and northern Tibet. This, in turn, caused severe cooling in eastern Canada together with the development of a vast, predominantly cold-based ice sheet. As the sea levels started to rise (about 19 ka B.P.), the East China Monsoon slowly redeveloped and a gradual warming took place on both continents. However, along the west side of the North American Cordillera, the Late Wisconsin glaciation only began in 29 ka B.P. but continued along the west coast until about 10 ka B.P. This paper explores the relationship of the Late Wisconsin history on the two continents, together with the mechanisms causing the landforms and climatic differences. Finally, the probable effects of these climatic changes on the early peopling of North America are discussed.

    Citation: Stuart A. Harris. The relationship of sea level changes to climatic change in northeast Asia and northern North America during the last 75 ka B.P.[J]. AIMS Environmental Science, 2019, 6(1): 14-40. doi: 10.3934/environsci.2019.1.14

    Related Papers:

    [1] Mohammad Aman YAQOBI, Hidehito Matayoshi, Natarajan Prabaharan, Hiroshi Takahashi, Ashraf M. Hemeida, Senjyu Tomonobu . Interconnected standalone DC microgrid fault protection based on Self-Adaptive DC fault current limiter with hybrid solid state circuit breaker. AIMS Energy, 2021, 9(5): 991-1008. doi: 10.3934/energy.2021045
    [2] Mohamed G Moh Almihat . An overview of AC and DC microgrid energy management systems. AIMS Energy, 2023, 11(6): 1031-1069. doi: 10.3934/energy.2023049
    [3] H. H. H. De Silva, D. K. J. S. Jayamaha, N. W. A. Lidula . Review on design and control of solid state transformer based microgrids. AIMS Energy, 2019, 7(6): 901-923. doi: 10.3934/energy.2019.6.901
    [4] Nagaraj C, K Manjunatha Sharma . Fuzzy PI controller for bidirectional power flow applications with harmonic current mitigation under unbalanced scenario. AIMS Energy, 2018, 6(5): 695-709. doi: 10.3934/energy.2018.5.695
    [5] Nan Liu, Thomas G. Habetler . Design of a parallel-LC-compensated on-board charger for universal inductive charging in electric vehicles. AIMS Energy, 2016, 4(5): 658-674. doi: 10.3934/energy.2016.5.658
    [6] Abdulrahman J. Babqi, NasimUllah, Ahmed Althobaiti, Hend I. Alkhammash, Asier Ibeas . Current model predictive fault-tolerant control for grid-connected photovoltaic system. AIMS Energy, 2022, 10(2): 273-291. doi: 10.3934/energy.2022015
    [7] Pooja Khandare, Sanjay Deokar, Arati Dixit . Relay Coordination and Optimization techniques using DWT-Differentiation Algorithms for Fault Detection in Microgrid. AIMS Energy, 2020, 8(4): 563-579. doi: 10.3934/energy.2020.4.563
    [8] Ahsan Iqbal, Ayesha Ayoub, Asad Waqar, Azhar Ul-Haq, Muhammad Zahid, Syed Haider . Voltage stability enhancement in grid-connected microgrid using enhanced dynamic voltage restorer (EDVR). AIMS Energy, 2021, 9(1): 150-177. doi: 10.3934/energy.2021009
    [9] T S S Senarathna, K T M Udayanga Hemapala . Review of adaptive protection methods for microgrids. AIMS Energy, 2019, 7(5): 557-578. doi: 10.3934/energy.2019.5.557
    [10] Awatif Nadia, Md. Sanwar Hossain, Md. Mehedi Hasan, Sinthia Afrin, Md Shafiullah, Md. Biplob Hossain, Khondoker Ziaul Islam . Determination of transmission reliability margin for brownout. AIMS Energy, 2021, 9(5): 1009-1026. doi: 10.3934/energy.2021046
  • The Arctic air mass is the cold, dry body of air slowly moving eastwards around the North Pole in the northern hemisphere. Its southern boundary consists of four planetary waves known as the Rossby waves that mark the interface with subtropical air bringing heat polewards. The Arctic air mass is constantly being modified by the addition of heat and moisture over the oceans, as well as by winter cooling over the land masses due to limited incoming solar radiation and constant reradiation of heat into the atmosphere. The coldest air in winter is located over northeastern Siberia and moves east, cooling Canada. Warm ocean currents add large quantities of heat to the air mass moving over them, but without this addition of heat, the Arctic air mass becomes significantly colder. Research in Tibet and Northeast Asia on depression of sea level shows that during the Late Wisconsin cold event (65–10 ka B.P.), vast quantities of sea water were sequestered on land primarily as ice sheets, exposing the sea bed in the Bering Strait from 50–10 ka B.P. together with the bottom of the South China Sea between 30–20 ka B.P.. The East China Monsoon failed to reach Tibet and much of Northeast China, resulting in severe cooling of northeast Siberia and northern Tibet. This, in turn, caused severe cooling in eastern Canada together with the development of a vast, predominantly cold-based ice sheet. As the sea levels started to rise (about 19 ka B.P.), the East China Monsoon slowly redeveloped and a gradual warming took place on both continents. However, along the west side of the North American Cordillera, the Late Wisconsin glaciation only began in 29 ka B.P. but continued along the west coast until about 10 ka B.P. This paper explores the relationship of the Late Wisconsin history on the two continents, together with the mechanisms causing the landforms and climatic differences. Finally, the probable effects of these climatic changes on the early peopling of North America are discussed.


    Skin Cutaneous Melanoma (SKCM) is a deadly and invasive kind of skin cancer that develops when melanocytes in the skin epidermis' basal layer undergo malignant transformation. The incidence of melanoma is rising faster in developed countries. There will be approximately 99,780 new patients of melanoma in the United States in 2022, nearly double the 53,600 cases in 2002 [1]. Environmental, immunological, and genetic factors contribute to melanoma development. The primary treatment for non-metastatic melanoma is surgery. However, not all patients with the disease are candidates for it due to anatomical location, number of lesions, and recurrence rate following surgical resection. For the aforementioned circumstance, the response rate has been enhanced by the development of new medications with specific targeting and immunotherapy. Effective melanoma treatment options include immunotherapy and targeted therapy. Currently, immune checkpoint medications that target CTLA-4 and PD-1 monoclonal antibodies are routinely utilized throughout the world to treat melanoma [2]. Relevant clinical studies have demonstrated good patient response to immunotherapy, such as immune checkpoint inhibitors. Unfortunately, many patients with the disease develop immune resistance to immunotherapy, which may be due to the complicated interactions between tumor cells and the tumor microenvironment [3]. The tumor microenvironment is heterogeneous and consists of various types of cells and extracellular matrix with unique Spatio-temporal interactions between these cellular components that influence the behavior of the tumor [4]. During melanoma development, the proliferation and apoptosis of tumor cells are determined by the activity of immune cells [5]. In the background of cancer immunotherapy, intratumoral heterogeneity influences the therapeutic effectiveness of anticancer medications. The tumor microenvironment (TME), which is heterogeneous and has unique spatiotemporal interactions between all cellular constituents, can be rationally explained as the source of therapeutic pressure and cancer adaptability [6]. It has been shown that tumor immune subtypes can be determined on account of the combined abundance of immune cells, which will help select suitable therapeutic strategies [7]. Therefore, constructing an immune cell risk score that can predict prediction is crucial for melanoma prevention and treatment.

    ScRNA-seq technology has recently made advancements that have made it easier to comprehend the heterogeneity of tumor cells. ScRNA-seq analysis is unquestionably a valuable method for learning the characteristics of various cell types within cells and is a critical component of single-cell sequencing technology [8]. For instance, a recent study found that scRNA-seq can analyze the genetic signature of drug-resistant clusters, find clusters that cause drug resistance, and assess cell clusters in tumor tissue [9]. ScRNA-seq can enhance the interpretation of how vital cellular constituents co-stimulate tumor emergence behavior and describe tumors' cellular composition [10]. This investigation field is also an essential part of cancer investigation. However, applying scRNA-seq in this field needs further in-depth studies.

    The RNA-Seq TPM (Transcripts Per Kilobase of exon model per Million mapped reads) data and clinical data of SKCM samples were downloaded from the TCGA database (https://portal.gdc.cancer.gov/https://www.gtexportal.org/home/index.html). The corresponding information for melanoma was downloaded from the Gene Expression Omnibus (GEO) [11] (https://www.ncbi.nlm.nih.gov/) database for scRNA-seq data: GSE72056 and bulk RNA-seq data: GSE65904. The dataset of TCGA was utilized as the training set, and the GSE65904 cohort was utilized as the verification set.

    The ssGSEA method was employed to calculate the abundance of 23 immune cells in skin cutaneous melanoma according to the expression data of tumor samples [12]. Prognosis-relevant immune cells were screened on account of univariate COX regression analysis, followed by constructing the ICRS with predictive value on account of prognosis-relevant immune cells using the Lasso [13]. The immune cell risk score was calculated as follows:

    ICRS=i=1nCoef_i×Xi

    where Coef represents the coefficient of prognostic immune cells and X denotes the abundance of predictive immune cells. The "glmnet" package was utilized to perform LASSO-COX regression analysis to remove excess prognostic immune cells in order to build an immune cell risk score subsequently. The entire sample was classified as different ICRS on account of the median risk score of the training cohort. Kaplan-Meier survival curves were plotted to determine the prognosis value of the immune cell risk score for melanoma. The "ESTIMATE" algorithm was utilized to assess the immune score, stromal score, and ESTIMATE score for each patient in the melanoma cohort [14]. CIBERSORT is a generalized deconvolution algorithm for quantifying the cellular fraction of a large number of tissues on account of gene expression matrices to evaluate the proportion of immune cells of 22 phenotypes. To confirm the precision of the inverse convolution algorithm, the threshold was set to a P value < 0.05.

    GSVA switched the analysis object from genes to gene sets for pathway-level differential analysis. GSVA portrays the overall differences between different immune cell risk scores on account of the training cohort. The Wilcoxon rank sum test was next utilized to calculate differential genes between different ICRS. |log2FoldChange| > 1 and P-Value < 0.05 were utilized as significance criteria to determine the differential genes. Gene Ontology (GO) enrichment analysis and Genomes (KEGG) pathway analysis of differential genes were performed using the "clusterProfiler" package of R software.

    Two machine learning methods were employed to bolster hub genes affecting melanoma prognosis. The LASSO is utilized to build a penalty function to obtain a more refined model by using the "glmnet" package. It is a biased estimation for data with multiple covariances that improve statistical models' prediction accuracy and understandability. The package "randomForest" [15] was utilized to build a random forest model to filter DEGs. In the random forest classifier, the number of random seeds and decision trees is set to 123,456 and 300, respectively, to obtain a stable model error and high accuracy. The differential genes with utmost significance were screened as essential genes affecting melanoma prognosis for subsequent analysis. For the accuracy of the results, the overlapping genes of LASSO and RF are regarded as hub genes affecting melanoma's prognosis [16].

    The scRNA-seq was downloaded from the GEO database, and the study included 4645 cells from 19 patients. Cells with more than 10,000 genes and less than 200 genes were discarded. These cells did not include mitochondrial genes. Ultimately, a total of 4612 cells were included in this study for follow-up analysis. First, we normalized the merged data by log-normalization and identified the first 2000 highly variable genes by the "FindVariableFeatures" function. A principal component analysis (PCA) was subsequently performed to select significant main components (PCs) using the "JackStraw" and "ScoreJackStraw" functions. The "FindAllMarkers" function is utilized to recognize characteristic genes for each cell population [17]. The "RunTSNE" function is then utilized for cell clustering and visualization analysis of TSNE. T-SNE is an algorithm that converts multidimensional data into two-dimensional or three-dimensional data by using accountable flow theory and topological algorithm while maintaining the original data distribution and similarity to achieve the effect of dimensionality reduction. Marker genes are subsequently annotated using the "singleR" package and corrected according to their features using CellMarker.

    Cellular communication between cell types is assessed by the "CellChat" package [18]. Specifically, Cellular communication is the process by which a portion of cells in vivo sends signals and the target cells receive them and translate them into changes in cellular function. The mechanism of communication between individual cell types was obtained by comparing the differences in ligand and receptor gene expression between cells of different sample groups across cell types. In this process, the ligand is considered the output signal and the receptor is considered the input signal. The ligand-receptor interaction score reflects the possible interactions between cells.

    We downloaded 812 normal skin samples from the GTEx database for joint analysis, with 463 melanoma samples from the TCGA database. Boxplot visualizes the differential expression levels of hub genes between tumor and normal groups. Kaplan-Meier curves allowed us to determine the impact of hub genes on melanoma prognosis.

    The flow chart of this study is shown in Figure 1. In the training cohort, 17 immune cells associated with survival were ascertained using univariate COX regression analysis and subsequently screened by LASSO (Figure 2A, B) to obtain the ICRS constructed from Activated CD8 T cells and Immature B cells. The ICRS was calculated on account of the formula: the abundance of Activated CD8 T cell × (-0.429966) + the abundance of Immature B cell × (-1.514149). The distribution of patients, state of survival, and expression abundance of Activated CD8 T cells and Immature B cells between the different ICRS are shown in Figure 2C. Kaplan-Meier curves indicated that patients in the high-ICRS group had great shorter survival times than those in the low-ICRS group (P < 0.05) (Figure 2D). Next, immune characteristics between the different ICRS groups were analyzed using ESTIMATE and CIBERSORT methods. We noted that the immune score, ESTIMATE score and stromal score were great higher in the low -ICRS group (Figure 2EG), while the tumor purity was considerably higher in the high-ICRS group (Figure 2H). Furthermore, it is worth noting that in the results of immune cell infiltration, both T cell CD8 and B cells native were discovered to be more highly expressed in the low immune group (Figure 2I, J), which is consistent with our previous analysis (P < 0.001).

    Figure 1.  Workflow of this study.
    Figure 2.  Construction and evaluation of the ICRS in the training cohort. (A, B) Coefficient path diagram and cross-verification curve of LASSO. (C) Distribution of patients, state of survival and abundance of immune cells between different ICRS groups. (D) Kaplan-Meier survival curve analysis. (E–H) Differential profiles of the immune score, ESTIMATE score, stromal score and tumor purity. (I) Heatmap showing the differences of 23 immune cells between different groups. (J) Boxplot showing the difference of 22 immune cells between different groups.

    To further determine the accuracy of the ICRS, GSE65904 was employed as the verification cohort for analysis in this study. The abundance of 23 immune cells was evaluated by ssGSEA and modeled using the same formula as the training cohort. Next, melanoma patients were divided into different ICRS groups on the strength of the same cut-off of the ICRS as the training cohort. The distribution of patients, state of survival and the abundance of Activated CD8 T cell and Immature B cell expression in the different ICRS groups are shown (Figure 3A). Survival curves show a better prognosis for patients in the high ICRS group (P < 0.05) (Figure 3B). Next, we noted that immune scores, ESTIMATE scores, and stromal scores were significantly higher in the low-ICRS group of the verification cohort (Figure 3CE). Meanwhile, the purity of tumors in the high subgroup significantly exceeded that of the ground subgroup. (Figure 3F). Furthermore, most immune cells were significantly more elevated in the low-ICRS group (Figure 3G, H). These results are consistent with the training cohort, showing that the ICRS model identifies immune subtypes of melanoma well and effectively predicts the prognosis of melanoma patients.

    Figure 3.  External verification of the ICRS accuracy in the verification cohort. (A) Distribution of patients, state of survival and abundance of immune cells between different ICRS groups. (B) Kaplan-Meier survival curve analysis. (C–F) Differential profiles of immune score, ESTIMATE score, stromal score and tumor purity. (G) Heatmap showing the differences of 23 immune cells between different groups. (H) Boxplot showing the difference of 22 immune cells between different groups.

    Analyses were executed in the training cohort to elucidate the functional differences between the different ICRS groups. First, GSVA was utilized to illuminate the differences in pathways between the two groups. The different ICRS groups are mainly enriched in immune-related pathways such as allograft rejection and antigen processing and presentation (Figure 4A). Next, differential analysis between different ICRS groups was identified using the Wilcoxon rank sum test, and 972 differential genes were obtained (Figure 4B). GO or KEGG enrichment analysis of 972 differential genes to identify the functions of these genes or the pathways involved in regulation. The results of GO enrichment analysis suggested that in terms of biological processes (BP), differential genes were mainly involved in processes such as leukocyte-mediated immunity and lymphocyte-mediated immunity, and in terms of cellular components (CC), differential genes were associated with T cell receptor complexes, MHC protein complex, and other components. In terms of molecular functions (MF), differential genes are mainly immunoglobulin receptor binding, cytokine activity, chemokine receptor binding and other procedures (Figure 4D). KEGG pathways are primarily enriched in Cell adhesion molecules and hematopoietic cell lineage (Figure 4C). These differential terms may be potentially responsible for prolonged survival time in low-ICRS groups.

    Figure 4.  Differential Analysis between the different ICRS groups in the training cohort. (A) Heatmap of the first 25 critical terms of GSVA. (B) Volcano of differential genes between the different ICRS groups. (C) GO enrichment analysis of differential genes. (D) KEGG pathway analysis of differential genes.

    We further screened these DEGs using Lasso and random forest classifiers to filter the hub genes with predictive values for melanoma. LASSO regression analysis screened 45 significant candidate hub genes (Figure 5A). In the random forest, the number of random seeds is set to 123,456. We chose 500 trees as the random forest model's parameters, representing the model's stable error (Figure 5B). Finally, we screened the 43 hub genes with the highest Gini index scores. Finally, we filtered the 43 genes with the highest Gini index scores, and we obtained 45 hub genes in the Lasso analysis. The intersection of five potential candidate genes in RF and LASSO was shown by a Venn diagram (Figure 5C), and five genes (CXCL13, FCRL3, NCF1C, PLA2G2D, and SLAMF6) were identified as hub signature genes for subsequent analysis. Next, correlations among the five hub genes were analyzed using correlation analysis (Figure 5D).

    Figure 5.  Machine learning screens candidate diagnostic biomarkers. (A) Biomarker screening in the Lasso model. (B) The relationship between the number of decision trees and the model error. (C) The Venn diagram shows that the two algorithms described above identified five candidate diagnostic genes. (D) Correlation analysis of five hub genes.

    To better investigate the role of 5 hub genes in melanoma prognosis, we analyzed scRNA-sequencing data, which included 4645 cells. A total of 4612 cells and 22,289 genes eventually passed the quality control filter and were utilized for further data analysis. Unsupervised clustering analysis was performed using the whole transcriptome of all single cells after batch effect correction, and the results were visualized using TSNE for downscaling. By using the "singR" package to annotate the cells, we obtained a total of 8 clusters of cells: T cells, B cell, Neurons, Monocyte, Tissue stem cells, NK cell, Fibroblasts, Endothelial cells, and visualized the top 5 significant differential genes in each cluster by heatmap (Figure 6A, B). Finally, we analyzed the distribution of the five hub genes in each cell cluster. We showed that SLAMF6 and FCRL3 were mainly distributed in T cells, B cells, and NK cell subpopulations, NCF1C was told primarily on B cells and Monocyte, CXCL9 was distributed in T cells, while PLA2G2D was not expressed in almost any cells (Figure 6CG).

    Figure 6.  Visualization of scRNA-seq. (A) TSEN plot of all melanoma cells. (B) Heatmap of the top five DEGs in cell subpopulations. (C–G) Distribution of five key genes in different cell clusters.

    To better comprehend the intercommunication of different cell types in melanoma, we analyzed the interactions between eight cell types using cellular communication analysis. And further extracted individual cells to observe the communication between this cell and other cells (Figure 7A), and it was found that cellular communication between T cells and B cells was more assertive (Figure 7B). Next, we analyzed the cellular communication involved in the core gene. The results found that the core gene CXCL13 is present in the CXCL pathway between T cells and b cells, where CXCL13 is the ligand and CXCR5 is the ligand, so we further investigated the communication role and cellular action type analysis between immune cells in the CXCL pathway (Figure 7C). It was found that B cells, Endothelial cells, monocytes and Fibroblasts can act as both ligand and receptor cells in this pathway, where Endothelial cells, monocytes, and Fibroblasts can send signals to other cells. In contrast, T cells, Neurons, Tissue stem cells, and NK cells, can only act as ligand cells (Figure 7D, E). These results elucidate the possible interactions between these cell types and the cellular pathways involved in core genes. These help us further investigate the combined role of cell types and hub genes in melanoma prognosis.

    Figure 7.  Analysis of cellular communication in melanoma. (A) The number of interactions between immune cells. (B) Communication between B cells and other cells. (C) Cellular communication at the level of ligand-receptor pairs. (D) Hierarchical map of cellular communication demonstrating pathways. (E) Analysis of cellular action types.

    To investigate the expression characteristics of hub genes between normal samples and melanoma, we downloaded 812 normal skin samples from the GTEx database for joint analysis with 463 melanoma samples from the TCGA database to ascertain the expression differences of hub genes between the tumor and normal groups. The five hub genes, CXCL13, FCRL3, NCF1C, PLA2G2D, and NCF1C, were highly expressed in normal samples with significant significance (Figure 8KO). Next, further survival curve analysis showed that the five hub genes, CXCL13, FCRL3, PLA2G2D, NCF1C, and SLAMF6, had longer survival times for high expression. Although the gene PLA2G2D was insignificant in the verification group (Figure 8AJ). The above results suggest that these five genes are good prognostic genes and may play a key role in influencing the recurrence of melanoma.

    Figure 8.  Verification of hub genes. (A–E) Survival analysis of five hub genes in the training cohort. (F–J) Survival analysis of five hub genes in the verification cohort. (K–O) Expression characteristics of CXCL13, FCRL3, PLA2G2D, NCF1C, and SLAMF6 in normal samples and melanoma.

    Melanoma is composed of many different cells and the interaction between these cells is associated with tumor formation, metastasis, treatment resistance and immune infiltration [19]. These different cells interact with each other through ligand-receptor interactions. Because of the importance of this intercellular communication for patient prognosis, therapeutic approaches targeting intercellular interactions are becoming increasingly popular in clinical research. However, there have not been sufficient studies to identify the genes involved in intercellular communication.

    In this study, the TCGA dataset as a training cohort, the abundance of 23 immune cells was assessed by ssGSEA, and a combination of univariate and LASSO COX regression analysis was utilized to screen for Activated CD8 T cells and Immature B cells, which were significantly relevant with melanoma prognosis. A validated ICRS was constructed on account of these two immune cells. We investigated the model's validity in predicting melanoma survival, showing that patients in the low-ICRS group had a better prognosis. The ICRS model also had an excellent diagnostic performance in the external verification cohort. These results indicate that prognostic models constructed on account of immune cells can precisely predict the survival rate of melanoma patients.

    In both cell types, the risk factors for T and B cells are less than 0, indicating that these two immune cells are protective factors in melanoma. Alterations in immune cell content play a significant impact in melanoma progression. Studies have reported that B cell in patients with advanced melanoma is pivotal for an efficient anti-tumor immune response and stable disease control. At the same time, B cells could play a crucial role in regulating early tumor progression in melanoma patients [20]. On the other hand, activated CD8 T cells were found in our study to play a crucial role in melanoma development. An increased number of CD8+ T cells can control tumor growth [21].

    With the development of bioinformatics, many risk models have been constructed to assess the prognostic outcome of melanoma patients. Song et al. created a predictive model that utilizes coagulation-related genes to predict patient survival and facilitate clinical treatment management [22]. Yang et al. reported that the analysis of cuproptosis-related lncRNAs in SKCM opens new directions for SKCM treatment [23]. Fu et al. constructed an immunogenic cell death(ICD) gene-related model to initially explore the ICD and possible signaling pathways related to melanoma prognosis [24]. Nonetheless, these studies have all been multi-gene combination analyses to predict the prognostic impact on melanoma. Still, constructing predictive models with two immune cells, T cells and B cells, to expect melanoma survival is not reported. Notably, a predictive model built with T and B cells could represent the immune profile of melanoma and afford a basis for assessing the response to clinical treatment of melanoma.

    In addition, the difference genes between the different ICRS groups were obtained by The Wilcoxon rank sum test. The different genes were discovered to be mainly enriched in immune-related pathways by GO and KEGG enrichment analysis. GSVA also elucidated the pathway differences between the different ICRS groups. On account of this, we further screened the hub genes by Lasso and random forest algorithms, and five hub genes were screened: CXCL13, FCRL3, NCF1C, PLA2G2D and SLAMF6. In a previous study, CXCL13 was found to be a prognostic biomarker in melanoma patients treated with immune checkpoint inhibition [25], which enrolls B cells into the tumor microenvironment, and CXCL13 decreases regulatory B cells in the tumor microenvironment and restrains tumor growth [26]. CXCL13 was identified as an immune-related biomarker relevant to tumor formation and prognosis in patients with melanoma [27]. These are consistent with our findings. FCRL3 accelerates TLR9-induced B-cell activation and inhibits plasma cell decomposition [28] and has also been validated as an underlying prognostic biomarker relevant to melanoma metastasis [29]. NCF1 has been reported to cause an autosomal recessive disorder, chronic sarcoidosis [30]. Its association with cancer and melanoma has not been reported and deserves further exploration in the future. PLA2G2D is associated with melanoma metastasis, and its more in-depth mechanisms need to be further explored. Anti-SLAMF6 was found to be effective in reclaiming CD8+ T cell function and directly influencing tumor progression [31]. SLAMF6 and SLAM-related protein levels in CD8+ T lymphocytes were lower in patients with severe repletion anemia than in normal controls and were inversely associated with CD8+ T lymphocyte functional molecular levels [32]. SLAMF6 is a suppressive immune receptor, and its deletion allows powerful CD8+ T cells to eliminate tumors [33]. Its relationship with T cells is closely related, but the mechanism of its effect on melanoma has not been explored.

    To further investigate the hub genes relevant to melanoma prognosis, we integrated bulk and scRNA sequencing data for analysis. In this report, the data set GSE72056 was obtained from the GEO database, including a total of 4645 cells from 19 melanoma patients [34]. The cells were annotated by singR software package, and we obtained eight cell populations of T cells, B cells, Neurons, Monocyte, tissue stem cells and NK cells. The present study used the interaction between T cells and B cells to construct the model. Tumor-associated B cells can attract effector T cells into the tumor and protect them. They can help activate the immune role of T cells, acting together with other immune components [35]. In addition, B cells may also activate and enroll other immune effector cells by secreting a range of cytokines (including TNF, IL-2, IL-6, and INF-γ) [36]. Several studies in cancer have now shown that the existence of TLS and B cells in tumor tissue is relevant to a better prognosis and that the anti-tumor effect of T cells is more substantial when B cells are present [37]. The tumor microenvironment also contains many immunosuppressive cells, such as mesenchymal cells, fibroblasts, regulatory T cells (Treg), macrophages (TAM), and myeloid-derived suppressor cells (MDSC). Treg cells can reduce anti-tumor immunity and accelerate tumor proliferation and metastasis by suppressing the functions of CD8+ T cells, NK cells, B cells and APC cells [38,39]. It has also been reported that melanoma growth and angiogenesis are precipitated after the inhibition of sensory neuron activity and decreased after the overstimulation of these neurons [40]. As a highly immunogenic tumor, the interrelationship between immune cells in melanoma has an essential impact on antitumor immune effects [41]. Therefore, studying the cellular composition of the tumor microenvironment and its dynamic changes is crucial for melanoma progression.

    Then, we predicted potential ligand-receptor pairs between immune cells and found that the core genes screened by machine learning are mainly involved in the CXCL pathway during cellular communication. Analysis of the pathway's receptor-ligand pairs revealed that CXCL13-CXCR5 receptor-ligand pairs are only present in cellular communication between T and B cells. It was reported that CXCR5 is mainly conveyed in mature B cells and follicular helper T cells and that the CXCL13/CXCR5 axis coordinates cell-cell interactions and regulates lymphocyte infiltration in the tumor microenvironment [42]. Altered chemokine receptor-ligand expression of CXCR5/CXCL13 has been reported to be relevant in establishing B-cell dysfunction during HIV-1 infection [43]. The present study hypothesized that altered chemokine receptor-ligand expression of CXCR5/CXCL13 influences the immune infiltration of T lymphocytes during melanoma formation, which is one of the highlights of this study and requires further verification by subsequent studies. The present study also has some limitations. First, although we provided a prognosis model to predict melanoma survival, more prospective studies are needed to confirm the model's reliability. We also need more experimental evidence to prove our conclusions and elucidate the exact mechanisms of hub genes in melanoma development.

    In this study, a novel ICRS was built to effectively predict the immune subtypes of SKCM patients. A machine learning algorithm further screened five hub genes that may be involved in regulating the immune microenvironment. In addition, on account of the analysis of scRNA-seq data from melanoma samples, this study described the cell clusters in melanoma, leading to a deeper understanding of underlying intercellular interactions in the immune microenvironment of melanoma, and identified ligand-receptor genes that determine intercellular interactions, suggesting that they may be underlying therapeutic targets affecting the prognosis of melanoma.

    This study was supported by the Guangdong Provincial Science and Technology Innovation Strategy Special Fund (Climbing Plan Special Fund) Project (grant No. pdjh2020a0300), Key Laboratory of Pharmacovigilance Technology Research and Evaluation of State Drug Administration and Innovation Team of Health Economics and Social Security Research.

    The authors declare that there is no conflict of interest.



    [1] Harris S (2013) Climatic change: Casual correlations over the last 240 Ma. Sci Cold Arid Reg 5: 259–274. doi: 10.3724/SP.J.1226.2013.00259
    [2] Imbrie J, Imbrie J (1980) Modeling climatic response to orbital variations. Science 4: 943–953.
    [3] Jin H, Cheng X, Lou D, et al. (2016) Evolution of permafrost in Northeast China since the Late Pleistocene. Sci Cold Arid Reg 8: 269–296.
    [4] Harris S, Jin, He R, et al. (2018) Tessellons, topography, and glaciations on the Qinghai-Tibet Plateau. Sci Cold Arid Reg 10: 187–206.
    [5] Strahler A (1969) Physical Geography, 3Eds., New York: J. Wiley and Sons.
    [6] Wallen R (1992) Introduction to Physical Geography. Dubuque: Wm. C. Brown Publishers.
    [7] Rossby CG, et al. (1939) Relation between variations in the intensity of the zonal circulation of the atmosphere and the displacements of the semi-permanent centers of action. J Mar Res 2: 38–55. doi: 10.1357/002224039806649023
    [8] Harris S (2010) Climatic change in Western North America during the last 15,000 years: The role of changes in the strengths of air masses in producing the changing climates. Sci Cold Arid Reg 2: 371–383.
    [9] Harris S, Brouchkov A, Cheng G (2017) Geocryology. Characteristics and use of Frozen Ground and Permafrost Landforms, Baton Rouge, Florida: CRC Press.
    [10] Held I (2001) The portioning the tropical of the poleward energy transport between the tropical ocean and atmosphere. J Atmos Sci 58: 943–948. doi: 10.1175/1520-0469(2001)058<0943:TPOTPE>2.0.CO;2
    [11] Farneti R, Vallis G (2013) Meridional energy transport in the coupled atmosphere-ocean system: Compensation and partitioning. J Climate 26: 7151–7165. doi: 10.1175/JCLI-D-12-00133.1
    [12] Kang S, Shin Y, Codron F (2018) The partitioning of poleward energy transport response between the atmosphere and Ekman flux to prescribed surface forcing in a simplified GCM. Geosci Lett 5: 22. doi: 10.1186/s40562-018-0124-9
    [13] Wu CY (1985) Flora Xizancica. Science Press, Volume 5, In Chinese.
    [14] Charkevicz S (1985) Plantae Vasculares Orienttis Sovietici. Leningrad: Nauka, 8 volumes, In Russian.
    [15] Krasnoborov I (1988) Flora Sibirae. Novosibirsk: Nauka. 13 volumes. In Russian.
    [16] Breckle SW, Hedge I, Rafiqpoor M, et al. (2013) Vascular Plants of Afghanistan. Bonn: Scientia Bonnensis.
    [17] Harris S (1982) Cold air drainage west of Fort Nelson, British Columbia. Arctic 35: 537–541.
    [18] NSIDC, New study explains Antarctica's coldest temperatures. 2018. Available from: https://nsidc.org/news/newsroom/new-study-explains-antarctica-coldest-temperatures.
    [19] Arnfield A (2003) Two decades of urban climate research: A review of turbulence, exchanges of energy and water and the heat island. Int J Climatol 23: 1–26. doi: 10.1002/joc.859
    [20] Østrem G (1966) The height of the glaciation limit in southern British Columbia and Alberta. Geogr Ann A 48: 126–138. doi: 10.1080/04353676.1966.11879734
    [21] Sherzer W (1913) Glacial history of the Huron-Erie Basin: Geological report on Wayne County.In: Michigan Geological and Biological Survey, Lansing, Michigan: Wyakoop, Hallenbeck, Crawford Co., State printers Publishers.
    [22] Shackleton N, Hall M, Pate D (1995) Pliocene stable isotope stratigraphy of site 846. Proc Ocean Drill Program: Sci Results 138: 337–353.
    [23] Harris S (1994) Chronostratigraphy of Glaciations and Permafrost episodes in the Cordillera of North America. Prog Phys Geog 18: 366–395. doi: 10.1177/030913339401800305
    [24] Zheng B, Shen Y, Jiao K, et al. (2014) New progress and problems of Quaternary moraine dating in the Tibetan Plateau. Sci Cold Arid Reg 6: 183–189.
    [25] Zhang H, Chang F, Li H, et al. (2018) OSL and AMS 14C age of the most complete mammoth fossil skeleton from northeastern China and itspaleoclimate significance. Radiocarbon 61: 347–358.
    [26] Vandenburghe J, Wang X, Vandenburghe D (2016) Very large cryoturbation structures of the last permafrost maximum age at the foot of the Qilian Mountains (NE Tibet Plateau, China). Permafrost Periglac 27: 138–143. doi: 10.1002/ppp.1847
    [27] Harris S, Jin H, He R (2017) Very Large cryoturbation structures of last permafrost maximum age at the foot of Qilian Mountains (NE Tibet Plateau, China): A discussion. Permafrost Periglac 28: 757–762. doi: 10.1002/ppp.1942
    [28] Voris H (2000) Maps of Pleistocene sea levels in Southeast Asia: Shorelines, river systems and time durations. J Biogeogr 27: 1153–1167. doi: 10.1046/j.1365-2699.2000.00489.x
    [29] Harris S, Jin H (2012) Tessellons and "sand wedges" on the Qinghai-Tibet Plateau and their palaeoenvironmental implications. Proceedings of the 10th International Conference on Permafrost, Salekhard, Russia, 1: 147–153.
    [30] Yu L, Lai Z (2012) OSL Chronology and palaeoclimatic implications of aeolian sediments in the eastern Qaidam Basin of the northeastern Qinhai-Tibetan Plateau. Palaeogeogr Palaeocl 337–338: 120–129.
    [31] Liu B, Jin H, Sun L, et al. (2013) Holocene climatic change revealed by aeolian deposits from the Gonghe Basin, northeastern Qinghai-Tibetan Plateau. Quatern Int 296: 231–240. doi: 10.1016/j.quaint.2012.05.003
    [32] Yang S, Jin H (2011) δ18O and δD records of inactive ice-wedges in Yitulihe, Northeast China and their paleoclimatic implications. Sci China Earth Sci 54: 119–126. doi: 10.1007/s11430-010-4029-5
    [33] Yang S, Cao X, Jin H (2015) Validation of wedge ice isotopes at Yituli'he, northeastern China as climate proxy. Boreas 44: 502–510. doi: 10.1111/bor.12121
    [34] Tarasov P, Bezrukova E, Karabanov E, et al. (2007) Vegetation and climate dynamics during the Holocene and Eemian interglacials derived from Lake Baikal pollen records. Palaeogeogr Palaeocl 252: 440–457. doi: 10.1016/j.palaeo.2007.05.002
    [35] Tarasov P, Bezrukova E, Krivonogov S (2009) Late Glacial and Holocene changes in vegetation cover and climate in Southern Siberia derived from a 15 kyr long pollen record from Lake Kotokel. Clim Past 5: 285–295. doi: 10.5194/cp-5-285-2009
    [36] Murton J, Edwards M, Lozhkin A, et al. (2017) Preliminary paleoenvironmental analysis of permafrost deposits at Batagnika megaslump, Yana Uplands, northeast Siberia. Quaternary Res 87: 314–330. doi: 10.1017/qua.2016.15
    [37] Lozhkin A, Anderson P (2018) Another perspective on the age and origin of the Berelyokh Mammoth site (northeast Siberia). Quaternary Res 9: 1–19.
    [38] Clague J, Curry B, Dreimanis A, et al. (1993) Initiation and development of the Laurentide and Cordilleran Ice Sheets following the last interglaciation. Quat Sci Rev 12: 79–114. doi: 10.1016/0277-3791(93)90011-A
    [39] Prest V (1990) Laurentide ice-flow patterns: A historical review, and implications of the dispersal of Belcher Island erratics. Geogr phys Quatern 44: 113–136.
    [40] Lemke R, Laird W, Tipton M, et al. (1965) Quaternary geology of the Northern Great Plains, In: Wright H.E., Jr., Frey, D.G. Editors, The Quaternary of the United States, Princeton: Princeton University Press, 15–27.
    [41] Koerner R (2010) Glaciers of the Hugh Arctic Islands, In: Williams, R.S., Jr., Ferriguo, J.G. Editors, Satellite Image Atlas of Glaciers of The World, United States Geological Survey Professional Paper, 1486–J–1, J111–J–146.
    [42] Gooding A (1963) Illinoian and Wisconsin Glaciations in the Whitewater Basin, Southeast Indiana, and adjacent areas. J Geo 71: 665–682. doi: 10.1086/626948
    [43] Frye J, Willman H, Black R (1965) Outline of glacial geology of Illinois and Wisconsin. In: Wright H.E., Jr., Frey, D.G. Editors, The Quaternary of the United States, Princeton: Princeton University Press, 43–61.
    [44] Wayne W, Zumberge J (1965) Pleistocene geology of Indiana and Michigan, In: Wright H.E., Jr., Frey, D.G. Editors, The Quaternary of the United States, Princeton: Princeton University Press, 63–83.
    [45] Goldthwait R, Dreimanis A, Forsyth J, et al. (1965) Pleistocene deposits of the Erie Lobe, In: Wright, H.E., Jr., Frey, D.G. Editors, The Quaternary of the United States, Princeton: Princeton University Press, 85–97.
    [46] Teller, JT, Fenton, MM (1980) Late Wisconsin Glacial stratigraphy and history of southeastern Manitoba. Can J Earth Sci 17: 19–35. doi: 10.1139/e80-002
    [47] Christiansen E (1992) Pleistocene stratigraphy of the Saskatoon area, Saskatchewan, Canada: An update. Can J Earth Sci 29: 1767–1778. doi: 10.1139/e92-139
    [48] SkwaraWoolf T (1980) Mammals of the Riddell Local Fauna (Floral Formation, Pleistocene, Late Rancholabrean) Saskatoon, Canada. Saskatoon: Culture and Youth Museum of Natural History, Regina.
    [49] Dyke A (2004) An outline of North American deglaciation with emphasis on Central and northern Canada. In: Ehlers, J., Gibbard P.L. Editors, Quaternary glaciations–Extent and Chronology, Part II. Amsterdam: Elservier Science and Technology Books.
    [50] Barendregt R, Irving E (1998) Changes in the extent of North American ice sheets during the Late Cenozoic. Can J Earth Sci 35: 504–509. doi: 10.1139/e97-126
    [51] Patton H, Hubbard A, Andreassen K, et al. (2017) Deglaciation of the Eurasian ice sheet complex. Quat Sci Rev 169: 148–172. doi: 10.1016/j.quascirev.2017.05.019
    [52] Monegato G, Ravizzi C (2018) The Late Pleistocene multifold glaciation in the Alps: Update and open questions. Alp Mediterr Quat 31: 225–229.
    [53] Patterson C (1998) Laurentide glacial landscapes: The role of ice streams. Geology 26: 643–646. doi: 10.1130/0091-7613(1998)026<0643:LGLTRO>2.3.CO;2
    [54] Dredge L, Thorleifson L (1987) The Middle Wisconsin history of the Laurentide ice sheet. Geogr phys Quatern 41: 215–235.
    [55] Liverman D, Catto N, Rutter N (1989) Laurentide glaciation in west-central Alberta: A single (Late Wisconsinan) event. Can J Earth Sci 26: 266–274. doi: 10.1139/e89-022
    [56] Young R, Burns J, Smith D, et al. (1994) A single, late Wisconsin Laurentide glaciation, Edmonton area and southwestern Alberta. Geology 22: 683–686. doi: 10.1130/0091-7613(1994)022<0683:ASLWLG>2.3.CO;2
    [57] Jackson L, Little E (2004) A single continental glaciation of Rocky Mountain Foothills, south-western Alberta, Canada. Dev Quatern Sci 2: 29–38.
    [58] Andriashek L, Barendregt R (2016) Evidence for Early Pleistocene glaciation from borehole stratigraphy in north-central Alberta, Canada. Can J Earth Sci 54: 445–460.
    [59] Marshall S, Clarke K, Dyke A, et al. (1996) Geologic and topographic controls on fast flow in the Laurentide and Cordilleran ice sheets. J Geophys Res 101: 17827–17839. doi: 10.1029/96JB01180
    [60] Margold M, Stokes C, Clark C (2015) Ice streams in the Laurentide Ice Sheet: Identification, characteristics and comparison to modern ice sheets. Earth-Sci Rev 143: 117–146. doi: 10.1016/j.earscirev.2015.01.011
    [61] Henderson E (1959) Surficial geology of Sturgeon Lake map-area, Alberta. Geol Surv Can Memoir 303.
    [62] Tokarsky O (1967) Geology and groundwater resources (Quaternary) of the Grimshaw area, Alberta (Canada). Unpublished M.Sc. thesis, University of Alberta, Edmonton.
    [63] St. Onge D (1972) La stratigraphie du quaternaire des environs de Fort-Assiniboine, Alberta, Canada. Rev Géogr Montreal 26: 153–163.
    [64] Alley N (1973) Glacial stratigraphy and limits of the Rocky Mountain and Laurentide ice sheets in southwestern Alberta, Canada. B Can Petrol Geol 21: 153–177.
    [65] Alley N, Harris S (1974) Pleistocene Glacial Lake sequence in the Foothills, southwestern Alberta, Canada. Can J Earth Sci 11: 1220–1235. doi: 10.1139/e74-115
    [66] Stalker A (1976) Quaternary stratigraphy of the southwestern Canadian Prairies. In: Mahoney, W.C. Editor, Quaternary Stratigraphy of North America, Strodsberg, Pennsylvania, 381–407.
    [67] Mathews W (1978) Quaternary stratigraphy and geomorphology of Charlie Lake (94A) map area, British Columbia. Geol Surv Can 76–20.
    [68] Duk-Rodkin A, Barendregt R, Tarnocai C, et al. (1995) Late Tertiary to late Quaternary record in the Mackenzie Mountains, Northwest Territories, Canada: Stratigraphy, paleosols, paleomagnetism, and chlorine-36. Can J Earth Sci 33: 875–895.
    [69] Bednarski J, Smith T (2007) Laurentide and montane glaciation along the Rocky Mountain Foothills of northeastern British Columbia. Can J Earth Sci 44: 445–457. doi: 10.1139/e06-095
    [70] Clayton L (1967) Stagnant glacier features of the Missouri Coteau. In: Clayton L., Freers T. Editors, Glacial Geology of the Missouri Coteau and adjacent areas. North Dakota Geological Survey Miscellaneous Papers, 30: 25–46.
    [71] Prest V, Grant D, Rampton V (1968) Glacial Map of Canada. Geological Survey of Canada Map 1253A.
    [72] Ryder J, Fulton R, Clague J (1991) The Cordilleran Ice Sheet and the glacial geomorphology of southern and central British Columbia. Geogr phys Quatern 45: 356–377.
    [73] Fulton R (1991) A conceptual model for growth and decay of the Cordilleran Ice Sheet. Geogr phys Quatern 45: 281–286.
    [74] Nichol C, Monahan P, Fulton R, et al. (2015) Quaternary stratigraphy and evidence for multiple glacial episodes in the north Okanagan valley, British Columbia. Can J Earth Sci 52: 338–356. doi: 10.1139/cjes-2014-0182
    [75] Jackson L, Clague, J (1991) The Cordilleran Ice Sheet: One hundred and fifty years of exploration and discovery. Geogr phys Quatern 45: 269–280.
    [76] Harris S (1985) Evidence for the nature of the early Holocene climate and paleogeography, High Plains, Alberta, Canada. Arct Alp Res 17: 49–67. doi: 10.2307/1550961
    [77] Jackson L (1977) Quaternary stratigraphy and terrain inventory of the Alberta portion of the Kananaskis Lakes 1:250,000 sheet (82-J), Unpublished Ph.D. thesis, University of Calgary, Calgary.
    [78] Jackson L (1980) Glacial history and stratigraphy of the Alberta portion of the Kananaskis Lake map area. Can J Earth Sci 17: 459–477. doi: 10.1139/e80-043
    [79] Borns H (1973) Late Wisconsin fluctuations of the Laurentide ice sheet in southern and eastern New England. Geol Soc Am Bull 139: 37–45.
    [80] Mulligan R, Bajc A (2018) The pre-Late Wisconsin stratigraphy of southern Simcoe County, Ontario: Implications for ice sheet buildup, decay and Great Lakes drainage evolution. Can J Earth Sci 55: 709–729. doi: 10.1139/cjes-2016-0160
    [81] Christiansen E (1979) The Wiscosin deglaciation of southern Saskatchewan and adjacent areas. Can J Earth Sci 16: 913–938. doi: 10.1139/e79-079
    [82] Upham W (1895)The Glacial Lake Agassiz; Monographs of the United States Geological Survey, Volume XXV, Washington.D.C.: Government Printing Office.
    [83] Teller J, Clayton L (1983) Glacial Lake Agassiz. Geological Association of Canada, St. Johns, Newfoundland.
    [84] Teller J (1990) Volume and routing of late-glacial runoff from the southern Laurentide Ice Sheet. Quaternary Res 34: 12–23. doi: 10.1016/0033-5894(90)90069-W
    [85] Fisher, T, Smith D, Andrews J (2002) Preboreal oscillation caused by a glacial Lake Agassiz flood. Quat Sci Rev 21: 873–878. doi: 10.1016/S0277-3791(01)00148-2
    [86] Teller J, Leverington D (2004) Glacial Lake Agassiz: A 5000 yr history of change and its relationship to the δ18O record of Greenland. GSA Bull 116: 729–742. doi: 10.1130/B25316.1
    [87] Churcher C (1968) Pleistocene ungulates from the Bow River gravels at Cochrane, Alberta. Can J Earth Sci 5: 1467–1488. doi: 10.1139/e68-145
    [88] Churcher C (1975) Additional evidence of Pleistocene ungulates from the Bow River gravels at Cochrane, Alberta. Can J Earth Sci 12: 68–76. doi: 10.1139/e75-008
    [89] Wilson M, Churcher C (1978) Late Pleistocene Camelops from the Gallelli Pit, Calgary, Alberta: Morphology and geologic setting. Can J Earth Sci 15: 729–740. doi: 10.1139/e78-080
    [90] Ritchie J (1976) The Late Quaternary vegetational history of the western interior of Canada. Can J Bot 54: 1793–1818. doi: 10.1139/b76-194
    [91] MacDonald G (1982) Late Quaternary paleoenvironments of the Morley Flats and Kananaskis Valley of southwestern Alberta. Can J Earth Sci 19: 23–35. doi: 10.1139/e82-003
    [92] Gryba EM (1983) Sibbald Creek: 11,000 years of Human Use of the Alberta Foothills. Archaeological Survey of Alberta, Occasional Paper #22.
    [93] Luckman B (1988) 8200-year-old-wood from the Athabasca Glacier, Alberta. Can J Earth Sci 14: 1809–1822.
    [94] Osborn G, Luckman B (1988) Holocene glacier fluctuations in the Canadian (Alberta and British Columbia). Quat Sci Rev 7: 115–128. doi: 10.1016/0277-3791(88)90002-9
    [95] Harris S, Howell J (1977) Chateau Lake Loiuse moraines–evidence for a new Holocene glacial event in southwestern Alberta. B Can Petrol Geol 25: 441–455.
    [96] Zoltai S, Tarnocai C, Pettapiece W (1978) Age of cryoturbated organic materials in earth hummocks from the Canadian Arctic. In: Proceedings of the 3rd International Conference on Permafrost, Edmonton, Alberta. Ottawa, National Research Council of Canada: 326–331.
    [97] Harris S (2002) Biodiversity of the vascular timberline flora in the Rocky Mountains of Alberta, Canada. In: Koerner, C., Spehn, E. Editors, Mountain Biodiversity: A global assessment, Lancashire: Parthenon Publishing group, United Kingdom, 49–57.
    [98] Fontanella F, Feldman C, Siddall M, et al. (2008) Phylogeography of Diadophis puntatus: Extension lineage diversity and repeated patterns of historical demography of a trans-continental snake. Mol Phylogenet Evol 46: 1049–1070. doi: 10.1016/j.ympev.2007.10.017
    [99] Harris S (2012) The role that diastrophism and climatic change have played in determining biodiversity in continental North America. In: Lameed, A. Editor, Biodiversity, Conservation and Utilization in a Diverse World, Intech Press, 233–260.
    [100] Zagura S (1984) The initial peopling of the Americas: an overview from the perspective of physical anthropology. Acta Anthropog 8: 1–21.
    [101] Kitchen A, Miyamoto M, Mulligan C (2008) A three stage colonization model for the peopling of the Americas. PLOS ONE 3: e1596. Available from: Doi.org/10.1371/journal.pone.0001596. doi: 10.1371/journal.pone.0001596
    [102] Fagan B (2016) Searching for the origins of the first Americans. Sapiens.
    [103] Anon (2018) Suspected first trace of Beringian people on the land bridge – now mostly sunken – joining Russia and North America. The Siberian Times.
    [104] Williams R, Steinberg A, Gershowitz H, et al. (1985) GM allotypes in Native Americans: Evidence for three distinct migrations across the Bering land bridge. Am J Phys Anthropol 66: 1–19. doi: 10.1002/ajpa.1330660102
    [105] Goebel T, Waters M, Dikova M (2003) The archaeology of Ushki Lake, Kamchatka, the Pleistocene peopling of the Americas. Science 301: 501–505. doi: 10.1126/science.1086555
    [106] Elias S, Crocker B (2008) The Bering Land Bridge: A moisture barrier to the dispersal of steppe-tundra biota? Quat Sci Rev 27: 2473–2483. doi: 10.1016/j.quascirev.2008.09.011
    [107] Murton J, Goslar T, Edwards M, et al. (2015) Palaeoenvironmental interpretation of Yedoma Silt (Ice Complex) deposition as cold-climate loess, Duvenny Yar, Northeast Siberia. Permafrost and Periglac 26: 208–288. doi: 10.1002/ppp.1843
    [108] Madsen D, Perreault C, Rhode D, et al. (2017) Early foraging settlement of the Tibetan Plateau Highlands. Archaeol Res Asia.
    [109] Sanchez G, Holliday V, Gaines E, et al. (2014) Human (Clovis)–gomphothere (Cuvieronius sp.) association ∼13,390 calibrated yBP in Sonora, Mexico. PNAS 111: 10972–10977.
    [110] Ferring C (2001) The Archaeology and Paleoecology of the Aubrey Clovis Site (41DN479), Denton County, Texas. U.S. Army Corps of Engineers, Fort Worth District. Center for Environmental Archaeology, Department of Geography, University of North Texas.
    [111] Goebel T, Waters M, O'Rourke D (2008) The late Pleistocene dispersal of modern humans in the Americas. Science 319: 1497–1502. doi: 10.1126/science.1153569
  • This article has been cited by:

    1. Ji-Sol Jeong, Hyo-Sang Choi, Interruption Characteristics of Transformer-Type Superconducting DC Circuit Breaker According to Transformer Turn-Ratio, 2022, 32, 1051-8223, 1, 10.1109/TASC.2022.3160967
    2. Muhammad Aurangzeb, Ai Xin, Sheeraz Iqbal, Flah Aymen, Michał Jasiński, Laura Jasińska, Utilizing Parallel Superconducting Element as a Novel Approach of Flux-Coupled Type SFCL to Limit DC Current in the System, 2022, 11, 2079-9292, 3785, 10.3390/electronics11223785
    3. Muhammad Aurangzeb, Ai Xin, Sheeraz Iqbal, Irfan Jamil, Aymen Flah, Ambe Harrison, Mohammed Alenezi, Mokhtar Shouran, A novel hardware implementation approach to enhanced stable island operation of hybrid distributed generation using superconducting fault current limiters, 2024, 12, 23524847, 1044, 10.1016/j.egyr.2024.06.060
  • Reader Comments
  • © 2019 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(6116) PDF downloads(796) Cited by(5)

Figures and Tables

Figures(9)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog