搜索结果：Tree

Understanding whether long-term vascular cambium vitality in ancient trees reflects progressive decline or adaptive reprogramming is central to grasping woody plant longevity. We performed an integrative analysis of functional traits, transcriptome profiles, and metabolomic landscapes of cambial zone enriched from 80-, 500-, and 1000-year-old Styphnolobium japonicum trees. With increasing tree age, the vascular cambium showed fewer cell layers, reduced thickness, and lower auxin, gibberellin, and IAA/ABA ratios, whereas bark thickness, malondialdehyde, abscisic acid, jasmonic acid, and salicylic acid contents increased. Transcriptomic and metabolomic analyses revealed that differentially expressed genes and metabolites were primarily enriched in the cell cycle, phytohormone signaling, and phenylpropanoid biosynthesis pathways. Specifically, genes associated with cell division were down-regulated in millennial trees, whereas phenolic acids, flavonoids, and lignin-related metabolites significantly accumulated. Piecewise structural equation modeling suggested associations among tree age, transcription factors, structural genes, metabolites, and cambial functional traits. These results indicate that cambial senescence is not a simple linear decay but a highly coordinated remodeling process, providing crucial evidence for delayed senescence in long-lived woody species.

The taxonomic status of Cipangopaludina chinensis and C. cathayensis remains controversial due to morphological overlap. We combined morphometrics (24 landmarks vs. 200 semilandmarks) with multi-locus phylogenetics (COI, 16S rRNA, H3, 28S rRNA) and haplotype networks on sympatric populations from Liuzhou and GenBank. Morphometrics completely separated C. chinensis_lz from C. cathayensis_lz in shell shape; key diagnostic traits include aperture size, spire height, and body whorl inflation. The high-density semilandmark method outperformed discrete landmarks in dimensionality reduction and fine-scale resolution, capturing subtle apex and lateral whorl variations. No sexual dimorphism occurred in C. cathayensis_lz, with only weak dimorphism in C. chinensis_lz. Molecular data were ambiguous: the species tree and mitochondrial gene tree recovered C. chinensis and C. cathayensis as monophyletic, but the nuclear gene tree showed mixing of C. chinensis_lz/C. cathayensis_lz with C. wisseli, C. chinensis, and C. cathayensis. Haplotype networks revealed haplotypes sharing between C. chinensis_lz and C. cathayensis_lz, yet neither population shared haplotype with GenBank sequences of the two nominal species. Morphology strongly supports C. chinensis_lz and C. cathayensis_lz as distinct species, whereas molecular evidence shows only low-level differentiation and discordant signals between mitochondrial and nuclear markers. We conclude they are likely valid species, yet mitonuclear discordance warrants further genomic investigation. Despite no evidence of microhabitat partitioning, the stable morphological divergence between these sympatric morphospecies supports separate management units to conserve phenotypic diversity and local adaptive potential. This study validates geometric morphometrics as an efficient frontline tool for biodiversity assessment in morphologically diverse yet genetically conserved freshwater snails, and reinforces the need for genome-wide data to resolve species boundaries within the Cipangopaludina complex.

To assess woody plant community functional diversity and its main environmental drivers in the northern Hainan volcanic lava region, a tropical volcanic landscape with marked habitat heterogeneity. Woody plant communities were surveyed in the field, leaf functional traits were measured, and nonparametric tests, correlation analysis, redundancy analysis (RDA), and structural equation modeling (SEM) were used to evaluate functional-diversity patterns and potential direct and indirect driver pathways. Community-weighted mean values of leaf area (LA), leaf dry matter content (LDMC), chlorophyll a/b ratio (Chl a/Chl b), leaf dry weight (LDW), and leaf tissue density (LTD) were higher in trees than in shrubs (p&#xa0;<&#xa0;0.001). Functional richness (FRic), functional dispersion (FDis), and Rao's quadratic entropy (RaoQ) were also higher in trees (p&#xa0;<&#xa0;0.001). Mean annual precipitation (MAP), soil pH, and aggregation index (AI) were the main drivers for the whole community, tree layer, and shrub layer, respectively. SEM suggested that meteorological and environmental factors, anthropogenic-disturbance, and landscape factors may affect functional diversity both directly and indirectly through soil properties. Functional diversity in this volcanic lava ecosystem is shaped by multiple environmental pathways, with soil properties playing an important mediating role. These findings provide evidence for conserving and managing woody plant communities in tropical volcanic landscapes.

In the Neotropical Savanna domain, few equations are available for estimating biomass stocks in Cerrado sensu stricto (CSS), despite the importance of such tools for estimating carbon stocks, understanding ecosystem functioning, and supporting conservation actions. We conducted forest inventories in 40 temporary 1000 m2 plots in southeastern Brazil to estimate total and compartmental biomass stocks and to model biomass using structural and edaphic predictors. Total biomass (TB) included aboveground woody biomass (AGWB), necromass, litter, and belowground biomass (BGB). AGWB was estimated for trees with basal diameter &#x2265; 5 cm using a previously fitted regional equation. Root biomass was sampled using a 1 m3 trench excavated at a single point adjacent to each plot. Necromass was quantified using the line-intersect method along a 50 m transect, considering debris with diameter &#x2265; 3 cm. Litter was sampled using a 0.25 m2 frame placed at the center of each plot. Biomass was modeled on an area basis using a hierarchical approach for TB, total tree biomass (TTB = AGWB + BGB), and AGWB. Mean stocks (Mg ha-1 &#xb1; s.d.) were 45.24 &#xb1; 17.32 (TB), 20.47 &#xb1; 11.26 (AGWB), 18.47 &#xb1; 10.88 (BGB), 5.49 &#xb1; 4.18 (litter), and 0.81 &#xb1; 1.62 (necromass). Models selected using the Akaike Information Criterion (AIC) and validated by repeated k-fold cross-validation achieved r&#x177;y = 0.76, 0.72, and 0.94 and RMSE = 24.40%, 27.06%, and 18.10% for TB, TTB, and AGWB. As the equations were calibrated for CSS under the environmental conditions of the Brazilian semiarid region, their transferability to other Cerrado regions should be considered with caution. The inclusion of soil variables (e.g., Al, Mg, and sand) improved predictions, reducing relative costs and taxonomic dependence, while incorporating nutritional adaptations to the acidic soils characteristic of the biome.

N. cadamba is a fast-growing and economically important tree species with significant industrial and medicinal values. Its spherical capitulum comprises hundreds of sessile florets densely aggregated on a receptacle. How this unique floral structure shapes the mating system of this species remains unknown, as this structure enhances the probabilities of both selfing and outcrossing. Here we clarify this question by investigating a natural population. We randomly sampled 11 half-sib families and 6 progenies per half-sib family and assayed samples with whole-genome resequencing. Population genomic analysis indicated that about 93.09% of 4,876,674 high-quality SNPs were under Hardy-Weinberg equilibrium, with a mean inbreeding coefficient of 0.0614 in the sample. Genome-wide nucleotide diversity (&#x3c0;) was 0.0024 on average. The linkage disequilibria (r2) decayed by half at approximately 2.31 kb between sites across genomes. Principal component analysis showed clear family structure among all samples. Analysis with independent SNPs and MLTR (Program for estimating multilocus outcrossing rate) showed that the multilocus outcrossing rate (tm) and single-locus outcrossing rate (ts) were 1.028 &#xb1; 0.048 and 1.015 &#xb1; 0.060, respectively, with an estimated difference between tm and ts of 0.013 &#xb1; 0.022. Analysis with BORICE (Bayesian outcrossing rate and inbreeding coefficient estimation) showed an outcrossing rate (t) of 0.948 &#xb1; 0.029 and an inbreeding coefficient (F) of 0.012 &#xb1; 0.010. Collectively, these results imply that N. cadamba could be essentially outcrossing, without selfing and biparental inbreeding. The potentially outcrossing system of N. cadamba implies differences from its close relative N. macrophylla (a mixed-mating system) in population structure, genome evolution, and natural geographical distribution. This finding also provided a useful reference for breeding and genetic improvement of this "miracle tree".

Interpretable organization of latent spaces remains a central challenge in generative modeling. Most generative models rely on continuous latent variables, but the learned space often lacks an explicit structure that explains how samples are organized or how semantic variation can be controlled. This paper proposes a hyperbolic prototype-residual autoencoder that organizes generative latent representations using a fixed prototype tree embedded in the Poincar&#xe9; ball. Each encoded sample is assigned to a prototype by hyperbolic distance, and the decoder reconstructs or generates images from a prototype-residual representation. The prototype serves as a semantic anchor, while the residual captures local instance-level variation around the selected prototype. The framework combines prototype semantic learning, MMD-based latent spreading, and structural regularizers to align encoder-decoder behavior with the predefined hyperbolic hierarchy. Prototype semantic learning encourages fixed prototypes to decode into representative visual anchors, while MMD encourages encoded samples to occupy broad regions of the hyperbolic latent space. Experiments on MNIST and CelebA show interpretable coarse-to-fine behavior through radial decoding and prototype decoding, suggesting that fixed hyperbolic prototype trees provide an effective scaffold for prototype-guided image reconstruction and generation.

Flatfoot, also known as pes planus, is a deformity of the foot characterized by a decrease or absence of the medial longitudinal arch, which may lead to postural and locomotion abnormalities. This study presents a comparative analysis of handcrafted feature-based machine learning (LBP with Random Forest, Decision Tree, Logistic Regression) and deep learning models (InceptionResNetV2, ResNet101V2, DenseNet201, DenseNet169, InceptionV3, Xception) for flatfoot classification. Monte Carlo cross-validation was employed with subject-wise splitting. SHAP and Grad-CAM were used for explainability. Random Forest achieved 70% accuracy with LBP features. ResNet101V2-RMSprop achieved 92.86%&#x2009;&#xb1;&#x2009;2.82% mean accuracy and 98.21% best single-run accuracy. Deep features with Decision Tree achieved 97% accuracy, 95% recall, and 100% specificity. SHAP and Grad-CAM confirmed that the model focuses on clinically relevant regions: the medial longitudinal arch and calcaneus for pes planus, and the talus-navicular region for normal feet. The hybrid approach is suitable for clinical screening.

The sustainable management of urban pruning residues is increasingly important for promoting circular bioeconomy practices and reducing green waste disposal. This study evaluated the composting potential of pruning biomass from 18 urban tree species (11 deciduous and 7 coniferous) and assessed the effects of biomass type and feedstock ratio on compost quality. Plant residues were composted with fresh cow manure using two compost ratios: V1 (700 g manure + 100 g plant biomass) and V2 (700 g manure + 700 g plant biomass). Compost quality was evaluated through pH, macroelement (N, P, K, Ca, Mg) and total microelement (Mn, Cu, and Zn) and heavy metal (Pb, Cd, and Ni) concentrations. The composting ratio was the primary factor controlling compost properties. V1 produced alkaline composts (pH 7.96-9.42) with higher nutrient concentrations, whereas V2 generated composts with pH values closer to neutrality (7.39-9.44). Increasing the proportion of plant biomass reduced macronutrient concentrations, with total N ranging from 2.23% in V1 to 0.72% in V2, and potassium decreasing by 50-80% relative to V1. In contrast, micronutrient concentrations increased in V2, reaching 1142.7 mg&#xb7;kg-1 Mn and 481.1 mg&#xb7;kg-1 Zn. Total heavy metal concentrations also increased with greater biomass incorporation, although values generally remained low; the highest Ni concentration (15.77 mg&#xb7;kg-1) was recorded in Paulownia tomentosa compost. Species-specific responses were observed for nutrient release and metal accumulation, highlighting differences between deciduous and coniferous feedstocks. The findings demonstrate that compost quality is primarily determined by the manure-to-biomass ratio, while tree species influence nutrient dynamics and trace element accumulation. Composts with higher manure proportions are suitable for nutrient enrichment and acidic soil amelioration, whereas biomass-rich composts provide enhanced micronutrient contents but require monitoring of heavy metal accumulation.

In applications pertaining to sensor network, the Internet of Things, industrial monitoring, and cyber-physical security, graphs are increasingly being outsourced to clouds, where similarity search should be supported without exposing graph content, query graphs, update contents, or database evolution. Existing privacy-preserving graph similarity schemes mainly target static encrypted databases, and as such struggle to handle insertions, deletions, label updates, and long-running index maintenance. This paper proposes DFB-PPGSQ, a dynamic forward-private and epochal backward-private graph similarity matching scheme that moves branch-based lower-bound filtering into a structured encryption framework. DFB-PPGSQ uses epoch-local feature tokens, per-record occurrence handles, one-time update labels, update buffers, deletion tombstones, and shuffle-based branch-tree re-randomization to preserve pruning efficiency while making same-epoch tombstone, traversal, size, timing, and refresh leakage explicit. We formalize the system model, leakage functions, algorithms, and security interpretation, then implement a reproducible Python prototype with HMAC-SHA256 token generation and multi-profile dynamic sensor-topology workloads. Across five random seeds, DFB-PPGSQ keeps server-side filtering latency close to the static branch-tree baseline (46.60 ms versus 44.72 ms at 4000 graphs), avoids immediate full-rebuild updates (0.091 ms insertion and 0.092 ms label update), and keeps metadata-assisted cross-epoch token linkage below 5.6% attack success after refresh in additional industrial and campus IoT stress workloads. Storage, communication, exact GED refinement, side-channel hardening, and verifiable-result protection are treated as deployment costs and limitations rather than being included in the headline latency.

Background: Machine learning (ML) is increasingly applied in medicine, underscoring the need for transparent and clinically relevant models. In gastrointestinal oncology, most ML studies rely on raw imaging data, which limits clinical adoption due to poor interpretability and the difficulty of collecting high-quality, large-scale video and image datasets in routine practice. Endoscopic ultrasound (EUS) plays a central role in the evaluation of pancreatic cancer; however, structured EUS features remain underused in predictive modeling. Objective: To assess the performance and interpretability of ML models for diagnosing pancreatic ductal adenocarcinoma (PDAC) using routinely collected EUS variables. Methods: We conducted a retrospective multicenter study using data from two Italian hospitals (n = 641) for model training and internal validation and from a third hospital (n = 120) for external validation, collected from 2015 to 2023. Decision trees, random forests, na&#xef;ve Bayes and other classifiers were developed and evaluated. Model performance was assessed in terms of discriminative ability, calibration, and selective prediction. Results: All models demonstrated high discriminative performance (AUC &#x2265; 0.90). Decision trees provided the most favorable balance between interpretability and accuracy (balanced accuracy = 0.87; sensitivity = 0.89). Calibration and selective prediction analyses confirmed the robustness of the models. Conclusions: These findings demonstrate the feasibility of implementing interpretable yet high-performing ML models for PDAC diagnosis in real-life endoscopic settings.

Rotaviruses (RVs) are important pathogens which induce gastroenteritis in different kinds of animals, including mammals and birds. Rotaviruses are divided into nine species (RVA-RVD and RVF-RVJ), and RVA-RVC and RVH can infect both humans and pigs. It is vital to understand the genetic diversity and evolution of porcine rotavirus (PoRV) for effective prevention and control of this disease. In this study, 5320 intestinal tissue samples and fecal swabs were collected from different pig farms in Guangxi Province, southern China, from 2022 to 2025. These samples were tested for PoRV species A (PoRVA), PoRVB, PoRVC, and PoRVH using the multiplex RT-qPCR. The positive samples of PoRVA were further selected to amplify and analyze the VP4, VP6, and VP7 gene sequences. The phylogenetic trees were constructed based on the PoRVA VP4, VP6, and VP7 gene sequences. Bayesian time-dynamic analysis and recombination analysis were performed for the PoRVA VP4 gene. The results indicated that the PoRVA, PoRVB, PoRVC, and PoRVH positivity rates were 16.92% (900/5320), 0.51% (27/5320), 12.71% (676/5320), and 6.22% (331/5320), respectively. Fifty-two VP4, VP6, and VP7 gene sequences were obtained from the 52 selected PoRVA-positive clinical samples. The nucleotide and amino acid identity analysis of the obtained PoRVA VP4, VP6, and VP7 genes indicated that the genetic diversity of the VP4 gene was higher than that of the VP6 and VP7 genes. The phylogenetic trees based on the VP4, VP6, and VP7 genes revealed that the predominant strains of PoRVA in Guangxi Province were the G9P[13]I5 genotype. Bayesian analysis indicated that the population size of PoRVA kept steady with no significant expansion from its discovery in the 1970s to approximately 2016, then exhibited gradual growth. Sequence analysis of the PoRVA VP4 gene revealed substitutions and recombination in the PoRVA strains, and one strain was derived from recombination of a porcine-originating strain and a human-originating strain. This study provided useful information on the molecular characteristics and genetic diversity of PoRVA and supplied important clues for in-depth research on the cross-species transmission of PoRVA.

Understanding crop genetic diversity is essential for conservation and breeding, yet farmer-maintained germplasm remains largely underrepresented in genomic studies. Theobroma cacao L. has a complex domestication history, extensive global diversity, and is currently cultivated in Central America. Costa Rican cacao has been understudied compared to South American and Mexican cultivars despite cultural and historical importance. In this study, we investigate the genetic diversity of cacao from farmer-managed systems across Costa Rica to search for Criollo germplasm while identifying and characterizing unique local genetic groups. Ninety-four trees were sampled for whole genome resequencing from 17 farms across four regions of the country. Farmer materials were analyzed alongside 166 previously characterized reference accessions representing major cacao genetic groups. Population structure analyses, phylogenetic reconstruction, and network approaches revealed that Costa Rican cacao encompasses multiple known genetic groups, including Criollo-derived lineages while also harbouring locally distinct diversity not represented in current global reference collections. Analyses revealed close kinship between many accessions with no clear geographic patterns corresponding to the observed population differentiation, reflecting the influence of farmers in generating the dominant patterns of gene flow through seed-saving, clonal propagation, and sharing of genotypes among farms. Heterozygosity levels varied substantially among individuals, consistent with a mixture of highly inbred Criollo trees and more heterozygous, admixed genotypes. We find that farmer-managed cacao systems are reservoirs of genetic diversity that can include rare or historically important lineages, underscoring the value of these farming systems for effective conservation and management of genomic resources for cacao resilience and improvement.

Persian walnut (Juglans regia L.) is an economically important tree species in the Juglandaceae family. Comparative analysis of its chloroplast genome sequences is of great significance for species identification and evolutionary research. In this study, high-throughput sequencing was performed using the Illumina HiSeq platform, followed by de novo assembly of the chloroplast genome, and then analyses of its repetitive sequences, codon usage bias, nucleotide polymorphism and population structure. The results showed that the chloroplast genome size of walnut ranged from 160,311 bp to 160,367 bp with a GC content of 36.11%. It exhibited a typical quadripartite circular structure, and a total of 78 distinct protein-coding genes (PCGs), 30 tRNA genes and four rRNA genes were annotated. A total of 16 distinct SSR motifs were identified, among which A/T motifs were the most abundant; 31-40 bp long tandem repeats were the most abundant in all samples, which increased genomic variability. Codon usage bias analysis revealed that all walnut samples preferentially use the codon UUA (Leu), and most codons end with A/U. Nucleotide sequence alignment identified 129 polymorphic sites. Single nucleotide variants (SNVs) were the dominant variation type, followed by InDels. Furthermore, the phylogenetic tree, principal component analysis (PCA), haplotype network and population structure classified the 41 walnut samples into three genetic clusters (C1, C2 and C3). This study elucidated the conserved characteristics and evolutionary variations in walnut chloroplast genomes, providing theoretical support for its phylogenetic research and germplasm resource utilization.

Boreal forest ecosystems constitute a large terrestrial reservoir of carbon. In these nitrogen-limited environments, release of nutrients through decomposition of soil organic matter is of fundamental importance. Fungi, particularly saprotrophic Agaricomycetes, are thought to drive this process using lignocellulolytic enzymes to degrade plant litter. However, some ectomycorrhizal fungal lineages have retained ancestral decomposition capabilities, yet evidence of their direct involvement in decomposition under field conditions is scarce. We used metatranscriptomics to examine the involvement of ectomycorrhizal fungi in the production of class II peroxidases in the soil of a Swedish boreal forest. We compared nutrient-poor plots with more fertile ones and related the peroxidase-expressing community to the total and cellulose-degrading fungal communities. We found that overall expression of class II peroxidase genes was upregulated in nutrient-poor soil, with ectomycorrhizal species in the Cortinariaceae family accounting for most of the transcripts. Among cellulose-degrading fungi, there was a shift from saprotrophic Agaricomycetes in nutrient-rich soil to dominance by Ascomycetes under nutrient-poor conditions. Symbiosis may enable ectomycorrhizal fungi to use tree photoassimilates to drive energetically costly oxidation belowground. Ectomycorrhiza-driven oxidation may, thereby, enable trees to indirectly regulate decomposition and nutrient cycling to maintain ecosystem productivity on unfertile soils.

Climate change is increasingly reshaping the distribution and long-term persistence of urban greening tree species, yet national-scale assessments of widely planted species in China remain limited. Salix babylonica is one of the most representative and extensively planted urban greening trees in China, but its future climate vulnerability and redistribution patterns are still poorly understood. In this study, we integrated 425 occurrence records and 16 environmental variables to project the potential distribution of S. babylonica under current and future (2050s, 2070s, and 2090s) climate scenarios across three shared socioeconomic pathways (SSP126, SSP370, and SSP585) using an optimized MaxEnt model. The optimized model demonstrated strong predictive performance and identified annual precipitation (&#x2265;309.64 mm), precipitation of the wettest month (&#x2265;82.01 mm), elevation (&#x2264;2251.24 m), mean temperature of the coldest quarter (&#x2265;-10.77 &#xb0;C), annual mean temperature (&#x2265;3.94 &#xb0;C), and minimum temperature of the coldest month (&#x2265;-19.11 &#xb0;C) as the dominant environmental factors (threshold) constraining species distribution. Under the current climate, suitable habitats of S. babylonica are primarily distributed in East, Central, Southwest, and South China, with highly suitable habitats concentrated in the eastern and central plains. Under future climate scenarios, high suitability areas are projected to contract significantly and become progressively fragmented, with reductions ranging from 2.81% (2070s-SSP126) to 40.01% (2090s-SSP585), whereas medium and low suitability areas are projected to expand substantially, with increases ranging from 6.76% to 48.10%. The overall centroid of suitable habitats exhibits a consistent northward shift across scenarios. These results indicate that climate change will not simply expand the potential distribution of S. babylonica, but will reorganize habitat quality and increase long-term risks for its continued use in urban greening. Our findings provide a spatially explicit basis for climate-adaptive urban forestry planning and support risk-differentiated management, including the conservation of stable core areas, monitoring of declining risk zones, and cautious introduction trials in newly emerging suitable areas.