Divergent contributions of coding and noncoding sequences to initial high-altitude adaptation in passerine birds endemic to the Qinghai-Tibet Plateau.

Early events in the evolution of an ancestral lineage can shape the adaptive patterns of descendant species, but the evolutionary mechanisms driving initial adaptation from an ancestor remain largely unexplored. High-altitude adaptations have been extensively explored from the viewpoint of protein-coding genes; however, the contribution of noncoding regions remains relatively neglected. Here, we integrate genomic and transcriptomic data to investigate adaptive evolution in the ancestor of three high-altitude snowfinch species endemic to the Qinghai-Tibet Plateau. Our genome-wide scan for adaptation in the snowfinch ancestor identifies strong adaptation signals in functions of development and metabolism for the coding genes, but in functions of the nervous system development for noncoding regions. This pattern is exclusive to the snowfinch ancestor compared to a control ancestral lineage subject to weak selection. Changes in noncoding regions in the snowfinch ancestor, especially those nearest to coding genes, may be disproportionately associated with the differential expression of genes in the brain tissue compared to other tissues. Extensive gene expression in the brain tissue can be further altered via genetic regulatory networks of transcription factors harbouring potential accelerated regulatory regions (e.g., the development-related transcription factor YEATS4). Altogether, our study provides new evidence concerning how coding and noncoding sequences work through decoupled pathways in initial adaptation to the selective pressure of high-altitude environments. The analysis highlights the idea that noncoding sequences may be promising elements in facilitating the rapid evolution and adaptation to high altitudes.

Genomic Landscape of Long Terminal Repeat Retrotransposons (LTR-RTs) and Solo LTRs as Shaped by Ectopic Recombination in Chicken and Zebra Finch.

Transposable elements (TEs) are nearly ubiquitous among eukaryotic genomes, but TE contents vary dramatically among phylogenetic lineages. Several mechanisms have been proposed as drivers of TE dynamics in genomes, including the fixation/loss of a particular TE insertion by selection or drift as well as structural changes in the genome due to mutation (e.g., recombination). In particular, recombination can have a significant and directional effect on the genomic TE landscape. For example, ectopic recombination removes internal regions of long terminal repeat retrotransposons (LTR-RTs) as well as one long terminal repeat (LTR), resulting in a solo LTR. In this study, we focus on the intra-species dynamics of LTR-RTs and solo LTRs in bird genomes. The distribution of LTR-RTs and solo LTRs in birds is intriguing because avian recombination rates vary widely within a given genome. We used published linkage maps and whole genome assemblies to study the relationship between recombination rates and LTR-removal events in the chicken and zebra finch. We hypothesized that regions with low recombination rates would harbor more full-length LTR-RTs (and fewer solo LTRs) than regions with high recombination rates. We tested this hypothesis by comparing the ratio of full-length LTR-RTs and solo LTRs across chromosomes, across non-overlapping megabase windows, and across physical features (i.e., centromeres and telomeres). The chicken data statistically supported the hypothesis that recombination rates are inversely correlated with the ratio of full-length to solo LTRs at both the chromosome level and in 1-Mb non-overlapping windows. We also found that the ratio of full-length to solo LTRs near chicken telomeres was significantly lower than those ratios near centromeres. Our results suggest a potential role of ectopic recombination in shaping the chicken LTR-RT genomic landscape.

Metabolic and microbial changes in light-vented bulbul during recent northward range expansion.

Endotherms recently expanding to cold environments generally exhibit strong physiological acclimation to sustain high body temperature. During this process, gut microbes likely play a considerable role in host physiological functions, including digestion and thermogenesis. The light-vented bulbul Pycnonotus sinensis represents one such species. It used to be restricted to the Oriental realm but expanded its distribution range north to the Palearctic areas during the past few decades. Here, we explored the seasonal dynamics of the resting metabolic rate (RMR) and microbiota for local and newly colonized populations of the species. Our results showed that the mass-adjusted RMR and body mass were positively correlated with latitude variations in both seasons. Consistently, the gut microbiota showed a corresponding variation to the northern cold environments. In the two northern populations, the alpha diversity decreased compared with those of the two southern populations. Significant differences were detected in dominant phyla, such as Firmicutes, Bacteroidetes, Proteobacteria, and Desulfobacterota in both seasons. The core microbiota showed geographic differences in the winter, including the elevated relative abundance of 5 species in northern populations. Finally, to explore the link between microbial communities and host metabolic thermogenesis, we conducted a correlation analysis between microbiota and mass-adjusted RMR. We found that more genera were significantly correlated with mass-adjusted RMR in the wintering season compared to the breeding season (71 vs. 23). These results suggest that microbiota of the lighted-vented bulbul linked with thermogenesis in diversity and abundance under northward expansion.

Bird Object Detection: Dataset Construction, Model Performance Evaluation, and Model Lightweighting.

The application of object detection technology has a positive auxiliary role in advancing the intelligence of bird recognition and enhancing the convenience of bird field surveys. However, challenges arise due to the absence of dedicated bird datasets and evaluation benchmarks. To address this, we have not only constructed the largest known bird object detection dataset, but also compared the performances of eight mainstream detection models on bird object detection tasks and proposed feasible approaches for model lightweighting in bird object detection. Our constructed bird detection dataset of GBDD1433-2023, includes 1433 globally common bird species and 148,000 manually annotated bird images. Based on this dataset, two-stage detection models like Faster R-CNN and Cascade R-CNN demonstrated superior performances, achieving a Mean Average Precision (mAP) of 73.7% compared to one-stage models. In addition, compared to one-stage object detection models, two-stage object detection models have a stronger robustness to variations in foreground image scaling and background interference in bird images. On bird counting tasks, the accuracy ranged between 60.8% to 77.2% for up to five birds in an image, but this decreased sharply beyond that count, suggesting limitations of object detection models in multi-bird counting tasks. Finally, we proposed an adaptive localization distillation method for one-stage lightweight object detection models that are suitable for offline deployment, which improved the performance of the relevant models. Overall, our work furnishes an enriched dataset and practice guidelines for selecting suitable bird detection models.

Comparison between short-term stress and long-term adaptive responses reveal common paths to molecular adaptation.

The phenotypic plasticity in responses to short-term stress can provide clues for understanding the adaptive fixation mechanism of genetic variation during long-term exposure to extreme environments. However, few studies have compared short-term stress responses with long-term evolutionary patterns; in particular, no interactions between the two processes have been evaluated in high-altitude environment. We performed RNA sequencing in embryo fibroblasts derived from great tits and mice to explore transcriptional responses after exposure to simulated high-altitude environmental stresses. Transcriptional changes of genes associated with metabolic pathways were identified in both bird and mice cells after short-term stress responses. Genomic comparisons among long-term highland tits and mammals and their lowland relatives revealed similar pathways (e.g., metabolic pathways) with that initiated under short-term stress transcriptional responses in vitro. These findings highlight the indicative roles of short-term stress in the long-term adaptation, and adopt common paths to molecular adaptation in mouse and bird cells.

Eyebirds: Enabling the Public to Recognize Water Birds at Hand.

Enabling the public to easily recognize water birds has a positive effect on wetland bird conservation. However, classifying water birds requires advanced ornithological knowledge, which makes it very difficult for the public to recognize water bird species in daily life. To break the knowledge barrier of water bird recognition for the public, we construct a water bird recognition system (Eyebirds) by using deep learning, which is implemented as a smartphone app. Eyebirds consists of three main modules: (1) a water bird image dataset; (2) an attention mechanism-based deep convolution neural network for water bird recognition (AM-CNN); (3) an app for smartphone users. The waterbird image dataset currently covers 48 families, 203 genera and 548 species of water birds worldwide, which is used to train our water bird recognition model. The AM-CNN model employs attention mechanism to enhance the shallow features of bird images for boosting image classification performance. Experimental results on the North American bird dataset (CUB200-2011) show that the AM-CNN model achieves an average classification accuracy of 85%. On our self-built water bird image dataset, the AM-CNN model also works well with classification accuracies of 94.0%, 93.6% and 86.4% at three levels: family, genus and species, respectively. The user-side app is a WeChat applet deployed in smartphones. With the app, users can easily recognize water birds in expeditions, camping, sightseeing, or even daily life. In summary, our system can bring not only fun, but also water bird knowledge to the public, thus inspiring their interests and further promoting their participation in bird ecological conservation.

Orthologous microsatellites, transposable elements, and DNA deletions correlate with generation time and body mass in neoavian birds.

The rate of mutation accumulation in germline cells can be affected by cell replication and/or DNA damage, which are further related to life history traits such as generation time and body mass. Leveraging the existing datasets of 233 neoavian bird species, here, we investigated whether generation time and body mass contribute to the interspecific variation of orthologous microsatellite length, transposable element (TE) length, and deletion length and how these genomic attributes affect genome sizes. In nonpasserines, we found that generation time is correlated to both orthologous microsatellite length and TE length, and body mass is negatively correlated to DNA deletions. These patterns are less pronounced in passerines. In all species, we found that DNA deletions relate to genome size similarly as TE length, suggesting a role of body mass dynamics in genome evolution. Our results indicate that generation time and body mass shape the evolution of genomic attributes in neoavian birds.

Comparative analysis of active retrotransposons in the transcriptomes of three species of heteromyid rodents.

Virtually every eukaryotic genome is replete with transposable elements (TEs). TE activity, or lack thereof, is of considerable evolutionary interest as TE insertions actively drive genome evolution by altering gene expression and/or function through different mechanisms. Herein, we take a comparative approach to better understand the variation in TE transcriptional activity. Our goals were to identify transcriptionally active TEs and to evaluate their relative expression levels in an effort to identify key determinants of TE activity. We do so in three related non-model rodent species (Dipodomys spectabilis, Chaetodipus baileyi, and Heteromys desmarestianus). We used these species (a) because rodents have long been productive models for the study of TEs; (b) because the known and dated relationships among these species permit strong phylogenetic inference (e.g., ancestral character states); and (c) because we have previously characterized key genes that underlie evolutionary adaptations in these species (e.g., osmoregulation). We used RNA-seq to characterize the transcriptomes of two different tissues--kidney and spleen--in each of these three species. Our data revealed a diversity of retrotransposons that were actively transcribed in these rodents (including LINEs, retroviruses/ERVs, and SINEs). We also identified tissue-specific differences in retrotransposon-activity in about half of the retrotransposons detected. We interpret these differences in TE activity in light of overall gene expression, with the transcribed SINEs enriched in differentially expressed sequences in all three species. These transcribed TEs may simply represent transpositionally active TEs. Alternatively, they may be TEs that have been co-opted to represent functionally important components of the host transcriptome (i.e., exaptations). Overall, our results contribute to a growing appreciation for the evolutionary interplay between TEs and their hosts.

The genome sequence of a widespread apex predator, the golden eagle (Aquila chrysaetos).

Biologists routinely use molecular markers to identify conservation units, to quantify genetic connectivity, to estimate population sizes, and to identify targets of selection. Many imperiled eagle populations require such efforts and would benefit from enhanced genomic resources. We sequenced, assembled, and annotated the first eagle genome using DNA from a male golden eagle (Aquila chrysaetos) captured in western North America. We constructed genomic libraries that were sequenced using Illumina technology and assembled the high-quality data to a depth of ∼40x coverage. The genome assembly includes 2,552 scaffolds >10 Kb and 415 scaffolds >1.2 Mb. We annotated 16,571 genes that are involved in myriad biological processes, including such disparate traits as beak formation and color vision. We also identified repetitive regions spanning 92 Mb (∼6% of the assembly), including LINES, SINES, LTR-RTs and DNA transposons. The mitochondrial genome encompasses 17,332 bp and is ∼91% identical to the Mountain Hawk-Eagle (Nisaetus nipalensis). Finally, the data reveal that several anonymous microsatellites commonly used for population studies are embedded within protein-coding genes and thus may not have evolved in a neutral fashion. Because the genome sequence includes ∼800,000 novel polymorphisms, markers can now be chosen based on their proximity to functional genes involved in migration, carnivory, and other biological processes.

Of contigs and quagmires: next-generation sequencing pitfalls associated with transcriptomic studies.

Molecular ecologists have good reasons to be excited about the newest DNA/RNA sequencing technologies. However, this exuberance should be tempered with a hefty dose of reality: new sequencing technologies come with significant new challenges. Herein, we offer a brief overview of some practical problems encountered during transcriptomics studies conducted in our laboratory, and of nontrivial issues that prospective practitioners should consider. These include template contamination (e.g. from xenobiotics) and the cutting-room floor problem, whereby most of the data are often unassembled, unannotated and unused. We also highlight computational requirements, including hardware, personnel time and associated skill sets. We are very optimistic about the future of molecular ecology, but we hope this cautionary overview will help neophytes better recognize some key challenges associated with new technologies.