Applications in Plant Sciences最新文献

Premise: Deep learning-based classification of herbarium images is hampered by background heterogeneity, which introduces noise and artifacts that can potentially mislead models and degrade their accuracy. Addressing these effects is essential to enhance overall performance.

Methods: We introduce PlantSAM, an automated segmentation pipeline that integrates YOLOv10 for plant region detection and the Segment Anything Model (SAM2) for segmentation. YOLOv10 generates bounding box prompts to guide SAM2, enhancing segmentation accuracy. Both models were fine-tuned on herbarium images and evaluated using intersection over union (IoU) and Sørensen-Dice coefficient metrics.

Results: PlantSAM achieved state-of-the-art segmentation performance, with an IoU of 0.94 and a Sørensen-Dice coefficient of 0.97. Incorporating segmented images into classification models led to consistent performance improvements across five tested botanical traits, with accuracy gains of up to 4.36% and F1 score improvements of 4.15%.

Conclusions: Our findings highlight the importance of background removal in herbarium image analysis, as it significantly enhances classification performance by enabling models to focus more effectively on the foreground plant structures.

Premise: Salt marshes in the North Atlantic United States are dominated by grasses from the genus Sporobolus, which are perennial C₄ plants known for tolerating saline, anoxic, and flooded coastal sediments. Establishing in vitro cultures of Sporobolus species remains a major challenge due to frequent seed contamination from ergot and endophytic bacteria in vegetative tissues.

Methods and results: Here, we demonstrate that in vitro axenic cultures of Sporobolus alterniflorus can be established in four weeks and micropropagation chains in three months with a 75% success rate. Additionally, this method is easily transferable to other salt marsh grasses, such as S. cynosuroides.

Conclusions: The strength of our method lies in bypassing adventitious regeneration or somatic embryogenesis. Instead, we remove the hypocotyls and endosperm, primary sources of fungal contamination, and induce adventitious roots in the epicotyl. We also detail culture- and PCR-based methods to screen for contamination in vitroplants of S. alterniflorus.

Premise: Studies comparing seed longevity between species or genotypes commonly measure seed longevity as the time it takes for seed lot viability to drop to 50% (p ₅₀). However, p ₅₀ is influenced by the initial viability. Although standard protocols for comparative studies thus recommend using seed lots with similar and high initial viability (>85%), variation in viability strongly influences p ₅₀.

Methods: We simulated seed viability decline across a range of initial viabilities and rates of probit viability loss (σ) to illustrate how variation in initial viability affects p ₅₀ estimates and to test approaches for reducing this bias.

Results: For hypothetical seed lots with identical rates of seed viability loss, variation in initial viability leads to a threefold variation in p ₅₀ estimates. Narrowing the initial viability (e.g., to 85-95%) reduced this bias. Alternatively, p ₅₀ can be recalculated to a standardized value of initial viability (e.g., 90%), which makes it proportional to the rate of probit viability loss. The most straightforward measure of seed longevity for comparative studies is the probit rate of viability loss itself, represented by σ from the viability equation.

Conclusions: p ₅₀ is confounded by variation in initial seed lot viability and is suboptimal for comparative studies of seed longevity among species or genotypes. Robust measures of seed longevity include the rate of probit viability loss (σ) or p ₅₀ standardized to a certain initial seed viability.

Premise: Due to their small size and lack of easily visible macroscopic characters, the identification of cryptogam species has always been challenging. Here, the use of a machine learning computer vision method is explored for the identification of species of lichens and bryophytes from Australian biocrusts.

Methods: Three models were trained using mostly images from herbarium specimens. The models were then evaluated based on statistics produced by Microsoft Azure Custom Vision and a bench-test with the CSIRO Horama ID mobile app.

Results: Despite the small size and reduced habit of lichens and bryophytes, the Cryptogam (lichens and bryophytes) model performance value is just slightly lower than the performance of a vascular plant model of similar scope (64% accuracy for the Cryptogam model versus 70.3% for vascular plants from Costa Rica).

Discussion: The performance of our models suggested opportunities for improvement, including for bias issues caused by imbalanced datasets, white background, and mixed specimens, as well as the difficulty in stabilizing live images at high magnification when using a mobile device to deploy the model. Further opportunities to improve model performance for these small and character-poor organisms, including data augmentation and image segmentation, are also discussed.

Premise: Extracting high-quality DNA from challenging plant tissues can be hindered by high levels of phenolics, carbohydrates, and other compounds that bind to or coprecipitate with DNA. Bryophytes, due to their small size, tendency to intermix, biochemical richness, and capacity to absorb environmental pollutants and heavy metals, pose unique challenges that further complicate proper DNA isolation.

Methods and results: We compared cetyltrimethylammonium bromide (CTAB)-based extraction protocols and evaluated step-by-step modifications aimed at maximizing DNA yield and purity for Orthotrichaceae, one of the most diverse families of bryophytes. The final protocol was also evaluated on representatives of other bryophyte families as a preliminary exploration of its broader applicability.

Conclusions: We present an optimized, rapid, and efficient DNA extraction protocol that yields high-quality DNA suitable for high-throughput sequencing and molecular analyses. The presented extraction protocol is likely to work well for other bryophyte taxa and may be of value for molecular analyses involving recalcitrant samples.