Bud Growth Stages Research Articles

Lightweight GAN-Assisted Class Imbalance Mitigation for Apple Flower Bud Detection

Multi-class object detectors often suffer from the class imbalance issue, where substantial model performance discrepancies exist between classes. Generative adversarial networks (GANs), an emerging deep learning research topic, are able to learn from existing data distributions and generate similar synthetic data, which might serve as valid training data for improving object detectors. The current study investigated the utility of lightweight unconditional GAN in addressing weak object detector class performance by incorporating synthetic data into real data for model retraining, under an agricultural context. AriAplBud, a multi-growth stage aerial apple flower bud dataset was deployed in the study. A baseline YOLO11n detector was first developed based on training, validation, and test datasets derived from AriAplBud. Six FastGAN models were developed based on dedicated subsets of the same YOLO training and validation datasets for different apple flower bud growth stages. Positive sample rates and average instance number per image of synthetic data generated by each of the FastGAN models were investigated based on 1000 synthetic images and the baseline detector at various confidence thresholds. In total, 13 new YOLO11n detectors were retrained specifically for the two weak growth stages, tip and half-inch green, by including synthetic data in training datasets to increase total instance number to 1000, 2000, 4000, and 8000, respectively, pseudo-labeled by the baseline detector. FastGAN showed its resilience in successfully generating positive samples, despite apple flower bud instances being generally small and randomly distributed in the images. Positive sample rates of the synthetic datasets were negatively correlated with the detector confidence thresholds as expected, which ranged from 0 to 1. Higher overall positive sample rates were observed for the growth stages with higher detector performance. The synthetic images generally contained fewer detector-detectable instances per image than the corresponding real training images. The best achieved YOLO11n AP improvements in the retrained detectors for tip and half-inch green were 30.13% and 14.02% respectively, while the best achieved YOLO11n mAP improvement was 2.83%. However, the relationship between synthetic training instance quantity and detector class performances had yet to be determined. GAN was concluded to be beneficial in retraining object detectors and improving their performances. Further studies are still in need to investigate the influence of synthetic training data quantity and quality on retrained object detector performance.

Accuracy Comparison of YOLOv7 and YOLOv4 Regarding Image Annotation Quality for Apple Flower Bud Classification

Object detection is one of the most promising research topics currently, whose application in agriculture, however, can be challenged by the difficulty of annotating complex and crowded scenes. This study presents a brief performance assessment of YOLOv7, the state-of-the-art object detector, in comparison to YOLOv4 for apple flower bud classification using datasets with artificially manipulated image annotation qualities from 100% to 5%. Seven YOLOv7 models were developed and compared to corresponding YOLOv4 models in terms of average precisions (APs) of four apple flower bud growth stages and mean APs (mAPs). Based on the same test dataset, YOLOv7 outperformed YOLOv4 for all growth stages at all training image annotation quality levels. A 0.80 mAP was achieved by YOLOv7 with 100% training image annotation quality, meanwhile a 0.63 mAP was achieved with only 5% training image annotation quality. YOLOv7 improved YOLOv4 APs by 1.52% to 166.48% and mAPs by 3.43% to 53.45%, depending on the apple flower bud growth stage and training image annotation quality. Fewer training instances were required by YOLOv7 than YOLOv4 to achieve the same levels of classification accuracies. The most YOLOv7 AP increase was observed in the training instance number range of roughly 0 to 2000. It was concluded that YOLOv7 is undoubtedly a superior apple flower bud classifier than YOLOv4, especially when training image annotation quality is suboptimal.

Sensitivity examination of YOLOv4 regarding test image distortion and training dataset attribute for apple flower bud classification

ABSTRACT Applications of convolutional neural network (CNN)-based object detectors in agriculture have been a popular research topic in recent years. However, complicated agricultural environments bring many difficulties for ground truth annotation as well as potential uncertainties for image data quality. Using YOLOv4 as a representation of state-of-the-art object detectors, this study quantified YOLOv4’s sensitivity against artificial image distortions including white noise, motion blur, hue shift, saturation change, and intensity change, and examined the importance of various training dataset attributes based on model classification accuracies, including dataset size, label quality, negative sample presence, image sequence, and image distortion levels. The YOLOv4 model trained and validated on the original datasets failed at 31.91% white noise, 22.05-pixel motion blur, 77.38° hue clockwise shift, 64.81° hue counterclockwise shift, 89.98% saturation decrease, 895.35% saturation increase, 79.80% intensity decrease, and 162.71% intensity increase with 30% mean average precisions (mAPs) for four apple flower bud growth stages. The performance of YOLOv4 decreased with both declining training dataset size and training image label quality. Negative samples and training image sequence did not make a substantial difference in model performance. Incorporating distorted images during training improved the classification accuracies of YOLOv4 models on noisy test datasets by 13 to 390%. In the context of apple flower bud growth-stage classification, except for motion blur, YOLOv4 is sufficiently robust for potential image distortions by white noise, hue shift, saturation change, and intensity change in real life. Training image label quality and training instance number are more important factors than training dataset size. Exposing models to test-image-alike training images is crucial for optimal model classification accuracies. The study enhances understanding of implementing object detectors in agricultural research.

UAV-Based Heating Requirement Determination for Frost Management in Apple Orchard

Frost is a natural disaster that can cause catastrophic damages in agriculture, while traditional temperature monitoring in orchards has disadvantages such as being imprecise and laborious, which can lead to inadequate or wasteful frost protection treatments. In this article, we presented a heating requirement assessment methodology for frost protection in an apple orchard utilizing unmanned aerial vehicle (UAV)-based thermal and RGB cameras. A thermal image stitching algorithm using the BRISK feature was developed for creating georeferenced orchard temperature maps, which attained a sub-centimeter map resolution and a stitching speed of 100 thermal images within 30 s. YOLOv4 classifiers for six apple flower bud growth stages in various network sizes were trained based on 5040 RGB images, and the best model achieved a 71.57% mAP for a test dataset consisted of 360 images. A flower bud mapping algorithm was developed to map classifier detection results into dense growth stage maps utilizing RGB image geoinformation. Heating requirement maps were created using artificial flower bud critical temperatures to simulate orchard heating demands during frost events. The results demonstrated the feasibility of the proposed orchard heating requirement determination methodology, which has the potential to be a critical component of an autonomous, precise frost management system in future studies.

Spring Freeze Damage of Pecan Bloom: A Review

Pecan is native to the United States. The US is the world’s largest pecan producer with an average yearly production of 250 to 300 million pounds; 80 percent of the world’s supply. Georgia, New Mexico, Texas, Arizona, Oklahoma, California, Louisiana, and Florida are the major US pecan producing states. Pecan trees frequently suffer from spring freeze at bud break and bloom as the buds are quite sensitive to freeze damage. This leads to poor flower and nut production. This review focuses on the impact of spring freeze during bud differentiation and flower development. Spring freeze kills the primary terminal buds, the pecan tree has a second chance for growth and flowering through secondary buds. Unfortunately, secondary buds have less bloom potential than primary buds and nut yield is reduced. Spring freeze damage depends on severity of the freeze, bud growth stage, cultivar type and tree age, tree height and tree vigor. This review discusses the impact of temperature on structure and function of male and female reproductive organs. It also summarizes carbohydrate relations as another factor that may play an important role in spring growth and transition of primary and secondary buds to flowers.

Interrelations of growth regulators, carbohydrates and expression of flowering genes (FT, LFY, AP1) in leaf and shoot apex of regular and alternate bearing mango (Mangifera indica L.) cultivars during flowering

Two mango cultivars, Amrapali (regular bearer) and Dashehari (alternate bearer), contrasting for bearing habit was assessed for endogenous phytohormone, carbohydrates and expression of floral inductive genes during flowering. Cultivar Amrapali had higher concentration of growth regulators like ABA, auxin, zeatin, salicylic acid and jasmonic acid and carbohydrates, whereas lower GA content compared to cv. Dashehari and the content increased during flowering when compared to vegetative stage. Gibberellic acid levels was higher in all the stages of bud growth in alternate bearer cv. Dashehari compared to the regular bearer cv. Amrapali. However, the starch content was higher in cv. Amrapali compared to cv. Dashehari. Studies on flower inducing genes indicated that the expression of MiLFY gene was found to be higher in the flowering panicles. Expression of MiFT gene was higher in regular bearer cv. Amrapali compared to the alternate bearer cv. Dashehari and the expression was observed only in leaves which are adjacent to the apical buds during flowering. Higher expression of MiAP1 gene was also observed in the leaves during the floral induction phase. MiFT and MiAP1 genes found to have a negative relation with gibberellic acid and its derivatives. The results indicate that positive regulators of flowering were higher in regular bearing cultivar Amrapali compared to alternate bearing cultivar Dashehari.

Polyamine and ethylene changes during floral initiation in response to paclobutrazol in mango (Mangifera indica L.)

Use of paclobutrazol is common strategy for inducing uniform and profuse flowering in mango. The possible mechanism by which paclobutrazol exert such responses are less understood. The present investigation was carried out to investigate possible role of polyamines and ethylene biosynthesis in the paclobutrazol induced floral induction in mango. Following paclobutrazol soil drenching application (1.25 g a.i. m-1) to mango cv. Totapuri, the free polyamine contents, ethylene production, 1-amino cyclopropane carboxylic acid (ACC) content and ACC oxidase activity were determined in the apical buds and leaves of growing shoots at 4 distinct bud developmental stages numerically characterized as 510 (initiation of bud swelling), 511 (swollen buds), 513 (bud burst) and 515 (panicle emergence) according to standard BBCH scale. The total free polyamines, spermidine and spermine contents increased and ethylene production, ACC content and ACC oxidase activity decreased in the buds and leaves of paclobutrazol treated as compared to untreated trees. In general under paclobutrazol treatment, buds accumulated more polyamines than the leaves. With respect to the bud growth stages, total free polyamines, spermidine and spermine were high at 510/511 stage both in the paclobutrazol treated and untreated trees which declined progressively as shoots approached panicle emergence stage (515). The ethylene production, ACC and ACC oxidase activity exhibited trends opposite to that of polyamines. The study showed that polyamine – ethylene balance may control paclobutrazol induced floral bud induction in mango and accumulation of polyamines-spermidine and spermine in buds appeared as an important factor in facilitating floral induction response.

Nutritional modulation of mouse and human liver bud growth through a branched-amino acid metabolism

Liver bud progenitors experience a transient amplification during the early organ growth phase, yet the mechanism responsible is not fully understood. Collective evidence highlights the specific requirements in stem cell metabolism for expanding organ progenitors during organogenesis and regeneration. Here, transcriptome analyses show that progenitors of the mouse and human liver bud growth stage specifically express the gene branched chain aminotransferase 1, encoding a known breakdown enzyme of branched-chain amino acids (BCAAs) for energy generation. Global metabolome analysis confirmed the active consumption of BCAAs in the growing liver bud, but not in the later fetal or adult liver. Consistently, maternal dietary restriction of BCAAs during pregnancy significantly abrogated the conceptus liver bud growth capability through a striking defect in hepatic progenitor expansion. Under defined conditions, the supplementation of L-valine specifically among the BCAAs promoted rigorous growth of the human liver bud organoid in culture by selectively amplifying self-renewing bi-potent hepatic progenitor cells. These results highlight a previously underappreciated role of branched-chain amino acid metabolism in regulating mouse and human liver bud growth that can be modulated by maternal nutrition in vivo or cultural supplement in vitro.

Bud development and shoot morphology in relation to crown location.

Plant architecture is shaped by endogenous growth processes interacting with the local environment. The current study investigated crown development in young black alder trees, assessing the effects of local light conditions and branch height on individual bud mass and contents. In addition, we examined the characteristics of parent shoots [the cross-sectional area (CSA) of stem and total leaf area, shoot length, the number of nodes, the number and total mass of buds per shoot] and leaf-stem as well as bud-stem allometry, as several recent studies link bud development to hydraulic architecture. We sampled shoots from top branches and two lower-crown locations: one subjected to deep shade and the other resembling the upper branches in light availability. Sampling was carried out three times between mid-July and late October, spanning from the early stages of bud growth to dormancy. Individual bud mass and shoot characteristics varied in response to light conditions, whereas leaf-stem allometry depended on branch height, most likely compensating for the increasing length of hydraulic pathways. Despite the differences in individual bud mass, the number of preformed leaves varied little across the crown, indicating that the plasticity in shoot characteristics was mainly achieved by neoformation. The relationship between total bud mass and stem CSA scaled similarly across crown locations. However, scaling slopes gradually decreased throughout the sampling period, driven by bud rather than by stem growth. This suggests that the allometry of total bud mass and CSA of stem is regulated locally, instead of resulting from crown-level processes.

Sugars take a central position in plant growth, development and, stress responses. A focus on apical dominance.

Sugars take a central position in plant growth, development and, stress responses. A focus on apical dominance.

Unsupervised Clustering of Subcellular Protein Expression Patterns in High-Throughput Microscopy Images Reveals Protein Complexes and Functional Relationships between Proteins

Protein subcellular localization has been systematically characterized in budding yeast using fluorescently tagged proteins. Based on the fluorescence microscopy images, subcellular localization of many proteins can be classified automatically using supervised machine learning approaches that have been trained to recognize predefined image classes based on statistical features. Here, we present an unsupervised analysis of protein expression patterns in a set of high-resolution, high-throughput microscope images. Our analysis is based on 7 biologically interpretable features which are evaluated on automatically identified cells, and whose cell-stage dependency is captured by a continuous model for cell growth. We show that it is possible to identify most previously identified localization patterns in a cluster analysis based on these features and that similarities between the inferred expression patterns contain more information about protein function than can be explained by a previous manual categorization of subcellular localization. Furthermore, the inferred cell-stage associated to each fluorescence measurement allows us to visualize large groups of proteins entering the bud at specific stages of bud growth. These correspond to proteins localized to organelles, revealing that the organelles must be entering the bud in a stereotypical order. We also identify and organize a smaller group of proteins that show subtle differences in the way they move around the bud during growth. Our results suggest that biologically interpretable features based on explicit models of cell morphology will yield unprecedented power for pattern discovery in high-resolution, high-throughput microscopy images.

Effect of the stage of maturity on the leaf percentage of lucerne and the effect of additives on silage characteristics

The first part of the study concerns the effect of the stage of maturity on nutrient content and leaf percentage of lucerne (<I>Medicago sativa</I>). The plants of lucerne were harvested and analysed at several stages of growth. The crop yield and changes in the concentration of nutrients were determined. The leaf percentage from whole herbage and leaf yield were determined as well. The leaf percentage at the growth stage of small buds was significantly (<I>P</I> < 0.05) higher than at the growth stage of bloom in all three cuts (1<sup>st</sup> cut 52.7% vs. 46.62%; 2<sup>nd</sup> cut 52.03% vs. 44.70%; 3<sup>rd</sup> cut 50.58% vs. 46.26%). Crude protein content of lucerne was decreased significantly (<I>P</I> < 0.05) from the large bud growth stage (small buds 219.6 g/kg DM and large buds 203.1 g/kg DM vs. bloom 173.5 and after bloom 154.2). In the second part of the study, the effect of silage additives on fermentation characteristics of lucerne was investigated. The bacterial inoculant (containing homo- and heterofermentative lactic acid bacteria), chemical additive (containing formic acid, propionic acid, ammonium formate and benzoic acid) and the bacterial inoculant with benzoic acid were used for the improvement of fermentation process.The addition of the inoculant with <I>L. buchneri</I> increased acetic acid (<I>P</I> < 0.05) concentration compared to the silage with chemical additive (group I with inoculant – 1.22%; group Ch with chemical additive – 0.84%; group ICh with chemical additive and inoculant – 1.43). Control silage (C) without additive contained 1.14% acetic acid in dry matter.

Subsurface Drip Irrigation and Fertigation of Broccoli

Production of broccoli (Brassica olearaceaL.Italica) in the southwestern USA is highly dependent on inputs of water and N fertilizer to achieve optimum yields and quality. The water and N‐response characteristics of subsurface drip‐irrigated broccoli have not previously been reported. Field experiments were conducted in southern Arizona during 1993 through 1996. The objectives were to determine: (i) subsurface drip‐irrigated broccoli response to a range of soil water tension (SWT), (ii) effects and interactions of water and N fertilizer inputs on crop yield and quality, and (iii) seasonal and daily N uptake. Experiments consisted of factorial combinations of three irrigation regimes (low, medium, and high) and four N rates (60–500 kg N ha−1). Irrigation was applied daily to maintain target SWT, and all N was applied by fertigation. With respect to marketable yield, the optimum SWT was ∼10 kPa in this sandy loam soil, as indicated by response surface models. Marketable yields across all treatments ranged from <3 to >18 Mg ha−1Marketable yield was significantly affected by N rate during all three seasons, and by SWT during two of three seasons. There were no significant SWT × N interactions for marketable yield. Quality parameters (head weight and diameter) were much more responsive to N rate than to SWT, and there were few significant SWT × N interactions for broccoli quality. Broccoli accumulated up to 320 kg N ha−1in the aboveground biomass, and N uptake fluxes were as high as 5 kg N ha−1d−1at the first bud growth stage (825–1000 heat units after planting [HUAP]).

Timing and Concentration of Hydrogen Cyanamide Affect Blueberry Bud Development and Flower Mortality

The effects of hydrogen cyanamide (H2CN2) sprays on vegetative and reproductive bud growth and development were evaluated for `Climax' rabbiteye (Vaccinium ashei Reade) and `Misty' southern highbush blueberry (V. corymbosum L. hybrid). `Climax' plants were sprayed with 0% or 1% H2CN2 (v/v) at each of several time intervals or flower bud growth stages following either 270 or 600 hours of artificial chilling. `Misty' plants were sprayed with 0%, 1%, or 2% H2CN2 (v/v) immediately after exposure to 0, 150, or 300 hours of artificial chilling. H2CN2 application to `Climax' plants at 3 days after forcing (DAF) and at 10% to 30% stage 3 flower bud development dramatically accelerated leafing, and only minimal flower bud damage was observed at these application times. For `Misty', vegetative budbreak was increased and advanced by both H2CN2 spray concentrations, regardless of pretreatment chilling levels; the number of vegetative budbreaks per plant increased with increased concentration. Timing of anthesis did not appear to be affected by H2CN2, but fruit maturity was hastened. Increased pretreatment chilling also hastened fruit development. This effect on maturity appears to be due primarily to increased and accelerated vegetative budbreak, which probably increased leaf: fruit ratios. Greater flower bud mortality from H2CN2 occurred in nonchilled plants than in those chilled for 150 or 300 hours, especially at 2% H2CN2. These results indicate that H2CN2 has potential value in stimulating vegetative bud development, which potentially hastens maturity in blueberries grown under the mild winter conditions of the Southeast. However, spray concentration and timing of application will be critical to successful use of this compound.

Allocation of current photosynthate and changes in tissue dry weight within northern red oak seedlings: individual leaf and flush carbon contribution during episodic growth

Relatively little is known about the changing carbon allocation patterns in species with episodic growth cycles such as northern red oak (Quercus rubra L.). To examine such changing allocation and growth patterns, northern red oak plants were grown from seed in controlled environment chambers through four cycles of growth. 14CO2 was supplied to leaves of the first, second, or third flushes at different Quercus morphological index growth stages within each flush, and the distribution of 14C within the plant was analyzed. Carbon allocation from source leaves of the first and second flush was primarily upward during the subsequent cycle of shoot growth and downward during lag and bud growth stages. All leaves within a flush did not respond the same. Upper leaves allocated most 14C-photosynthate upward during leaf and shoot growth while lower leaves supplied more 14C to lower stem and roots. During the third and fourth flushes, differential allocation from leaves within a flush resulted in essentially equal upward and downward carbon allocation. Growth and allometric relationships reflected these changes in carbon allocation.

Partitioning of current photosynthate to different chemical fractions in leaves, stems, and roots of northern red oak seedlings during episodic growth

The episodic or flushing growth habit of northern red oak (Quercus rubra L.) has a significant influence on carbon fixation, carbon transport from source leaves, and carbon allocation within the plant; however, the impact of episodic growth on carbon partitioning among chemical fractions is unknown. Median-flush leaves of the first and second flush were photosynthetically labeled with 14CO2, and partitioning of 14C into lipids and pigments, sugars, amino acids, organic acids, protein, starch, and structural carbohydrates of source leaves, stem, and roots was determined. In addition, four chemical fractions (sugars, starch, amino acids, and total structural carbohydrates) were quantitatively analyzed in leaves, stems, and roots. Chemical changes in source leaves reflected leaf maturation, changing sink demand during a growth cycle, and leaf senescence. Starch and sugar storage in leaves, stems, and roots during lag and bud growth stages indicate a feedback response of these tissues to decreasing sink strength and temporary storage of both starch and sugar in these plant tissues. Northern red oak, with episodic shoot growth patterns, provides an experimental system in which large changes in sink strength occur naturally and require no plant manipulation. Metabolic changes in leaf, stem, and root tissue of red oak have broad application for other oak species and for both temperate and tropical tree species with cyclic growth habits.

Adjusting Growth Stage Values to Develop a Linear Scale for Apricot Flower Bud Phenology

Integer values used to represent apricot (Prunus armeniaca L.) flower bud growth stages in a phenological scale were adjusted by a simple technique based on cumulative counts of bud observations. Adjusted stage values on a new continuous scale were calculated so that differences between consecutive values were proportional to the frequency with which buds were observed in each growth stage class during the entire assessment period. This meant that adjusted scale values were linearly related to bud development rate at 20 °C. The method was applied to a scale describing flower development from budbreak to petal fall for three cultivars of apricot growing under orchard conditions.

Seasonal Frost Hardiness Changes in the Cranberry Plant

In Wisconsin, the cranberry plant (Vaccinium macrocarpon Ait.) is protected from freezing temperatures by flooding and sprinkle irrigation. Due to the high value of the crop, growers typically overprotect by taking action at relatively warm temperatures. Our goal is to provide recommendations for improved frost protection strategies by studying seasonal hardiness changes in different parts of the cranberry plant (leaves, stems, buds, flowers, fruit). Stages of bud growth were defined and utilized in the hardiness determinations. Samples were collected from mid-April to mid-Oct. 1996 and cuttings were subjected to a series of freezing temperatures in a circulating glycol bath. Damage to plant parts was assessed by visual scoring and observation, ion leakage, and evaluation of the capability to regrow. The following results were obtained: 1) Overwintering structures, such as leaves, stems, and buds, can survive temperatures <–18°C in early spring, and then deacclimate to hardinesses between 0 and –2°C by late spring. 2) In the terminal bud floral meristems are much more sensitive to freeze–thaw stress than are the vegetative meristems. 3) Deacclimation of various plant parts occurred within 1 week, when minimum canopy temperatures were above 0°C, and when the most numerous bud stage collected stayed the same (bud swell). 4) Fruits >75% blush can survive temperatures of –5°C for short durations. By collecting environmental data from the same location we are attempting to relate plant development, frost hardiness, and canopy temperatures (heat units).

Infection Efficiency of Venturia inaequalis Ascospores as Affected by Apple Flower Bud Developmental Stage.

The average infection efficiency of ascospores of Venturia inaequalis deposited on cluster leaves of apple flower buds was 6 to 16%, 3 to 9%, and 0.4 to 0.6% at tight cluster, first pink, and full pink-to-bloom, respectively. No lesions were observed on flower bud cluster leaves at petal fall. However, the leaves on the vegetative shoot emerging from the flower bud were highly susceptible; the average infection efficiency of ascospores on these leaves was 6 to 21%. The infection efficiency was more variable on young cluster and vegetative shoot leaves than on developing and mature cluster leaves. Results from this study indicate that differences in infection efficiency of V. inaequalis ascospores could be identified by apple bud growth stages.

Weed and Corn (Zea mays) Responses to a Hairy Vetch (Vicia villosa) Cover Crop

Field studies were conducted in 1990 and 1991 to determine the effects of corn planting date and hairy vetch control method on the efficacy of fall-planted hairy vetch as a weedsuppressive cover crop for no-till corn. Glyphosate controlled hairy vetch when applied at the early bud growth stage (April), but hairy vetch residue provided no weed control compared to the weedy check. Mowing was not an effective means of suppressing hairy vetch at the early bud stage. Untreated hairy vetch reduced weed biomass 96% in 1990 and 58% in 1991 but reduced yield over 76% in April-planted corn. There was no competition of untreated hairy vetch with corn when corn planting was delayed until May or June (mid- or late-bloom growth stages of hairy vetch). Corn planted in May into untreated hairy vetch yielded similarly to corn planted in a no-cover weed-free check.