The rising age of the population in rural China and the labor intensity of grafting have resulted in a decrease in the number of grafters and a subsequent increase in their wages. Manual grafting can no longer satisfy the increasing demand for watermelon-grafted transplanting; thus, machine grafting will be an effective alternative. In order to accelerate the implementation of machine grafting in China, a comparative analysis between the automatic grafting machine (model 2TJGQ-800) and traditional hand grafting was conducted. The reliability and feasibility of machine grafting were evaluated through a comprehensive evaluation of the production capacity and grafting seedling quality. This study focuses on the grafting application of watermelon plug-tray seedlings. The scion and rootstock seeds were sown on 9 November 2022. Grafting experiments using an automatic grafting machine, skilled workers, and ordinary workers were conducted with the root-pruned one-cotyledon grafting method on 24 November 2022. The results showed that the machine grafting had a high uniformity and grafting speed. The grafting speed of the grafting machine was 774 plant·h⁻¹ and 1.65–2.55-fold higher than the hand grafting. With training, workers can improve their grafting speed, but it will still be slower than machine grafting. In addition, there was no significant difference in the grafting survival rate between the machine grafting and hand grafting. However, using machine grafting, the success rate decreased from 100% to 90.07% and the rootstock regrowth rate increased from 18.44% to 72.69%. Incomplete rootstock cutting, clip supply failure, and grafting drop failure are the three main factors that result in machine grafting failure. In conclusion, the grafting machine has advantages in terms of grafting speed and uniformity. Upon improving the accuracy of the cutting mechanism and grafting success rate, it will be adopted by commercial nurseries.

The grafting machine cutting step is the core step of the grafting process. The existing grafting machine cutting mechanism adopts fixed angle cutting and manually adjusts the cutting angle based on experience, and the cutting angle is not definite for rootstock cutting in different growth periods. In this paper, we proposed a method to guide the precise cutting mechanism according to internal and external parameters of melon rootstock at a specific period. First, we constructed a cutting model based on internal and external characteristics of rootstock cutting in the growth period of “two leaves and one core” and clarified the safe cutting area. Second, we designed a rotary precision cutting mechanism for rootstock cutting with automatic angle adjustment and constructed the displacement equation of the cutting trajectory of the cutter according to the cutting model. Last, we examined the cutting effect of the precision cutting mechanism and determined the precise cutting angle of the rootstock cutting mechanism in the growth period. Finally, the cutting effect of the precision cutting mechanism was examined, and the precision cutting angle of the rootstock during the growth period was determined. A comparison test between the precision cutting mechanism and the traditional cutting mechanism was carried out, and visual images of the incision were captured and analyzed. The results show that under the five horizontal cutting angles in the safe cutting area of rootstock, the length of the cut surface is inversely proportional to the cutting angle, and the flatness of the cut surface is directly proportional to the cutting angle. Comprehensive evaluation of the length of the cut surface, the cutting success rate, and the quality of cutting revealed that the average cutting angle of the precision cutting mechanism in the safe cutting area (26°) is better than that of the traditional cutting mechanism. It meets the technical requirements of the cutting technology for mechanically grafted rootstocks. The results provide a reference for studying new rootstock precision cutting mechanisms and cutting angle adaptive control models.

The labor-efficient automation of grafting has been recognized as a key factor in the wider adoption of grafting. In growing cucurbits, the root pruned one-cotyledon grafting method is the most commonly used method with grafting machines. The cutting angle, which affects the matching of the rootstock and scion, is key to the survival of the graft seedling. In the production of cucurbit graft seedlings, the cutting parameters are established based on experience, leading to low grafting success. To determine accurate cutting parameters, the watermelon cv. ‘Zaojia84-24′ was used as the scion and the pumpkin cv. ‘Zhenzhuang’ was used as rootstock, and two one-way experiments investigating the cutting angle of the watermelon scion and the rootstock as factors were conducted. The cutting angle of the rootstock and scion had no significant effect on the xylem reconnection rate or the grafting survival rate. A larger cutting angle for the rootstock and scion led to a delay in the reconnection of the phloem. Different cutting angles for the scion significantly affected the growth of the scion after grafting. Compared with a scion graft cutting angle of 10° (SL), graft cutting angles of 14° (SM) and 20° (SS) led to significantly greater scion dry weights, with increases of 16.00% and 18.61%, respectively. Different cutting angles of the rootstock significantly affected the growth of roots after grafting. Compared with a rootstock graft cutting angle of 10° (RL), graft cutting angles of 17° (RM) and 27° (RS) led to significantly greater root dry weights, with increases of 29.33% and 22.54%, respectively. The results of this study can provide a reference for the design of cutting mechanisms for cucurbit grafting robots, improving the cutting precision of grafting robots.

Graft healing is a complex process affected by environmental factors, with temperature being one of the most important influencing factors. Here, oriental melon grafted onto pumpkin was used to study changes in graft union formation and sugar contents at the graft interface under night temperatures of 18 °C and 28 °C. Histological analysis suggested that callus formation occurred 3 days after grafting with a night temperature of 28 °C, which was one day earlier than with a night temperature of 18 °C. Vascular reconnection with a night temperature of 28 °C was established 2 days earlier than with a night temperature of 18 °C. Additionally, nine sugars were significantly enriched in the graft union, with the contents of sucrose, trehalose, raffinose, D–glucose, D–fructose, D–galactose, and inositol initially increasing but then decreasing. Furthermore, we also found that exogenous glucose and fructose application promotes vascular reconnection. However, exogenous sucrose application did not promote vascular reconnection. Taken together, our results reveal that elevated temperatures improve the process of graft union formation through increasing the contents of sugars. This study provides information to develop strategies for improving grafting efficiency under low temperatures.

Rice seed vigor is an important indicator for evaluating seed quality and germination capability. Near-infrared hyperspectral imaging technology has shown great potential for seed vigor assessment. However, significant spectral distribution differences often exist among different rice varieties, resulting in degraded transfer classification performance. Moreover, under small-sample conditions, models have difficulty in sufficiently learning shared discriminative information among different varieties, leading to limited transfer generalization capability and restricting their application in cross-variety agricultural scenarios. In this paper, we propose a novel Dual-Level Probability Consistency FS-PFCE (DPC-FS-PFCE) model to achieve robust transfer classification of rice seed vigor across multiple varieties using near-infrared hyperspectral imaging. A total of 900 seed samples from three rice varieties were used, with 600 Chunyou83 samples serving as the master dataset and 150 samples each from Zhongzu53 and Zhongzao39 serving as slave datasets; transfer performance was evaluated on Zhongzu53, Zhongzao39, and their mixed dataset (MIX). Different probability consistency loss strategies were investigated to achieve joint utilization of global and local probability information. Furthermore, a multi-class weight direction consistency constraint was introduced to enhance parameter alignment and knowledge transfer between the master and slave models. The results demonstrated that the master model established using the Chunyou83 dataset achieved classification accuracies of 93.33%, 88.89%, and 87.78% on the Zhongzu53, Zhongzao39, and mixed MIX datasets, respectively, through the DPC-FS-PFCE transfer model. Compared with the Direct Transfer method, the proposed approach improved classification accuracy by 11.11%, 13.33%, and 8.89%, respectively. In addition, the proposed model exhibited more stable transfer classification performance under different master dataset proportions and complex data scenarios. The proposed DPC-FS-PFCE method effectively improves the cross-variety transfer classification capability for rice seed vigor assessment and provides a promising technical approach for nondestructive quality evaluation of crop seeds.