“WaterMan”: an on-farm water management game with heuristic capabilities

A modern computer-based simulation tool (WaterMan) in the form of a game for on-farm water management was developed for application in training events for farmers, students, and irrigators. The WaterMan game utilizes an interactive framework, thereby allowing the user to develop scenarios and test alternatives in a convenient, risk-free environment. It includes a comprehensive soil water and salt balance calculation algorithm. It also employs heuristic capabilities for modeling all of the important aspects of on-farm water management, and to provide quantitative performance evaluations and practical water management advice to the trainees. Random events (both favorable and unfavorable) and different strategic decisions are included in the game for more realism and to provide an appropriate level of challenge according to player performance. Thus, the ability to anticipate the player skill level, and to reply with random events appropriate to the anticipated level, is provided by the heuristic capabilities used in the software. These heuristic features were developed based on a combination of two artificial intelligence approaches: (1) a pattern recognition approach and (2) reinforcement learning based on a Markov decision processes approach, specifically the Q-learning method. These two approaches were combined in a new way to account for the difference in the effect of actions taken by the player and action taken by the system in the game world. The reward function for the Q-learning method was modified to reflect the suggested classification of the WaterMan game as what is referred to as a partially competitive and partially cooperative game.     

Promoting sustainable intensification in precision agriculture: review of decision support systems development and strategies

Precision agriculture provides important issues toward a more sustainable agriculture. Many farmers have the necessary technology to operate site-specifically, but they do not use it in practice, and thus available information and communications technology (ICT) systems are not used to their full potential. This paper addresses how to reduce the so-called “problem of implementation”, based on the knowledge that participatory approaches during the design and development process is one of the most important factors to frame technology adoption. The development of sustainable ICT systems through theories and methodologies from the fields of human computer interaction and user-centered design (UCD) is presented and an ongoing Swedish project for development of an agricultural decision support system (AgriDSS) for nitrogen fertilization is used as an example to frame the issue. The overreaching aim is to develop AgriDSSs that are sustainable in design as well as through design by stressing the importance of participatory approaches for the successful development of AgriDSSs. The Swedish project has the intention to apply a UCD approach, and some pitfalls on starting to use this way of working is identified as well as some suggestions on how to reduce them through co-learning processes. Despite the challenges presented in this paper, ICT can contribute significantly to long-term sustainable development. Thus, several competences and scientific disciplines need to act in concert to help develop a sustainable development of agriculture via a transdisciplinary approach that can make an impact on society at many levels.     

Inconspicuous restraint of rice seedling growth by root-infecting fungi in soil of a rice paddy field

 Rice seedling growth, estimated by plant height and root development and discoloration, was better in pasteurized soil than in unpasteurized soil obtained from a flooded rice field. Rice seedlings also grew better in sterilized soil modified by adding roots harvested from the pasteurized soil than in soil modified by adding roots harvested from the unpasteurized soil. The results demonstrate that seedling growth in the rice field soil was inhibited by soil microorganisms, even though no typical symptoms such as seedling blight or damping-off appeared. Pythium aristosporum is suggested to be involved in the inhibition. Thus, it appears that inconspicuous restraint of rice seedling growth could occur in soils of rice paddy fields. Received: May 20, 2002 / Accepted: October 16, 2002 Acknowledgments The authors thank Dr. T. Ichitani, former professor at Osaka Prefectural University, for providing an isolate of Pythium aristosporum for comparison, and Mr. Mitsuaki Sato of Akita Prefectural College of Agriculture for technical assistance.     

Soybean oil content QTL mapping and integrating with meta-analysis method for mining genes

Oil content of soybean was a valuable quantitative trait controlled by multiple genes. Eleven QTLs were detected by both CIM and MIM method with the population crossed between Charleston and Dong nong594 in recent 3 years (2007, 2008, 2009). Combining the QTLs collected over the past 20 years, an integrated map of oil-content major QTLs in soybean was established using soymap2, which was published in 2004, as a reference. Using the software BioMercator ver.2.1, QTLs were projected from their own maps onto the reference map. In total, ninety-eight QTLs were integrated into soymap2. A meta-analysis method was used to narrow down the confidence interval, and 20 consensus QTLs and their corresponding markers were obtained. Using a local version of GENSCAN, 10,137 sequences in the consensus QTL intervals were predicted. With BLAST, these predicted genes were compared to the International Protein Index database to mine the related genes. The results offer a basis for gene mining and molecular breeding in soybean.     

Heterosis,molecular diversity,combining ability and their interrelationships in short duration maize (<Emphasis Type="Italic">Zea mays</Emphasis> L.) across the environments

Five inbred lines, 10 single cross maize (Zea mays L.) hybrids and two standard cultivars as check were used to study the combining abilities and heterosis under three environmental conditions. Random amplified polymorphic DNAs (RAPDs) markers were used to study the genetic diversity (GD) and further to analyze relationship of RAPDs based GD with combining ability and heterosis in short duration maize. Spearman’s rank correlation coefficients and linear regressions were analyzed to identify the most important factor determining heterosis and per se performance of the hybrids. Variances due to general combining ability (GCA), specific combining ability (SCA) and their interactions with environment were found to be significant. Twenty random primers generated 179 RAPD fragments. Of these, 102 RAPD fragments were polymorphic. GD was determined using Jaccard’s similarity coefficient, and a dendogram was constructed by UPGMA cluster analysis. The RAPDs based GD exhibited non-significant negative or positive association, non-significant linear regression along with very low coefficient of determination (R ²) with SCA, high and mid parent heterosis (HP and MP) and per se performance of the hybrids. Significant positive correlations and regressions along with high coefficients of determination were recorded for SCA with HP, MP and per se performance of the hybrids. The HP and MP also established significant positive association and linear regression along with high coefficient of determination with per se performance of hybrids whereas the parental mean did not establish any significant correlations with the GD, HPH, MPH and grain yield of F₁s. The present investigation, therefore, did not find any role of RAPDs based GD in determining hybrid heterosis and hybrid performance in short duration sub-tropical maize. The SCA, however, has emerged as the most important factor in determination of heterosis as well as per se performance of the hybrids in short duration maize.     

基于SE模块和ResNet的番茄病虫害识别

番茄病虫害是引起番茄减产的重要因素。精确识别病虫害种类是当前的国际热点问题之一,有助于及时有效的采取针对性的病虫防治办法,减少和避免因番茄减产导致的经济上的损失。本文针对传统虫害识别方法存在的效率和精确率低的问题,利用Kaggle网站上的Tomato数据集,构建基于压缩和激励（Squeeze-and-Excitation,SE）模块的深度残差网络模型（ResNet）,优化番茄病虫害识别方法。结果表明：（1）通过Pytorch框架下的迁移学习,改进后的网络模型对番茄病虫害图像的平均识别准确率最高为97.96%;（2）基于SE模块的ResNet网络模型有助于增强特征区分能力,增加了模型的通用性和鲁棒性。本文研究结果对番茄病虫害的及时监测和处理、提高番茄产量具有重要意义。     

Effects of carbon and nitrogen amendment on soil carbon and nitrogen mineralization in volcanic immature soil in southern Kyushu,Japan

Soil N mineralization is affected by microbial biomass and respiration, which are limited by available C and N. To examine the relationship between C and N for soil microbial dynamics and N dynamics, we conducted long-term laboratory incubation (150 days) after C and N amendment and measured changes in C and N mineralization, microbial biomass C, and dissolved C and N throughout the incubation period. The study soil was volcanic immature soil from the southern part of Japan, which contains lower C and N compared with other Japanese forest soils. Despite this, the area is covered by well-developed natural and plantation forests. Carbon amendment resulted in an increase in both microbial biomass and respiration, and net N mineralization decreased, probably due to increasing microbial immobilization. In contrast, N amendment resulted in a decrease in microbial respiration and an increase in net N mineralization, possibly due to decreased immobilization by microbes. Amendment of both C and N simultaneously did not affect microbial biomass and respiration, although net N mineralization was slightly increased. The results suggested that inhibitory effect on microbial respiration by N amendment should be reduced if carbon availability is higher. Thus, soil available C may limit microbial biomass and respiration in this volcanic immature soil. Even in immature soil where C and N substrate is low, soil C, such as plant root exudates and materials from above- and belowground dead organisms, might help to maintain microbial activity and N mineralization in this study site.     

番茄抗叶霉病基因Cf19的抗性特点及遗传连锁分析

番茄抗叶霉病基因Cf19的抗性范围基本涵盖了中国目前分化出的所有叶霉菌生理小种,且抗性显著;台盼蓝染色结果显示该基因介导的抗性应答中出现明显的超敏反应,强度介于Cf4和Cf9基因之间;遗传连锁分析说明该基因的遗传特性符合单基因显性遗传规律;通过进一步SSR和AFLP分子标记连锁分析,共找到6个与Cf19基因连锁的分子标记,并将该基因初步定位于番茄1号染色体短臂区域。     

采用改进Mask R-CNN算法定位鲜食葡萄疏花夹持点

为实现鲜食葡萄疏花机械化与自动化,该研究提出了一种鲜食葡萄疏花夹持点定位方法。首先基于ResNeXt骨干网络并融合路径增强,改进Mask R-CNN模型,解决鲜食葡萄花穗、果梗目标较小难以检测的问题;进而针对花穗、果梗生长姿态的复杂性与不确定性,提出一种集合逻辑算法,该算法采用IoU函数剔除重复检测的花穗与果梗,建立花穗、果梗对,并对果梗掩模进行形态学开运算,利用集合关系获取主果梗掩模,确定以主果梗质心附近的中心点为果梗的夹持点。最后,随机选取测试集中的图像进行试验。试验结果表明：果梗夹持点平均定位准确率为83.3%,平均定位时间为0.325 s,夹持点x、y方向定位误差及定位总误差最大值分别为10、12和16像素,能够满足鲜食葡萄疏花的定位精度与速度要求,可为实现鲜食葡萄疏花机械化与自动化提供理论支撑。     

Handling missingness when modeling the force of infection from clustered seroprevalence data

Modeling infectious diseases data is a relatively young research area in which clustering and stratification are key features. It is not unlikely for these data to have missing values. If values are missing completely at random, the analysis on the complete cases is valid. However, in practice this assumption is usually not fulfilled. This article shows the effect of ignoring missing data in modeling the force of infection of the bovine herpesvirus-1 in Belgian cattle and proposes the use of weighted generalized estimating equations with constrained fractional polynomials as a flexible modeling tool.