采用双模态联合表征学习方法识别作物病害

基于深度卷积神经网络的视觉识别方法在病害诊断中表现出色,逐渐成为了研究热点。但是,基于深度卷积神经网络建立的视觉识别模型通常只利用了图像模态的数据,导致模型的识别准确率和鲁棒性,都依赖训练数据集的规模和标注的质量。构建开放环境下大规模的病害数据集并完成高质量的标注,通常需要付出巨大的经济和技术代价,限制了基于深度卷积神经网络的视觉识别方法在实际应用中的推广。该研究提出了一种基于图像与文本双模态联合表征学习的开放环境下作物病害识别模型(bimodalNet)。该模型在病害图像模态的基础上,进行了病害文本模态信息的嵌入,利用两种模态病害信息间的相关性和互补性,实现了病害特征的联合表征学习。最终bimodalNet在较小的数据集上取得了优于单纯的图像模态模型和文本模态模型的效果,最优模型组合在测试集的准确率、精确率、灵敏度、特异性和F1值分别为99.47%、98.51%、98.61%、99.68%和98.51%。该研究证明了利用病害图像和病害文本的双模态表征学习是解决开放环境下作物病害识别的有效方法。     

Habitat associations of the Threatened pugnose minnow (Opsopoeodus emiliae) at the northern edge of the species range

Pugnose minnow (Opsopoeodus emiliae) is a small, reclusive species that is widespread in North America, but is one of the rarest fishes in Canada, found in less than 12 known localities in southwestern Ontario. In contrast to most pugnose minnow populations across the global range, Canadian populations are primarily found in turbid systems, potentially indicating persistence in suboptimal conditions. We used data from a multi-gear species and habitat survey in the Canard River, Ontario, a system dominated by agricultural inputs and the best-known capture site of the species in Canada, to parameterise multi-gear occupancy models for understanding the relationship between pugnose minnow occupancy and microhabitat features, including the role of turbidity. Almost 300 pugnose minnow were captured, representing the largest single collection of the species in Canadian history. The best occupancy model indicated that the probability of pugnose minnow occupancy was highest in the deepest sites with the lowest water clarity (i.e. high turbidity); however, competing models suggested that occupancy was highest at sites with wild celery (Vallisneria americana) and higher water clarity, signifying that habitats with low turbidity may be utilised if sufficient physical cover exists. Together, our results suggest that Canadian pugnose minnow populations occupy and potentially favour turbid conditions, possibly to avoid visual predators in clearer habitats. It remains uncertain whether this abiotic association represents a long-term, viable, local adaptation or whether persistence of pugnose minnow in the Canard River is at risk unless significant water quality improvements can be made.     

Distribution and diversity of leaf-cutting ants in Northeastern Argentina: species most associated with forest plantations

Abstract

Leaf-cutting ants (LCA) are considered one of the main herbivores and one of the most destructive pest insects of the Neotropics. Northeastern Argentina harbors the greatest species richness of these ants and in turn comprises the highest surface with forest plantations. Our aim was to establish which species of leaf-cutting ants are most commonly associated with forest plantations by analyzing their geographic distribution using published and unpublished species occurrence data. Also, estimate their potential areas of distribution along a latitudinal gradient that entirely encompasses northeastern Argentina using Ecological Niche Modeling. Only seven of the 20 species recorded were strongly associated with productive systems along the gradient, but only 2–3 species in each region could be considered high-risk species for forest plantations. High-risk species composition shows a turnover between regions. Our models show the potential distribution areas where LCA could become more abundant and dominant, and possibly causing a detrimental effect on the forest plantations in the studied region. We find that ecological niche models are useful tools to assess the environmental suitability of important LCA.     

采用多元统计分析方法构建南极磷虾粉品质评价体系

南极磷虾是一种重要的战略性海洋生物资源，营养价值高，资源量大，开发利用前景广阔。为了构建南极磷虾粉品质评价体系，为南极磷虾粉品质的科学评价提供依据。通过检测15项南极磷虾粉品质指标，采用因子分析筛选品质评价指标，应用概率分布分级品质评价指标，通过层次分析确定品质评价指标权重，利用K-均值聚类分析和判别分析建立南极磷虾粉品质判别函数。结果表明：1）南极磷虾粉品质指标之间离散度有差异，变异系数在8.89%～84.46%；2）从15项指标中选择6项作为南极磷虾粉品评价指标（蛋白质、脂质、盐分、赖氨酸、生育酚、挥发性盐基氮）；3）建立了南极磷虾粉品质评价指标的分级标准和评分标准；4）建立了南极磷虾粉品质判别函数，判别率较高。该研究构建的南极磷虾粉品质评价体系为南极磷虾粉品质评价提供了科学依据，可以用于南极磷虾粉的品质定性判别和评价。     

麻山药种植田沙壤土流动性离散元模型颗粒放尺效应

为简化麻山药-沙壤土复合体离散元模型，提高离散单元法在农业领域中的计算效率，以沙壤土为研究对象，在EDEM离散元软件中构建非球形颗粒，进行双目标参数标定试验，采用放大颗粒粒径方法，利用转鼓、坍塌与FT4流变仿真试验，从颗粒群动态堆积角、流动质量、流动速率以及能量等方面探究了颗粒放尺效应对颗粒群物理特性的影响。试验结果表明：干燥处理后的2 mm粒径沙壤土基质静态堆积角和动态堆积角平均值分别为32.16°和35.02°；与独立标定试验相比，双目标标定试验获得的仿真参数更具准确性与唯一性；在动态堆积角试验中，真实沙壤土颗粒在转鼓中所形成的动态堆积角随粒径和旋转速度的增大而减小，而在仿真试验中，转鼓转速相同情况下，非球形颗粒群在增大粒径的情况下所产生的动态休止角差异较小；坍塌试验中，不同粒径颗粒群在流动过程中的流动质量与平均流速变化趋势基本一致，但误差随粒径增大而增大；颗粒质量相同时，将粒径分别放大2倍及4倍，颗粒数量同比减少87.24%、98.92%，仿真时间明显缩短，计算效率显著提高；FT4流变试验表明，当放尺因子S为2时，阻力FV及其力矩T随时间变化的拟合曲线值约为原尺时的2倍，而当放尺因子S为4时，与原尺相比，拟合曲线斜率差异显著，相关性明显降低。研究结果可为构建沙壤土离散元放尺模型提供理论依据，同时也可为农业工程离散元放尺仿真计算提供一定参考。     

基于SEBAL模型评估干旱半干旱区人工灌丛植被对陆表蒸散的影响

评估人工灌丛植被重建对干旱半干旱区陆地生态系统蒸散的影响,不仅能揭示植被变化与水文过程的耦合机理,又可为区域生态治理与水资源管理提供科学指导。该研究利用Landsat-8OLI/TIR遥感影像及气象数据等驱动SEBAL模型,反演宁夏盐池县的年内不同日期的陆表蒸散,结合目视解译选取的人工灌丛区与对照草地,评估了人工灌丛植被对陆表蒸散的影响。结果表明:1)SEBAL模型的蒸散反演精度与站点观测结果较为一致,可用于干旱半干旱区蒸散反演及空间特征研究;2)盐池县人工灌丛植被区日平均蒸散为1.20mm/d,高于对照草地1.17mm/d的日平均蒸散量,即干旱半干旱区人工种植灌木林增加了生态系统水分消耗,但不同季节和不同生物地理条件下的蒸散增强作用存在差异,蒸散增强在8月份最大,而3、4月份呈现负效应;3)人工灌丛的密度越大、植被盖度越高,对陆表蒸散的增强作用越强,特别在NDVI0.4的高盖度情况下蒸散增强作用更加明显。由此可知,在水资源紧缺的干旱半干旱区开展以灌木树种为主的植被重建需在合理的生态水文阈值范围内开展,才能构建出稳定可持续的人工生态系统。     

土壤外源钴对大麦根伸长的毒害及其预测模型

选取我国11种不同性质的农田土壤，通过外源添加重金属钴（Co），研究其对大麦（Hordeum vulgare L.）根伸长的毒性阈值及土壤性质对Co毒性的影响。结果发现，Co对大麦根伸长10%抑制效应（EC₁₀）在11种土壤中的变化范围为37.1~3 914 mg·kg^-1土（105.5倍），50%抑制效应（EC₅₀）的变化范围为166.1~6 030 mg·kg^-1土（36.3倍）。建立土壤性质与毒性阈值的回归方程，结果表明土壤pH是影响土壤Co毒性阈值最重要的因子，作为单因子时分别可以解释77.6%、72%的EC₁₀和EC₅₀的变异（P≤0.001）。当在EC₁₀预测模型中引入土壤pH和土壤黏粒（Clay）双因子时，可以解释83.9%的EC₁₀的变异（P<0.001），EC₅₀预测模型中引入土壤pH和总碳（TC）双因子时，可以解释86.1%的EC₅₀的变异（P<0.001）。将我国土壤中得到的Co毒性阈值预测模型和欧洲北美10种土壤的预测模型进行比较验证，结果发现基于我国土壤得到的预测模型可以较为准确地预测欧洲北美土壤中Co的大麦根伸长毒性阈值，但基于欧洲北美土壤的预测模型不能准确预测我国土壤中Co的毒性阈值。研究表明，我国土壤性质对Co毒性有显著的影响，基于土壤性质建立的预测模型可为土壤中Co生态风险评价提供参考依据。     

基于电磁感应仪数据的南疆棉田土壤电导率反演模型研究

土壤盐渍化是南疆棉田优质高产的主要障碍因素,定额灌溉可有效改良土壤盐渍化,EM38-MK2可快速、动态获取土壤盐渍化数据,适时监测土壤盐渍化,为棉田定额灌溉提供数字依据。针对电磁感应仪(EM38-MK2)的发射圈和接收圈设计了不同采样方案,在同一条田内采集了6个不同时期的土壤表观电导率数据及相应的剖面土样,分析了不同土壤采集方案及土壤含水量对表观电导率模型精度的影响,对比了以单一时期数据建模的局部模型和6个时期整体数据建模的全局模型的反演精度。结果表明,由单点采集电磁感应仪发射线圈位置土样所构建的模型精度更高,稳定性更好,能有效减少由采样带来的误差。当土壤含水量低于10%时,表观电导率与实测电导率之间的相关性较低,决定系数为0.58,反演模型只具备粗略估计实测电导率的能力;当土壤含水量高于10%时,表观电导率与实测电导率具有很好的相关性,决定系数达到0.80以上,反演模型具有较好的预测能力。EMH+EMV多测定模式下表观电导率与实测电导率之间的模型精度高于EMH或EMV单一测定模式,确立EC_(h0.375)+(EC_(h0.75)+EC_(v0.75))/2+EC_(v1.5)为建模因子,有效提高了反演模型的精度。不同深度土层的局部模型反演精度均高于全局模型精度,局部模型的RPD均大于2.0,具有较好的预测能力。     

北京松山自然保护区主要树种冠幅预测模型研究

冠幅作为林业研究的关键因子，开展冠幅的研究对于精准林业发展具有重要意义。为了更好地利用冠幅对树木其他因子进行预测，选取了北京市松山自然保护区油松、白蜡、蒙古栎等14个常见树种作为研究对象，基于6种回归方程，借助样地实测数据分别建立每个树种的胸径 冠幅和树高 冠幅模型，并针对每个树种筛选出的最优方程进行了精度检验。结果表明，研建的14个树种的模型具有较好的拟合精度，决定系数R2绝大多数都大于0.5，其中，核桃的胸径 冠幅模型最高，R2=0.951，树高-冠幅模型R2=0.887，蒙古栎树高-冠幅模型最低，R2=0.506，也超过了0.5。所建模型的精度评价指标具有一定的统计学意义，对北京市松山地区大面积林分参数的预测预估有重要的参考价值。     

Linking small pelagic fish and cetacean distribution to model suitable habitat for coastal dolphin species,Delphinus delphis and Tursiops truncatus,in the Greek Seas (Eastern Mediterranean)

1. A large‐scale assessment of the summertime suitable habitat for Delphinus delphis (short‐beaked common dolphin) and Tursiops truncatus (common bottlenose dolphin) in Greek Seas (Eastern Mediterranean) was conducted using data from dedicated and opportunistic cetacean surveys and published data records.
2. Using a presence/absence approach, generalized additive models were applied to define a suite of environmental, bathymetric and biotic factors that best describe common and bottlenose dolphin spatial distribution, during early (May, June, July) and late (August, September) summer.
3. A geographic information system (GIS) was used to integrate sightings data with environmental characteristics, distance from the coast and sardine probability of presence. These variables were considered as good proxies for defining species‐suitable habitat within the study area's coastal environment.
4. The final selected models were used to produce annual probability maps of the presence of the species in the entire Greek Seas, as a measure of habitat suitability. Based on the mean probability and standard deviation maps for the study period GIS techniques were subsequently used to determine the persistent (areas with high mean and low variation) and occasional (high mean and high variation) habitat of each species.
5. Results showed that there was a high probability of common dolphin presence in areas with a high probability of sardine presence. For bottlenose dolphin, higher probability of the presence of species occurred in areas closer to the shore, with a high probability of sardine presence and with high concentrations of chlorophyll‐a.
6. In both seasons, the North Aegean Sea and the Inner Ionian Sea Archipelago were indicated as the most suitable areas for common dolphin distribution. Persistent habitat areas of the bottlenose dolphin included enclosed seas, continental shelf waters, and waters surrounding islands. The indicated suitable areas are discussed along with deficiencies of the models and future implications for conservation.

Copyright © 2016 John Wiley & Sons, Ltd.