不拘于形，不役于物——欧阳修的园林活动与园林书写

欧阳修自幼与园林结缘，仕宦四十年，所到之处植花造园，笔耕不辍，留下数座园林景观、二十多篇园亭记及百余首园林诗词，撰写了现存最早的牡丹专著《洛阳牡丹记》，为后人研究园林提供了珍贵的史料。欧阳修的园林活 动和园林书写与其宦海沉浮、文学创作紧密相关。欧阳修热爱但不耽溺于园林，其崇尚自然、开放自适，不拘于形，不役于物的园林观，拓宽了园林审美的视域，启发了后人对“物”的审美思考，为园林理论注入了新鲜血液。欧阳修进一步打开了园林的大门，推动了园林的开放，为园林的繁荣与普及做出贡献。     

含刚性沉水植物明渠水流结构的试验研究

【目的】得到含刚性沉水植物明渠的水流结构。【方法】采用粒子图像测速仪(PIV),用有机玻璃棒模拟刚性植物,在不同来水流量、植物密度条件下对明渠水流结构进行了试验研究。【结果】无植物时平均流速沿垂向呈对数分布规律,有植物时则呈现明显的分区分布特性。无植物时紊动强度值沿垂向变化不大,有植物时在植物顶端位置处紊动最剧烈。流量或植物密度越大,植物顶端位置的流速梯度就越大,植物层上方的流速最大值也越大。植物密度越大,植物顶端位置的紊动强度越大,植物密度对水流紊动强度由最大值减小到最小值的区域影响很大。【结论】刚性沉水植物的存在会改变水流结构,增强紊动掺混,增强流体质点交换和能量传递,且上述影响会随着流量或植物密度的增大而增强。     

畜牧养殖污染现状及综合治理对策

畜牧养殖业污染问题是畜牧养殖生产过程中的常见问题,对于养殖业也有一定的影响,直接影响畜牧产品质量及动物疫病防治工作。该文针对畜牧养殖业污染现状进行分析研究,阐述畜牧养殖污染问题现状,并综合新疆地区的畜牧养殖业发展状态,提出畜牧养殖污染问题综合治理的有效对策。     

右江流域水生态补偿机制研究初探

为保护右江流域生态环境,协调郁江乃至珠江上下游各方经济利益,更好的激励上游地区生态环境保护行为,通过分析右江流域水生态现状、存在问题、现有生态补偿政策环境及其在郁江流域、珠江流域中的地位和作用,利用外部性理论,分析提出了右江流域水生态补偿模式。通过对标对表有关生态保护规划和区划成果,立足百色市实际,提出了右江流域水生态补偿内容、补偿主客体、补偿标准和补偿方式,初步构建右江流域水生态补偿机制框架,并提出有针对性的补偿政策建议,为水生态补偿实施提供了参考。     

基于卷积神经网络的拖拉机工况识别

农机工况识别在细化农机作业状态和帮助掌握区域污染物排放趋势方面有着重要的研究价值。基于拖拉机不同运行状态下的行驶速度、发动机转速以及实时油耗等时间序列,首次提出将图像识别方法引入到拖拉机工况识别中的思路,并分别应用参数优化的支持向量机与卷积神经网络对实际作业拖拉机工况进行研究。结果表明:(1)基于参数优化的支持向量机可以较好地实现样本点的工况识别且识别准确度达到99.851 9%,但无法实现农机工况的连续性识别,同时无法对农机工况转换阶段进行有效识别。(2)以拖拉机运行速度与发动机转速等信息构建样本图像来描述农机工况变化的数据表达,并在此基础上应用卷积神经网络可以有效实现农机工况的连续性识别,且识别准确率可以达到93.3%。本研究在农机工况识别方面具有一定参考价值,并为后续农机不同工况下区域污染物排放研究提供技术支持。     

面向养殖水体分布差异的COD光谱法检测研究

面向养殖水体，传统光谱法对化学需氧量（Chemical Oxygen Demand，COD）检测模型构建的基础：源域（现有样本库）与目标域（检测地水体）间光谱数据独立同分布。但是当源域与目标域分布间存在差异时，由源域得到的低误差模型常在目标域上表现下滑。针对该问题，提出面向UV Vis光谱的域对抗训练网络（DAUVwpNet），将分布不同的源域和目标域数据映射至相同分布的特征空间中，使其在该空间的分布距离尽可能接近，从而在特征空间中对源域训练的目标函数也可以迁移至目标域上，以降低模型在目标域的误差。试验表明：面向同一批测试数据，DAUVwpNet的预测误差为0.78，要低于传统模型的预测误差（0.85）；DAUVwpNet预测值与实测值间相关系数为0.95，要高于传统模型的相关系数（0.89）。表明了该网络能够较好对齐两域特征空间数据分布，降低因分布差异带来的COD检测误差。     

Cohort‐specific variation in juvenile coho salmon habitat use

In this study, we examined summer and fall freshwater rearing habitat use by juvenile coho salmon (Oncorhynchus kisutch) in the quickly urbanising Big Lake drainage in south‐central Alaska. Habitat use was assessed by regressing fish count data against habitat survey information across thirty study sites using generalised linear mixed models. Habitat associations were examined by age‐0 and age‐1+ cohorts separately, providing an opportunity to compare habitat use across different juvenile coho salmon life stages during freshwater rearing. Regression results indicated that the age‐0 cohorts were strongly associated with shallow, wide stream reaches with in‐stream vegetation, whereas age‐1+ cohorts were associated with deeper stream reaches. Furthermore, associations between fork length and habitat characteristics suggest cohort‐specific habitat use patterns are distinct from those attributable to fish size. Habitat use information generated from this study is being used to guide optimal fish passage restoration planning in the Big Lake drainage. Evidence for habitat use partitioning by age cohort during freshwater juvenile rearing indicates that pooling age cohorts into a single “juvenile” stage for the purposes of watershed management may mask important habitat use dynamics.     

Measuring trail users’ perception of crowding in a peri-urban nature park: A best-worst scaling experiment

This paper evaluates the perception of crowding, its determinants and relationship with trail users’ experience in a peri-urban nature park. Existing crowding surveys rely heavily on a single-item 9-point crowding scale which was originally developed for backcountry areas. Receiver operating characteristic (ROC) analysis was used to test the validity and reliability of this measure in detecting crowding issues and formulating visitor capacity limits. Low validity and reliability raised concerns about the applicability of this scale in similar frontcountry areas. Inability of this simple crowding measure to detect a visitor threshold brings to the fore its another important limitation – ignorance of the complexity of crowding phenomenon. The perception of crowding varies with various social, psychological and situational factors, and a best-worst scaling (BWS) experiment was designed to simultaneously evaluate different aspects of crowding in the Medvednica Nature Park, a peri-urban nature park near Zagreb, Croatia. The findings of BWS experiment revealed that occupancy of mountain huts, number of other hikers on the trail and amount of litter along the trail on average contributed most to the trail users’ sense of crowding. A latent class analysis identified a large heterogeneity in the perception of crowding; while one group of trail users (63%) was more disturbed by the direct social impacts, the others (37%) were more disturbed by the secondary impacts of other visitors, especially by the negative externalities from road traffic in the park. Older visitors had a greater tendency to primarily associate crowding with the environmental conditions. A higher impact of trail use level in the perception of crowding, younger age and more frequent use were associated with a more negative evaluation of crowding. The park management should closely monitor road traffic in the park and visitor activity on highly used trails, especially on those popular among younger trail users.     

Influence of soil phosphorus status,texture, pH and metal content on the efficacy of amendments to pig slurry in reducing phosphorus losses

Cost‐effective strategies for using chemically amended organic fertilizers need to be developed to minimize nutrient losses in surface and groundwater. Coupling specific soil physical and chemical characteristics with amendment type could increase their effectiveness. This study investigated how water‐extractable phosphorus (P) was affected by chemical amendments added to pig slurry and how this effect varied with soil properties. A 3‐month incubation study was conducted on 18 different mineral soils, stored at 10 °C and 75% humidity and treated with unamended and amended slurry which was incorporated at a rate equivalent to 19 kg total P (TP )/ha. The amendments examined were commercial‐grade liquid alum, applied at a rate of 0.88:1 [Al:TP ], and commercial‐grade liquid poly‐aluminium chloride (PAC ), applied at a rate of 0.72:1 [Al:TP ]. These amendments were previously identified by the authors as being effective in reducing incidental losses of P. The efficacy of the amendments varied with the soil test P, the degree of P saturation (DPS ) and the Mehlich aluminium, iron and calcium, but not soil texture. Chemical amendments were most effective in soils with DPS over approximately 20%. Due to their high cost, the incorporation of amendments into existing management practices can only be justified as part of a holistic management plan where soils have high DPS .