The rice soils of Cambodia. II. Statistical discrimination of soil properties by the Cambodian Agronomic Soil Classification system (CASC)

Abstract. A novel agronomic system of soil classification for managing rice soils in Cambodia (CASC) was developed recently. It integrates local knowledge about the soil with its conceptual and taxonomic understanding by soil scientists. Using quantitative soil survey data we evaluated the agronomic efficacy of the classification system. Although the CASC is based on simple field criteria it explained 25 to 44% of the variation in soil prop-erties relevant for agronomic decision making. It failed, however, to differentiate soil types based on pH (6% of variances explained). Despite its simplicity it performed as well or better than the most widely used soil map in Cambodia (Crocker, 1962), and unlike this small-scale soil map it allows classification on a field-specific basis. The average values of soil properties were often significantly different between groups. Organic carbon, clay content and plant available magnesium and calcium contributed most to the discrimination of CASC soil groups. The predictive accuracy when allocating new samples on the basis of quantitative survey data to soil groups of the CASC was 50% to 100%, except for soil groups Kein Svay (0%) and Kampong Siem (20%). The CASC is valuable for managing Cambodian rice soils and may also be used for up-scaling and mapping of soil information.     

不同日平均气温计算方法对气温统计值的影响

日平均气温是反映气候特征的主要指标之一,计算日平均气温的方法有很多种,为了弄清不同的计算方法对日平均气温的影响,笔者对河北省5 个国家基准站的多年逐时气温资料分别进行极值平均、4 次平均、8 次平均3 种计算方法,统计日平均气温并与实际日平均气温进行对比分析。结果表明：3 种计算方法得出的平均气温序列与实际平均气温序列相关性较高,相关系数高达0.9 以上。气温平均值8 次平均最接近于实际日平均气温,平均偏低0.0052℃;极值平均比24 次平均偏高0.4568℃;4 次比24 次平均偏低0.1789℃。增温速率24 次平均最小,24 次平均比4 次平均的气温增温速率偏低0.001~0.03℃/10 a。3 种计算方法的估计精度在10-7和10-4之间,精度较高。在保证资料长度及测站数量的前提下,几种方法均能很好地代表气候的变化趋势。     

云南降水的5个统计指标及其空间分布

利用云南29个气象站1960-2014年的日降水数据,计算集中指数(Concentration Index,CI)、降水集中指数(Precipitation Concentration Index,PCI)、降水集中度(Precipitation Concentration Degree,PCD)、降水集中期(Prec...     

基于改进YOLO v5s的复杂环境下蔗梢分叉点识别与定位

甘蔗蔗梢分叉点的精确识别与高度定位是实现甘蔗收获机切梢器实时控制的关键技术之一,也是提高甘蔗收获机械化水平和降低甘蔗含杂率的重要途径。针对甘蔗地环境复杂、光照变化大、蔗梢分叉点相互遮挡等问题,首先通过田间调查,并现场测试、分析甘蔗生长点、甘蔗分叉点及相互关系的特征规律,采集图像的甘蔗分叉点的统计分析,并结合现场对甘蔗分叉点高度的测量统计分析,发现其均具有明显的正态统计特征。接着,提出了一种基于改进YOLO v5s的蔗梢分叉点识别方法。该方法采用单目和双目相机在广西大学扶绥农科基地采集甘蔗图像数据,并进行数据预处理和标注,构建了甘蔗蔗梢分叉点数据集。然后,在YOLO v5s中引入BiFPN特征融合结构和CA注意力机制,以增强不同层次特征的交互和表达能力,并使用GSConv卷积和Slim-Neck范式设计,在原始模型主干网络中引入Ghost模块替换原始普通卷积,来降低模型的计算量和参数量,提高模型的运行效率。最后,通过在现场采集的数据集上进行训练和测试,验证了该方法的有效性和优越性。实验结果表明,该方法在甘蔗蔗梢分叉点数据集上平均精确率达到92.3%、召回率89.3%和检测时间19.3 m...     

浙江毛竹林分非空间结构特征及其动态变化

利用2004与2009年浙江省两期森林连续清查毛竹样地数据,应用因子分析、聚类分析、相关分析、多元方差分析及核密度估计方法,采用7个指标(样地生物量,样地株数,林分平均胸径,1,2,3和≥4度竹所占比例)描述毛竹林分非空间结构特征。结果表明:毛竹林分非空间结构特征的主导因素依次为年龄结构因子、样地生物量因子、林分平均胸径因子;浙江现有毛竹林老龄化问题较突出,林分中各龄级所占比极不合理,现有经营措施存在较大缺陷,从而使林分生产力衰退;样地株数较少时,林分平均胸径与样地株数几乎没有相关性,但随着样地毛竹株数的增加,表现出一定的负指数关系;影响毛竹林分生物量的主要因素依次是样地株数与林分平均胸径;7个指标中除3度竹所占比例没有显著差异外,其他指标都存在显著差异,使得毛竹林分非空间结构(由7个指标构成)在2004年与2009年存在显著性差异;平均胸径小于8.7 cm的林分随着毛竹的生长所占比重在减少,而平均胸径大于8.7 cm的林分在增加,这也说明2009年林分平均胸径比2004年显著增大,样地株数和林分生物量与林分平均胸径有类似的趋势。研究结果对毛竹林的可持续经营与固碳功能的提高有一定的参考价值。     

Avoiding edge effects in agroforestry experiments; the use of neighbour-balanced designs and guard areas

A number of different types of edge effects on experimental plots are defined, and it is shown that they are frequently important in agroforestry experiments. Methods of eliminating edge effects, either at the design or the analysis stage, are described. Computer simulations, based on uniformity data, are used to investigate the efficacy of neighbour-balanced designs, with and without neighbour terms, for reducing the problem of neighbour effects between different treatments. It is concluded that, while neighbour-balanced designs may be useful in certain situations or in combination with other methods, it is preferable to eliminate edge effects by using adequate guard rows. Recommendations are therefore made concerning the correct use of guard areas in agroforestry experimentation.     

Modeling susceptibility to deforestation of remaining ecosystems in North Central Mexico with logistic regression

Determining underlying factors that foster deforestation and delineating forest areas by levels of susceptibility are of the main challenges when defining policies for forest management and planning at regional scale. The susceptibility to deforestation of remaining forest ecosystems (shrubland, temperate forest and rainforest) was conducted in the state of San Luis Potosi, located in north central Mexico. Spatial analysis techniques were used to detect the deforested areas in the study area during 1993-2007. Logistic regression was used to relate explanatory variables (such as social, investment, forest production, biophysical and proximity factors) with susceptibility to deforestation to construct predictive models with two focuses: general and by biogeographical zone.In all models, deforestation has positive correlation with distance to rainfed agriculture, and negative correlation with slope, distance to roads and distance to towns. Other variables were significant in some cases, but in others they had dual relationships, which varied in each biogeographical zone. The results show that the remaining rainforest of Huastecaregion is highly susceptible to deforestation. Both approaches show that more than 70% of the current rainforest area has high and very high levels of susceptibility to deforestation. The values represent a serious concern with global warming whether tree carbon is released to atmosphere. However, after some considerations, encouraging forest environmental services appears to be the best alternative to achieve sustainable forest management.     

1998～2003年河北省林业科技成果统计与分析

为了使广大林业科技工作者了解河北省林业科学研究进展,有针对性地开展科学研究,通过对1998—2003年河北省林业科技成果的统计分析,对河北省林业科技工作的现状、特点及存在的主要问题进行了分析,并在此基础上提出了加速河北省林业科技发展的若干建议。     

国有林场木材产、销、存计算机辅助管理系统

阐述国有林场木材产、销、存计算机辅助管理系统的设计目标、运行环境、运行特点、数据流程图及系统功能。运用该系统可进行进出仓材积、价格、金额计算 ,各种报表生成和打印 ,完全避免木材生产销售中的计算差错和损失 ,从而提高林场的经营管理水平 ,也便于快速获取各种数据 ,为经营决策提供科学依据。