基于距离相关系数和KNN回归模型的森林蓄积量估测研究

【目的】探究Landsat8 OLI数据和KNN算法在森林蓄积量估测中的潜力。【方法】以湖南省湘潭县为研究区,采用Landsat8 OLI数据和同时期的二类调查数据,通过距离相关系数筛选特征,分别采用线性回归模型(MLR)、K-近邻模型(KNN)、距离加权KNN模型(DW-KNN)和优化欧式KNN模型(FW-KNN)对森林蓄积量进行估测。使用十折交叉方法进行精度检验,对检验结果进行对比分析。【结果】3种KNN模型的估测结果均高于传统的线性模型,并且在3种KNN模型中,FW-KNN算法效果最好,决定系数达到0.69,为3种模型中最高;3种KNN模型中,本研究优化欧氏距离KNN模型的估测精度最高,其均方根误差为30.3%,相比于传统KNN模型的均方根误差降低了5.1%,相比于DW-KNN模型降低了3.3%。【结论】采用DW-KNN蓄积量估测结果明显优于其他两种模型,说明通过特征与蓄积量的相关性优化样本间的距离是一种可行的KNN优化方法。     

Variability in the performance of juvenile Chinook salmon is explained primarily by when and where they resided in estuarine habitats

Estuarine habitats provide rearing opportunities for the juvenile life stage of anadromous fishes. Because survival is positively correlated with juvenile performance, these estuarine habitats play an important role in population abundance and productivity. To provide information for the recovery of several depressed stocks of Chinook salmon in the Columbia River Basin, we sought to identify the factors that explain variability in performance. Using otolith‐derived estimates of juvenile somatic growth rate as an index of recent performance, we observed a negative nonlinear relationship between growth rate and day of year, and a decreasing and increasing trend of growth rate over the 8 years of this study and distance from the river mouth respectively. Using a generalised linear modelling approach, we found that variability in juvenile somatic growth rate was best explained by where and when individuals were collected, their body size, contaminant loads, stock of origin, and whether a fish was hatchery produced or unmarked. Lastly, we argue that a considerable improvement to the growth rate of juveniles in estuarine habitats is physiologically possible. The results of this 8‐year study provide a baseline of the performance of juvenile Chinook salmon to evaluate habitat restoration programs and to compare against future anthropogenic conditions.     

库存本阅览室与外借库的合并管理

我馆在2005年将中文图书库存本阅览室与中文图书外借库合并成中文书库,进行统一管理。本文结合工作实际,分析了书库合并前分开管理的弊端,阐述了二者合并的意义,提出了合并后的具体管理措施。     

黑杨无性系日CO2固定总量的测定及与苗木生产潜势的相关性研究

通过对不同基因型黑杨无性系植株净光合速率日变化曲线与光合主体叶面积在时间和光强上进行累计积分来计算植株日CO2固定总量P。由于净光合速率日变化曲线不易准确测得，故以光响应曲线和光强的日变化来推测净光合速率日变化。对P值和生物量的相关性研究表明，二者的相关性很高。且认为可以用计算出的P值作为评估苗木生产潜势的指标之一，这为育种中生产力的早期选择提供了有力的证据。     

红肉脐橙(Citrus sinensis L.)果肉中特征色素提取方法探索

红肉脐橙是目前为止惟一果肉呈粉红色或红色的脐橙,其呈色色素尚未见报道。为深入研究其果肉中特征色素的构成及在果实生长发育和贮存过程中色素含量的变化,本研究在对该色素初步定性的基础上对提取方法进行了探索。在分析红肉脐橙果肉中色素萃取液的紫外可见吸收光谱时发现,该色素有类胡萝卜素的特征吸收峰,将萃取液进行薄层层析则进一步证实其内含有番茄红素。在色素提取的前处理方式上,本研究认为冷冻干燥样品的浸提效果较真空干燥的样品稳定(变异系数小),且浸提率相差无几,所以选择冷冻干燥作为合适的样品前处理方式。色素浸提条件的正交实验结果表明,最佳浸提条件为50℃、1.5h内浸提4次,浸提的料液比为1:60。此外,浸提温度、浸提时间、浸提料液比和浸提次数对浸提率都有显著影响,影响最大的因素是浸提温度,最佳温度为50℃,随温度降低浸提率显著降低。其次是侵提时间的长短,最佳浸提时间为1.5h,延长或缩短0.5h都会使浸提率下降。浸提次数对浸提率的影响是最小的。     

Temperature-dependent stock-recruitment model for walleye pollock (Theragra chalcogramma) around northern Japan

Changes in fish year‐class strength have been attributed to year‐to‐year variability in environmental conditions and spawning stock biomass (SSB). In particular, sea temperature has been shown to be linked to fish recruitment. In the present study, I examined the relationship between sea surface temperature (SST), SSB and recruitment for two stocks of walleye pollock (Theragra chalcogramma) around northern Japan [Japanese Pacific stock (JPS) and northern Japan Sea stock (JSS)] using a temperature‐dependent stock‐recruitment model (TDSRM). The recruitment fluctuation of JPS was successfully reproduced by the TDSRM with February and April SSTs, and February SST was a better environmental predictor than April SST. In addition, the JPS recruitment was positively related to February SST and negatively to April SST. The JSS recruitment modeled by the TDSRM incorporating February SST was also consistent with the observation, whereas the relationship between recruitment and February SST was negative, that is the opposite trend to JPS. These findings suggest that SST in February is important as a predictor of recruitment for both stocks, and that higher and lower SSTs in February act favorably on the recruitment of JPS and JSS respectively. Furthermore, Ricker‐type TDSRM was not selected for either of the stocks, suggesting that the strong density‐dependent effect as in the Ricker model does not exist for JPS and JSS. I formulate hypotheses to explain the links between SST and recruitment, and note that these relationships should be considered in any future attempts to understand the recruitment dynamics of JPS and JSS.     

莫比朗在柑桔和土壤中的残留动态研究

采用田间试验方法,研究了莫比朗在柑桔和土壤中的消解动态和最终残留。样品用乙腈提取,过Envi-Florial固相小柱净化,液相色谱紫外检测器检测,外标法定量。残留动态试验结果表明,施药浓度为推荐剂量的两倍时(有效成分120g·hm-2),莫比朗在柑桔中的半衰期分别为7.00～8.90d,土壤中为4.00～6.38d。在有效成分为60g·hm-2的剂量下,施药3次,施药后第21d柑桔中莫比朗残留量低于0.10mg·kg-1。     

清家脐橙在贵州清水江河谷的栽培适应性

目前,我国的柑橘品种以中熟品种居多,成熟期多集中在12月上中旬,大量的鲜果集中上市造成产品在短期内供过于求,影响柑橘生产的经济效益。生产上选用早熟的优质脐橙品种,使熟期提早以调整柑橘果品上市时间和拉长销售期,对我国柑橘品种结构的调整和优化、提高脐橙生产的经济效益和增加农民经济收入都具有重要意义。     

重庆开县3个锦橙品种留树贮藏试验初报

对重庆市开县3个锦橙品种留树贮藏效果进行探索,结果表明,‘开陈72-1’、‘开盛76-8’和‘北碚447’3个锦橙品种留树贮藏至2月下旬,能维持较高的坐果率、商品果率、外观色泽和可溶性固形物含量;‘开陈72-1’最适宜做开县锦橙留树贮藏品种。