Herpetofauna diversity and microenvironment correlates across a pasture-edge-interior ecotone in tropical rainforest fragments in the Los Tuxtlas Biosphere Reserve of Veracruz, Mexico

We evaluated the relationship between amphibian and reptile diversity and microhabitat dynamics along pasture-edge-interior ecotones in a tropical rainforest in Veracruz, Mexico. To evaluate the main correlation patterns among microhabitat variables and species composition and richness, 14 ecotones were each divided into three habitats (pasture, forest edge and forest interior) with three transects per habitat, and sampled four times between June 2003 and May 2004 using equal day and night efforts. We measured 12 environmental variables describing the microclimate, vegetation structure, topography and distance to forest edge and streams.After sampling 126 transects (672 man-hours effort) we recorded 1256 amphibians belonging to 21 species (pasture: 12, edge: 14, and interior: 13 species), and 623 reptiles belonging to 33 species (pasture: 11, edge: 25, and interior: 22 species). There was a difference in species composition between pasture and both forest edge and interior habitats. A high correlation between distance to forest edge and temperature, understorey density, canopy cover, leaf litter cover, and leaf litter depth was found. There was also a strong relationship between the composition of amphibian and reptile ensembles and the measured environmental variables. The most important variables related to amphibian and reptile ensembles were canopy cover, understorey density, leaf litter cover and temperature.Based on amphibian and reptile affinity for the habitats along the ecotone, species were classified into five ensembles (generalist, pasture, forest, forest edge and forest interior species). We detected six species that could indicate good habitat quality of forest interior and their disappearance may be an indication of habitat degradation within a fragment, or that a fragment is not large enough to exclude edge effects. Different responses to spatial and environmental gradients and different degrees of tolerance to microclimatic changes indicated that each ensemble requires a different conservation strategy. We propose to maintain in the Los Tuxtlas Biosphere Reserve the forest remnants in the lowlands that have gentler slopes and a deep cover of leaf litter, a dense understorey, and high relative humidity and low temperature, to buffer the effects of edge related environmental changes and the invasion of species from the matrix.     

小麦模型算法集成平台在华北平原区的适应性评价

为评价小麦模型算法集成平台（wheat model algorithms integration platform, WMAIP）在华北平原区的适应性,该研究利用华北平原区4个典型试验站多年试验数据,对WMAIP组成的16个模型进行调参和验证,并利用归一化均方根误差（normalized root mean squared error, NRMSE）选择最优模型,最后评价WMAIP集成模型在华北平原区的适应性。WMAIP中组合的16个模型均能有效地模拟土壤水分动态和冬小麦生长发育指标。发育期模拟误差小于4.2%;2 m土层土壤贮水量模拟误差小于7.0%;生物量和产量模拟误差分别在17.3%～23.7%和10.8%～20.8%之间。单个模型的模拟性能不稳定,调参与验证结果的最优模型存在差异。模型集成可降低华北平原区冬小麦产量的模拟误差,用于集成的模型数量越多,模拟误差越小,选择6个模型进行集成就可获得近似田间试验的模拟误差。以16个组合模型模拟结果的均值作为集成模型的结果,得到生物量和产量的模拟误差分别为18.7%和11.8%。结果表明,WMAIP在华北平原区有较好的适应性,可用于华北平原区小麦生产管理和气候变化影响评估。     

基于模糊神经网络集成的土壤资源评价性能的改进

Land evaluation factors often contain continuous-, discrete- and nominal-valued attributes. In traditional land evaluation, these different attributes are usually graded into categorical indexes by land resource experts, and the evaluation results rely heavily on experts＇ experiences. In order to overcome the shortcoming, we presented a fuzzy neural network ensemble method that did not require grading the evaluation factors into categorical indexes and could evaluate land resources by using the three kinds of attribute values directly. A fuzzy back propagation neural network （BPNN）, a fuzzy radial basis function neural network （RBFNN）, a fuzzy BPNN ensemble, and a fuzzy RBFNN ensemble were used to evaluate the land resources in Guangdong Province. The evaluation results by using the fuzzy BPNN ensemble and the fuzzy RBFNN ensemble were much better than those by using the single fuzzy BPNN and the single fuzzy RBFNN, and the error rate of the single fuzzy RBFNN or fuzzy RBFNN ensemble was lower than that of the single fuzzy BPNN or fuzzy BPNN ensemble, respectively. By using the fuzzy neural network ensembles, the validity of land resource evaluation was improved and reliance on land evaluators＇ experiences was considerably reduced.     

基于Boosting的决策树集成土地评价

传统的土地评价方法易受人为因素的限制,探索更科学合理的土地资源评价方法,对土地利用与规划具有重要意义.由于决策树具有分类精度高、分类器可解释性强的优点,特别是C5.0采用了提高决策树分类精度的Boosting技术,提出利用Boosting技术的决策树集成C5.0进行土地评价的方法.采用C5.0算法对广东省七地资源进行了评价,对不使用Boosting的决策树和使用Boosting决策树集成的评价结果进行了分析和比较.研究结果表明利用决策树进行土地质量评价能够得到较高的评价精度,且Boosting决策树集成的土地评价精度高于不使用Boosting的决策树的精度.     

国内外气候模式对黑龙江省汛期降水预测分析

为了评估季节气候模式及多模式集合对黑龙江省汛期降水的跨季节预测能力,最终提高黑龙江省汛期气候预测准确率。基于1983—2017年中、美、欧三种季节气候模式的资料,将多模式集合预报技术应用于黑龙江省汛期降水预测,采用距平相关系数(ACC)、趋势异常综合检验(Ps)评估、分级评分(Pg)评估和距平符号一致率(Pc)4种定量评估方法全面评估了上述3种季节气候模式及多模式集合对黑龙江省汛期降水的跨季节预测能力,并最终给出适合于黑龙江省汛期降水的客观预测方法。结果表明：各家模式对黑龙江省汛期降水有一定的跨季节预报能力,但对于降水趋势的异常量级预测能力相对较差。各家模式预测评分比较来看,EC模式预测评分相对更好,在预测业务中可以重点考虑;多模式超级集合预测评分高于日常业务质量评分和多模式等权集合平均的预测评分,可以在汛期气候预测中参考。