Various edge response of ground beetles in edges under natural versus anthropogenic influence: A meta-analysis using life-history traits
Various edge response of ground beetles in edges under natural versus anthropogenic influence: A meta-analysis using life-history traits
复制标题
自然与人为影响下边缘地面甲虫的各种边缘响应：使用生活史特征的荟萃分析
DOI：
10.17109/azh.65.suppl.3.2019
复制
发表时间：
2019
影响因子：
0.8
通讯作者：
B. Tóthmérész
中科院分区：
文献类型：
--
作者：
T. Magura;G. Lövei;B. Tóthmérész
由于自然栖息地的日益碎片化和丧失，世界范围内的边缘正在增加。在形成后，边缘通过各种过程(自然和持续的人为干预：林业、农业、城市化)来维持，这些过程影响个体物种对边缘效应的反应(基于历史的边缘效应假说)，这将反映在组合的多样性上。然而，多样性并不是边缘效应的最合适指标，因为具有不同特征的物种对边缘的反应可能不同。为了进一步阐明基于历史的边缘效应假说，我们评估了边缘效应对森林边缘最常用的生活史特征之一--地面甲虫的取食习性的影响。一项基于28篇出版物和422项比较的荟萃分析表明，自然干预与持续的人为干预作为维持边缘的过程反映在性状水平上。自然边缘的草食性、杂食性和捕食性地甲物种的丰度均高于林内，而在持续人为影响的边缘则没有类似的格局。这些结果表明，反复的人为影响所承受的边缘的结构和环境变化对生态系统功能产生不利影响，对病虫害或杂草防治等生态系统服务产生负面影响。对于异质性，Q和I2的互补度量计算的Renstein Q是数据集内的加权平方和。对于显著的异质性，可以根据预期偏差进行测试，假设所有研究都有共同的效应大小。I2衡量的是观察到的差异的比例，它反映了不同研究之间效果大小的实际差异。总方差(Q合计)被分成组内(Q内)和组间(Q组间)方差，并对这些方差进行统计学意义检验(Borenstein等人)。2009年)。组间显著差异(Q介于)意味着边缘对丰度的影响因边缘维持过程的不同而显著不同。在计算过程中，病例少于5例的亚组被排除在分析之外。使用漏斗图和Egger检验来测试发表偏倚(Borenstein等人。2009年)。在显著不对称的情况下，使用修剪和填充方法(Duval&Twedie 2000)。在计算过程中，在R编程环境(版本3.4.3；R Core Team 2017)中使用了MAD(版本0.8-2，Del Re&Hoyt 2014)和Metafor包(版本1.9-9，Viechtbauer 2010)。
Edges are on the increase world-wide due to increasing fragmentation and loss of natural habitats. After formation, edges are maintained by various processes (natural vs. contin ued anthropogenic interventions: forestry, agriculture, urbanization) which influence the reaction of individual species to edge effects (history-based edge effect hypothesis), and this will be reflected in the diversity of assemblages. Diversity, however, is not the most appropriate indicator of the edge effect because species with different traits may respond differently to the edges. To further articulate the history-based edge effect hypothesis, we evaluated the edge effect on one of the most commonly used life-history traits, the feed ing habit of ground beetles in forest edges. A meta-analysis based on 28 publications and 422 comparisons showed that natural vs. continued anthropogenic interventions as edge-maintaining processes reflected at the trait level. Abundance of herbivorous, omnivorous, and predatory ground beetle species were all higher in the natural edges than in the forest interiors, while no similar pattern occurred in edges with continued anthropogenic influ -ence. These results suggest that structural and environmental changes at edges sustained by repeated anthropogenic influence adversely influencing ecosystem functions, with neg ative effects on ecosystem services like pest or weed control. To heterogeneity, complementary measures of Q and I 2 calculated renstein Q is the weighted sum of squares within a data set. For significant het erogeneity, it can be tested against the expected deviation assuming that all studies share a common effect size. I 2 measures the proportion of the observed variance that reflects real differences in effect size between studies. Total variance ( Q total ) was partitioned into within- ( Q within ) and between group ( Q between ) variances and these were tested for statistical significance (Borenstein et al. 2009). Significant variance between groups ( Q between ) means that edge effect on abundance significantly differed according to the maintaining processes of edges. During the calculations, subgroups with less than five cases were excluded from analyses. Publication bias was tested using funnel plots and the Egger test (Borenstein et al. 2009). In the case of significant asymmetry, the trim and fill method was used (Duval & Tweedie 2000). During the calculations, the MAd (version 0.8-2, Del Re & Hoyt 2014) and metafor packages (version 1.9-9, Viechtbauer 2010) were used in R programming environ ment (version 3.4.3; R Core Team 2017).