van Dobben Han F, de Vries Wim
Wageningen University and Research Wageningen The Netherlands.
Ecol Evol. 2016 Dec 18;7(1):214-227. doi: 10.1002/ece3.2485. eCollection 2017 Jan.
We evaluated effects of atmospheric deposition of nitrogen on the composition of forest understorey vegetation both in space and time, using repeated data from the European wide monitoring program ICP-Forests, which focuses on normally managed forest. Our aim was to assess whether both spatial and temporal effects of deposition can be detected by a multiple regression approach using data from managed forests over a relatively short time interval, in which changes in the tree layer are limited. To characterize the vegetation, we used indicators derived from cover percentages per species using multivariate statistics and indicators derived from the presence/absence, that is, species numbers and Ellenberg's indicator values. As explanatory variables, we used climate, altitude, tree species, stand age, and soil chemistry, besides deposition of nitrate, ammonia and sulfate. We analyzed the effects of abiotic conditions at a single point in time by canonical correspondence analysis and multiple regression. The relation between the change in vegetation and abiotic conditions was analyzed using redundancy analysis and multiple regression, for a subset of the plots that had both abiotic data and enough species to compute a mean Ellenberg value per plot using a minimum of three species. Results showed that the spatial variation in the vegetation is mainly due to "traditional" factors such as soil type and climate, but a statistically significant part of the variation could be ascribed to atmospheric deposition of nitrate. The change in the vegetation over the past c. 10 years was also significantly correlated to nitrate deposition. Although the effect of deposition on the individual species could not be clearly defined, the effect on the vegetation as a whole was a shift toward nitrophytic species as witnessed by an increase in mean Ellenberg's indicator value.
我们利用欧洲范围监测项目“ICP森林”的重复数据，评估了大气氮沉降在空间和时间上对森林林下植被组成的影响，该项目聚焦于正常管理的森林。我们的目标是评估，在相对较短的时间间隔内，利用管理森林的数据，通过多元回归方法是否能够检测到沉降的空间和时间效应，在此期间，树木层的变化是有限的。为了描述植被特征，我们使用了基于每个物种覆盖百分比的多元统计指标，以及基于物种存在/不存在情况得出的指标，即物种数量和埃伦贝格指示值。作为解释变量，除了硝酸盐、氨和硫酸盐的沉降外，我们还使用了气候、海拔、树种、林分年龄和土壤化学性质。我们通过典范对应分析和多元回归分析了单个时间点非生物条件的影响。对于既有非生物数据又有足够物种以使用至少三个物种计算每个样地平均埃伦贝格值的样地子集，我们使用冗余分析和多元回归分析了植被变化与非生物条件之间的关系。结果表明，植被的空间变化主要归因于“传统”因素，如土壤类型和气候，但有统计学意义的一部分变化可归因于硝酸盐的大气沉降。过去约10年植被的变化也与硝酸盐沉降显著相关。尽管沉降对单个物种的影响无法明确界定，但对整个植被的影响是向嗜氮物种的转变，这表现为平均埃伦贝格指示值的增加。
