Towards benthic ecosystem functioning maps: Quantifying bioturbation potential in the German part of the Baltic Sea
Towards benthic ecosystem functioning maps: Quantifying bioturbation potential in the German part of the Baltic Sea
复制标题
DOI：
10.1016/j.ecolind.2016.10.025
复制
发表时间：
2017-02-01
影响因子：
6.9
通讯作者：
Zettler, Michael L.
中科院分区：
文献类型：
--
作者：
Gogina, Mayya;Morys, Claudia;Zettler, Michael L.
本研究的目的是估计，模型和地图的生态系统功能的空间差异表示的社区生物扰动潜力（BPc），一个指标的底栖动物功能的基础上生物扰动，在德国的波罗的海的一部分。生物扰动潜力计算的有用性是合理的，其温和的，但显着的相关性与来自成岩模型拟合到原位测得的沉积物深度剖面的天然叶绿素-a示踪剂的生物扰动率的估计。评估了BPc的季节和年际变化，并确定了影响研究区生物扰动的关键物种。生成最准确的BP c图并研究其可预测性。在非生物参数方面，我们测试了3种不同的方法：i）最初计算每个站点的底栖大型动物群落生物扰动潜力，并将其作为物种分布建模技术（RandomForest，RF）的响应变量，并将相关的可用环境层用作预测因子，ii）选择35个关键物种作为对BPc贡献最大的物种（负责任）。90%的总BPc），他们的人口生物扰动潜力（BPP）估计被用作响应变量和RF模型拟合在他们每个人，以预测其全覆盖分布，随后总结到BPc，iii）BPc值在站内插到光栅表面使用自然邻居技术。与观测值的BPc的比较表明，自然相邻插值得到的地图是最准确的一个考虑的分辨率为1 × 1公里。(C)2016爱思唯尔有限公司版权所有
This study aims to estimate, model and map the spatial differences of ecosystem functioning expressed by community bioturbation potential (BPc), an indicator of benthic faunal function based on bioturbation, in the German part of the Baltic Sea. The usefulness of bioturbation potential calculations was justified by its moderate but significant correlation with estimates of bioturbation rates derived from diagenetic models fitted to the in situ measured sediment depth profiles of naturally-occurring chlorophyll-a tracer. Seasonal and interannual variations of BPc were assessed and key species contributing to bioturbation in the study area were identified. To generate the most accurate map of BPc and to investigate its predictability based. on abiotic parameters, we have tested 3 different methodological approaches: i) benthic macrofauna community bioturbation potential was initially calculated per station and treated as response variable for species distribution modelling technique (RandomForest, RF) with relevant available environmental layers used as predictors, ii) 35 key species were selected as the most contributing to BPc (responsible. for 90% of total BPc), their population bioturbation potential (BPp) estimates were used as response variables and RF models were fitted on each of them to predict their full coverage distributions, that were subsequently summed up to the BPc, iii) BPc values at stations were interpolated to a raster surface using a natural neighbour technique. The comparison with observed values of BPc indicated that map derived by natural neighbour interpolation was the most accurate one given the considered resolution of 1 x 1 km. (C) 2016 Elsevier Ltd. All rights reserved.