کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
4374148 | 1303162 | 2011 | 7 صفحه PDF | دانلود رایگان |
![عکس صفحه اول مقاله: Analyzing the (mis)behavior of Shannon index in eutrophication studies using field and simulated phytoplankton assemblages Analyzing the (mis)behavior of Shannon index in eutrophication studies using field and simulated phytoplankton assemblages](/preview/png/4374148.png)
The Shannon index of diversity H′ is a commonly used metric in ecology. The tendency of this index to show a unimodal relationship with productivity has been the subject of several studies. In the present work, the behavior of H′ and three related ecological indices (Simpson, Hill, and Evenness) was investigated using phytoplankton assemblages along a eutrophication gradient. We used both natural and simulated assemblages, whereby the comparison enabled us to assess the role of environmental ‘noise’ on index behavior. We developed simulated assemblages based on phytoplankton distributions predicted by two model types: the log series statistical model and the random fraction niche-based model. Using field data, H′ and the related Simpson index showed expected unimodal relationships with eutrophication. The same unimodal relationships were reproduced with simulated assemblages. Comparing the simulations with natural assemblages along a eutrophication gradient showed that there was much unexplained variance in the real-world data, suggesting that these diversity indices are sensitive to stochastic processes. An analysis of the simulated assemblages using relative abundance distributions suggested that increasing H′ and Simpson index values in the low range of the eutrophication gradient were due to increasing species richness, and that decreasing index values in the high range of the eutrophication gradient were due to decreasing evenness. In addition, this analysis revealed how assemblages of identical H′ values arose from contrasting community structures found in the low- and high-range of eutrophication. The high variability and non-linearity of the Shannon and Simpson indices along a eutrophication gradient suggests that these measures of diversity are inappropriate for use as water quality monitoring assessment tools. Indeed, when calculating ecological quality ratios that are employed by the European Water Framework Directive, unreliable (non-monotonic) predictions of water quality resulted.
Journal: Ecological Indicators - Volume 11, Issue 2, March 2011, Pages 697–703