Abstract Temporal variability in phytoplankton is driven by a range of biotic and abiotic factors. In this study, we aimed to identify and estimate the important explanatory variables affecting the diatom species and their abundance in the coastal waters of Bangladesh. For this, species interaction and six explanatory variables were taken into consideration to develop a density dependent Bayesian regression model for quantifying the relative importance of different explanatory variables causing the variability in diatom abundance. Thus, three nutrients such as silicate, nitrate and phosphate, and zooplankton were identified as the most important predictors for 19–25 species. Moreover, the effect of three nutrients and salinity was positive for most diatoms, while temperature and zooplankton had a negative effect. Species interaction and environmental variables combinedly explained 77% of the variability in diatom abundance, with the latter being the dominant factor. Specifically, six environmental variables combinedly explained 57% of the variability in diatom abundance and the remaining 20% of the variability was due to species interaction. This is the first modelling study on phytoplankton community describing the dynamics of diatom species in the coastal waters of northern Bay of Bengal.

中文翻译：

---

亚热带沿海水域硅藻的驱动因素：贝叶斯建模方法

摘要 浮游植物的时间变异是由一系列生物和非生物因素驱动的。在这项研究中，我们旨在确定和估计影响孟加拉国沿海水域硅藻种类及其丰度的重要解释变量。为此，考虑了物种相互作用和六个解释变量来开发密度相关的贝叶斯回归模型，用于量化导致硅藻丰度变化的不同解释变量的相对重要性。因此，硅酸盐、硝酸盐和磷酸盐以及浮游动物等三种营养物质被确定为 19-25 个物种的最重要预测因子。此外，三种养分和盐度对大多数硅藻的影响是正面的，而温度和浮游动物的影响则是负面的。物种相互作用和环境变量共同解释了 77% 的硅藻丰度变异性，后者是主导因素。具体来说，六个环境变量共同解释了 57% 的硅藻丰度变异性，其余 20% 的变异性是由于物种相互作用造成的。这是第一次对浮游植物群落进行建模研究，描述了孟加拉湾北部沿海水域硅藻物种的动态。