Plankton is one of the key elements for biological monitoring of riverine ecosystem. It is of paramount importance to understand plankto-environmental interaction for river ecological assessment. This article has presented an alternative model-based analytical framework to quantify plankto-environmental interactions and prediction thereof for a tropical river system. Analysis of secondary data on the Narmada River in India has made it possible to accomplish this. Hierarchical cluster analysis, based on water quality similarity, has been applied for grouping of sites. The resulted clusters slightly deviate from subjective division of the river, but their formation have been coherently ordered downstream-wise. Generalized additive modeling (GAM) with Poisson error assumption of plankton abundance has been the key modeling approach. The GAM, which has explained 66% and 88% deviance of phytoplankton and zooplankton density, respectively, has identified four (transparency, nitrate, temperature, alkalinity) out of thirteen water quality attributes in order of their relative influence on the phytoplankton and zooplankton abundance. The study has established that, in riverine system, application of GAM model under Poisson error assumption has been very effective; and the model has provided reasonably accurate prediction of plankton abundance along hydrologic distance—with noise reduction due to water quality parameters.

浮游生物是河流生态系统生物监测的关键要素之一。理解浮游生物与环境之间的相互作用对于河流生态评估至关重要。本文提出了一种基于模型的替代分析框架，以量化热带河流系统的浮游生物与环境的相互作用及其预测。通过对印度纳尔默达河上的辅助数据进行分析，可以实现这一目标。基于水质相似性的层次聚类分析已应用于站点分组。产生的星团略微偏离了河流的主观划分，但是它们的形成在下游是连贯的。浮游生物丰度具有Poisson误差假设的广义加性建模（GAM）是关键的建模方法。GAM，分别解释了浮游植物和浮游动物密度偏差的66％和88％，已从13种水质属性中确定了四种（透明度，硝酸盐，温度，碱度），以它们对浮游植物和浮游动物丰度的相对影响为顺序。研究表明，在河流系统中，在Poisson误差假设下应用GAM模型非常有效。该模型提供了合理准确的沿水文距离的浮游生物丰度预测，并且由于水质参数而降低了噪声。碱度）从13种水质属性中按其对浮游植物和浮游动物丰度的相对影响顺序排列。研究表明，在河流系统中，在Poisson误差假设下应用GAM模型非常有效。该模型提供了合理准确的沿水文距离的浮游生物丰度预测，并且由于水质参数而降低了噪声。碱度）从13种水质属性中按其对浮游植物和浮游动物丰度的相对影响顺序排列。研究表明，在河流系统中，在Poisson误差假设下应用GAM模型非常有效。该模型提供了合理准确的沿水文距离的浮游生物丰度预测，并且由于水质参数而降低了噪声。