Global change will disturb the frequency, scale and distribution of harmful algal blooms (HABs), but we are unable to predict future HABs due to our limited understanding of how physicochemical changes in the environment affect interspecific competition between dinoflagellates. Trait-based mechanistic modelling is an important tool to unravel and quantify various direct and indirect interactions between species. The present study explores whether MacArthur's consumer-resource model can be used as a viable base model to predict dinoflagellate growth in closed multispecies systems. To this end, two batch culture experiments (294 cultures in total) with monocultures and multispecies cultures of Alexandrium minutum, Prorocentrum lima, P. micans, Protoceratium reticulatum and Scrippsiella trochoidea were performed. Despite changes to the relative (different nitrate concentrations) and absolute nutrient availability (dilutions of L1 medium), P. micans outcompeted all other species in mixed cultures. Consumer-resource modelling parameterized using monoculture growth correctly predicted this species dominance (R² between 0.80 and 0.95). Parameter estimates revealed that P. micans had a faster uptake of nitrogen when compared to its competitors, but did not differ in resource efficiency and natural mortality rate. Yet, while the model accurately predicted community dynamics during the growth phase, it was not able to predict their dynamics beyond the point of quiescence. Consumer-resource modelling was shown to differentiate the roles of resource assimilation, resource efficiency, and natural mortality rates in batch culture experiments with minimal data requirements beyond common measurements. The results suggest that consumer-resource models provide a promising basis for trait-based modelling of interspecific competition between (harmful) algae.

全球变化将扰乱有害藻华（HAB）的频率，规模和分布，但是由于我们对环境中的理化变化如何影响鞭毛藻种间竞争的了解有限，因此我们无法预测未来的藻华。基于特征的机制建模是阐明和量化物种之间各种直接和间接相互作用的重要工具。本研究探讨了麦克阿瑟的消费者资源模型是否可以用作预测封闭多物种系统中鞭毛虫生长的可行基础模型。为此，进行了两个分批培养实验（总共294个培养物），分别进行了亚历山大藻，利马原螯虾，米克假单胞菌，网纹原菌和Trochoideella trochoidea的单培养和多物种培养被执行。尽管相对（不同硝酸盐浓度）和绝对养分利用率（L1培养基的稀释度）发生了变化，但在混合培养中，P。micans的竞争优势超过了所有其他物种。使用单一养殖生长参数化的消费者资源模型正确地预测了该物种的优势地位（R²在0.80至0.95之间）。参数估计表明，P。micans与竞争对手相比，其氮的吸收速度更快，但资源效率和自然死亡率没有差异。但是，尽管该模型可以准确预测增长阶段的社区动态，但它无法预测超出静态点的动态。消费者资源模型被证明可以区分批处理培养实验中资源同化，资源效率和自然死亡率的作用，而所需的数据量却很少，超出了常规测量范围。结果表明，消费者资源模型为基于特征的（有害）藻类间竞争研究提供了有希望的基础。