Diatoms are powerful biomonitoring indicators for contemporary and past limnological conditions, provided that their environmental optima are well-defined. Surface-sediment-calibration (or training) sets are widely used to develop quantitative transfer functions linking species distributions to limnological variables. Accurate estimates of each taxon’s environmental optimum benefit from a large dataset spanning a broad gradient of the variable of interest. Here, we compiled data from surface-sediment-diatom analyses from over 450 lakes in Ontario (Canada) to identify ecological optima for the province’s most common taxa. The data were collected from the same laboratory using similar methods and spanned more than three decades of research (1987–2019) on 546 samples (representing 464 unique sites). We describe the ecological responses of 52 common diatom taxa to three limnological variables that are key to lake management: pH, total phosphorus (TP), and maximum lake depth. The gradients were broad (pH: 4.2–9.3; TP: 1.1–96.8 µg L⁻¹; maximum depth: 0.7–95.0 m), enabling robust statistical analyses. We used hierarchical logistic regression modelling to identify statistically significant ecological response curves for all common taxa, and weighted averaging to identify the optima. In our dataset, 96% of common diatom taxa had significant ecological responses and optima for pH, 81% for TP, and 83% for depth. Aside from the benthic cosmopolitan Achnanthidium minutissimum complex (occurring in 88.1% of samples), the most frequently recorded taxa were planktonic, including Discostella stelligera/pseudostelligera (83.2%), Asterionella formosa (77.8%), and the Lindavia bodanica complex (76.6%). Our results illustrate the responses of diatom taxa across key limnological gradients, providing important autecological information to advance their use as bioindicators in environmental assessments. These data can be used to address management and knowledge gaps in understudied areas of Ontario where calibration sets are not available and can support research in regions with similar limnological environments and diatom communities.

硅藻是当代和过去湖泊学条件的强大生物监测指标，前提是它们的最佳环境得到明确定义。表面沉积物校准（或训练）集广泛用于开发将物种分布与湖泊变量联系起来的定量传递函数。从跨越感兴趣变量的广泛梯度的大型数据集准确估计每个分类单元的环境最佳效益。在这里，我们收集了来自加拿大安大略省 450 多个湖泊的表面沉积物硅藻分析数据，以确定该省最常见类群的最佳生态环境。这些数据是从同一实验室使用类似的方法收集的，跨越了三十多年的研究（1987 年至 2019 年），对 546 个样本（代表 464 个独特地点）进行了研究。我们描述了 52 种常见硅藻类群对湖泊管理关键的三个湖泊学变量的生态反应：pH、总磷 (TP) 和最大湖泊深度。梯度很宽（pH：4.2–9.3；TP：1.1–96.8 µg L^-1 ; 最大深度：0.7–95.0 m），实现稳健的统计分析。我们使用分层逻辑回归模型来确定所有常见分类单元具有统计显着性的生态响应曲线，并使用加权平均来确定最佳值。在我们的数据集中，96% 的常见硅藻类群对 pH 值具有显着的生态响应和最佳值，81% 对 TP 具有显着的生态响应，83% 对深度具有最佳值。除了底栖世界性Achnanthidium minutissimum复合体（存在于 88.1% 的样本中）之外，最常记录的分类群是浮游生物，包括Discostella stelligera / pseudostelligera (83.2%)、Asterionella formosa (77.8%) 和Lindavia bodanica复杂（76.6%）。我们的结果说明了硅藻类群在关键湖泊学梯度上的反应，提供了重要的自生态学信息，以促进其在环境评估中作为生物指示剂的使用。这些数据可用于解决安大略省尚未获得校准集的未充分研究地区的管理和知识差距，并可支持具有类似湖泊环境和硅藻群落的地区的研究。