<jats:title>Abstract</jats:title><jats:p>Lakes cover large parts of the climatically sensitive Arctic landscape and respond rapidly to environmental change. Arctic lakes have different origins and include the predominant thermokarst lakes, which are small, young and highly dynamic, as well as large, old and stable glacial lakes. Freshwater diatoms dominate the primary producer community in these lakes and can be used to detect biotic responses to climate and environmental change. We used specific diatom metabarcoding on sedimentary DNA, combined with next-generation sequencing and diatom morphology, to assess diatom diversity in five glacial and 15 thermokarst lakes within the easternmost expanse of the Siberian treeline ecotone in Chukotka, Russia. We obtained 163 verified diatom sequence types and identified 176 diatom species morphologically. Although there were large differences in taxonomic assignment using the two approaches, they showed similar high abundances and diversity of Fragilariceae and Aulacoseiraceae. In particular, the genetic approach detected hidden within-lake variations of fragilarioids in glacial lakes and dominance of centric <jats:italic>Aulacoseira</jats:italic> species, whereas <jats:italic>Lindavia ocellata</jats:italic> was predominant using morphology. In thermokarst lakes, sequence types and valve counts also detected high diversity of Fragilariaceae, which followed the vegetation gradient along the treeline. Ordination analyses of the genetic data from glacial and thermokarst lakes suggest that concentrations of sulfate (SO<jats:sub>4</jats:sub><jats:sup>2−</jats:sup>), an indicator of the activity of sulfate-reducing microbes under anoxic conditions, and bicarbonate (HCO<jats:sub>3</jats:sub><jats:sup>−</jats:sup>), which relates to surrounding vegetation, have a significant influence on diatom community composition. For thermokarst lakes, we also identified lake depth as an important variable, but SO<jats:sub>4</jats:sub><jats:sup>2−</jats:sup> best explains diatom diversity derived from genetic data, whereas HCO<jats:sub>3</jats:sub><jats:sup>−</jats:sup> best explains the data from valve counts. Higher diatom diversity was detected in glacial lakes, most likely related to greater lake age and different edaphic settings, which gave rise to diversification and endemism. In contrast, small, dynamic thermokarst lakes are inhabited by stress-tolerant fragilarioids and are related to different vegetation types along the treeline ecotone. Our study demonstrated that genetic investigations of lake sediments can be used to interpret climate and environmental responses of diatoms. It also showed how lake type affects diatom diversity, and that such genetic analyses can be used to track diatom community changes under ongoing warming in the Arctic.</jats:p>