Chironomid-climate continentality conundrum

PLOS One | https://doi.org/10.1371/journal.pone.0327780 August 5, 2025 3 / 18 Larvae of chironomids, non-biting midges from the family Chironomidae, are recognised as one of the most reliable palaeoclimate proxies [30]. The taxonomic composition of subfossil chironomid assemblages is known to be responsive to environmental conditions [31,32], such as lake water trophic state [33,34], dissolved oxygen concentrations [35,36], pH [34,37], and depth [37,38]; warm season temperatures [39–42], and heat accumulation expressed as growing degree days (GDD; [43]). Chironomids are often assumed to be non-responsive directly to changes in winter temperatures as they experience diapause during the winter season in temperate and boreal climate zones [44]. However, several studies have shown a significant dependency of chironomid assemblages on winter temperatures [30,45]. An indirect impact of winter temperatures has been observed in several recent studies, showing that chironomid assemblages from boreal and temperate zones can be affected by the duration of ice cover, which is inversely correlated with dissolved oxygen levels and the warm season duration [24,46], as well as with continentality [47]. Self et al. [24] show that chironomid assem- blages in northern Russia are influenced by continentality (Gorzynski continentality index), which is thought to have an indirect effect through variations in ice-cover period length. It is commonly recommended to use chironomid training sets

only within the biogeographic area from which they originated [48,49]. Therefore, the training set developed by Self et al. [24] is applicable only in northern Russia. To expand our understanding of the relationship between chironomids and continentality in areas with transitional climates in northern and eastern Europe, we have assembled a new chironomid dataset that represents a wide range of climatic and environmental variations along a longitudinal continentality gradient (Fig 1) in northern and eastern Europe – from the oceanic Norwegian coast to the continental Ural Mountains. The objectives of our study are to determine (1) possible confounding factors to continentality environmental variables; (2) the potential influence of continentality and related climate variables on chironomid assemblages; and (3) the indicator taxa representative of different parts of the continentality gradient. This paper aims to serve as a prerequisite and justification for the increased use of chironomids as a continentality proxy and as a starting point for developing a more extensive training set applicable in northern and east- ern Europe. Data derived from continentality reconstructions will enhance our understanding of how continentality varies over time and how it impacts natural ecosystems. Materials and methods Climate data Hourly temperature (oC) and lake ice-thickness (mm) data for each lake were extracted from the ERA5 dataset with hourly temporal and 0.25° x 0.25°spatial resolution [50], which was downloaded from the Copernicus Climate Data Store. Using

the downloaded climatic data, the following variables were calculated based on a 30-year mean: (1) Mean January, April, July and October temperatures (oC). (2) Continentality indices: annual temperature range (ATR; the difference between the warmest month’s mean tempera- ture (July for all sites) and the coldest month’s mean temperature (January for all sites)); Kerner Oceanity Index (KOI; [4]), reflecting not only annual temperature variation, but also the warmth of spring and autumn, calculated following the equation: KOI = 100×(To-Ta)/ATR, where To is the mean October air temperature (oC), Ta is the mean April air tem- perature (oC), and ATR is the annual air temperature range (oC)). (3) Annual sum of Growing Degree Days at a base temperature of 5 °C (GDD5) was calculated by applying the daily tem- perature data following the equation [51]: ∑365 i=1 Tmin+Tmax 2 – Tbase, where Tbase equals 5 °C and i refers to day of the year. (4) Ice conditions (number of ice-cover days): the number of ice-cover days was estimated using the ice-thickness dataset by summing the number of days with a minimum ice thickness across the lake > 0 mm. Based on the above-listed datasets a set of thematic maps covering Northern and Eastern Europe was created. The dataset design strategy was guided by the longitudinal (east to west) continentality (ATR) gradient observed in northern Europe (Fig 1).