The study focused on Volkhov, home to an aluminium smelter launched in 1932 and later replaced by a fertiliser plant, and Syasstroy, where a pulp and paper mill has operated since 1928. Both enterprises were designated regional pollution hot spots in the early 1990s by the Baltic Marine Environment Protection Commission. Volkhov’s emissions were dominated by sulphur dioxide and fluoride compounds, while Syasstroy released dust, sulphur dioxide, and nitrogen oxides. By the mid-1950s, peak sulphur dioxide concentrations in Volkhov had reached roughly 750 times the currently recommended health-protective level, and local ecosystems near Syasstroy were classified as critical or even catastrophic in the 1990s.

Over the study period, emissions fell dramatically. Sulphur dioxide output at Volkhov dropped fourfold between 1988 and 2021, fluorine emissions fell fifteenfold in the early 1990s, and dust emissions were about fifty percent lower in the 2020s than in the late 1980s. At Syasstroy, dust and sulphur dioxide emissions plummeted a hundredfold in the 1990s following an abrupt production decline. The researchers documented corresponding environmental recovery: fluoride concentrations in birch leaves near Volkhov fell from a peak of 200 micrograms per gram in 1988 to 32 micrograms per gram in 2025, and the needle longevity of Norway spruce, a reliable bioindicator of pollution stress, rebounded from roughly half of background values in 1989 to normal or even above-normal levels by 2025.

Against this backdrop of dramatic environmental change, the fate of the insects was surprisingly static. Between 1988 and 2025, the team counted more than 70,000 observation units of birch-feeding herbivores, representing over 60 species grouped into 24 recognisable taxonomic units, across 27 study sites and 9,674 sampled birch trees and branches. The community included external defoliators such as geometrid moths and sawflies, leafminers from several moth families, and leafrollers. Yet when the researchers correlated insect densities with distance from the polluters across all taxonomic-unit-by-year combinations, only 8.1 percent of 383 correlations were statistically significant, barely exceeding the 5 percent expected by chance alone.

A meta-analysis of the spatial patterns revealed that only six of the twenty analysed taxonomic units increased in density near the polluters, while the remaining fourteen showed no association with pollution at all. This variation was not explained by the insects’ abundance or by their feeding guild. During the high-emission years of 1988 to 1996, overall herbivore abundance did increase with proximity to the pollution sources, but the effect size was a mere 0.10, which is considered small by conventional statistical standards. Remarkably, during the low-emission period from 1997 to 2025, abundance continued to rise near the polluters at the same rate as before, even as the emissions that supposedly caused the pattern had largely vanished.

The persistence of elevated insect densities near the towns after emissions collapsed raises a thorny question: if pollution is not driving the pattern, what is? The authors considered urbanisation as an alternative explanation, but the towns are small, and existing meta-analyses suggest urbanisation has no significant overall effect on insect herbivore abundance. The results also cast doubt on earlier syntheses. A previous meta-analysis had reported elevated herbivore abundances near industrial sites, but the authors of the new study point out that mean effect sizes based on randomly chosen years from multiyear datasets did not differ from zero, whereas single-year studies and pseudoreplicated designs produced larger effects. In other words, earlier work may have systematically overestimated the true impact of pollution on herbivorous insects.

Perhaps the most counterintuitive finding concerns plant quality and leaf damage. Larvae of the autumnal moth, Epirrita autumnata, grew 45 percent faster on birch leaves collected in 2017 from sites closest to the polluters compared with leaves from the most distant sites, with the effect reaching 78 percent near Volkhov. Yet the density of this moth did not change with pollution level, and overall foliar damage by insects near both towns was approximately half of that recorded at unpolluted sites. This decoupling between plant quality, herbivore abundance, and plant damage echoes previous reports that improved host plant quality does not necessarily translate into more insects or more defoliation, because better foliage may simply allow insects to eat less while growing faster.

The study also speaks directly to the heated debate over global insect decline. Contrary to predictions, the researchers found no evidence that birch-feeding insects declined over the 38 years, either in polluted or unpolluted sites. Overall abundance remained flat, while diversity increased substantially at both polluted and background sites. The compositional shifts that did occur were guild-specific: leafminer densities decreased over time, defoliators remained stable, and leafrollers tended to increase, a pattern consistent with data from a copper-nickel smelter at Monchegorsk some 900 kilometres to the north. The rise in diversity is most plausibly explained by climate warming, with mean ambient temperature in the St. Petersburg region climbing from 5.7 to 6.7 degrees Celsius between 1985 and 2024, allowing southern species to expand northward.

The authors caution that their findings carry important limitations. The temporal resolution of sampling was uneven across years and taxa, emission and deposition data are incomplete for earlier decades, and the observational design cannot fully disentangle the effects of pollution decline from co-occurring drivers such as climate warming. The use of recognisable taxonomic units rather than species-level identifications may also mask responses of individual species that differ in pollution sensitivity. Nevertheless, the sheer duration and spatial extent of the dataset, with roughly 85 percent of the data collected by the two authors themselves, minimises the observer biases that plague shorter studies.

The broader implications are twofold. First, the study exposes a critical weakness in bioindication: conclusions about ecosystem health depend heavily on which indicator is chosen. Pollution gradients were clearly reflected in fluoride accumulation and spruce vigour, but insect density and diversity told a different and far noisier story, and a researcher relying on a single taxon in a single year could have concluded almost anything. Second, the work challenges the generality of the view that insects are universally sensitive to environmental change and questions the role of moderate industrial pollution as a driver of insect decline, at least for herbivores. The authors describe several of their findings as ecological surprises, unexpected observations that force a re-evaluation of established views. In an era when alarming insect decline headlines dominate the news, this nearly four-decade record from the boreal forests of northwestern Russia offers a sobering reminder that nature’s responses to human pressures are rarely as simple as we predict.

Subject of Research: Long-term effects of industrial air pollution and emission decline on birch-feeding insect communities in northwestern Russia

Article Title: Weak and inconsistent associations between birch-feeding insects and pollution levels around two industrial towns during substantial emission decline (1988–2025)

Article References: Kozlov, M. V., & Zverev, V. (2026). Weak and inconsistent associations between birch-feeding insects and pollution levels around two industrial towns during substantial emission decline (1988–2025). Environmental Monitoring and Assessment, 198(11), Article 1157. https://doi.org/10.1007/s10661-026-15971-1

Image Credits: AI Generated

DOI: 10.1007/s10661-026-15971-1

Keywords: industrial pollution, insect herbivores, birch, emission decline, ecosystem recovery, bioindicators, insect decline, leafminers, Epirrita autumnata, boreal forest, Norway spruce, long-term monitoring