Phenological shifts enable species to adjust the timing of life-history events to required resources. Climate change alters the spatiotemporal association between temperature and photoperiod, modifying the scope for temperature regulation. Here, we propose a model hypothesizing that dissimilar light requirements and constraints differently impact phenological responses in diurnal versus nocturnal ectotherms. Next, we investigate temporal shifts and latitudinal trends in phenology for 363 Lepidoptera species using 44 years of citizen science data. In agreement with model predictions, long-term shifts in the estimated onset, peak, termination, and duration of the flight period differed qualitatively between diurnal and nocturnal species, even after accounting for voltinism and overwintering stage, supporting that diel activity is a key regulator of phenology. Phenology showed intraspecific latitudinal trends, and the peak occurred later in the north in diurnal species but was independent of latitude in nocturnal species. These contrasting phenological shifts may impact community composition and ecosystem functioning.

1. Movement is fundamental to population persistence and metapopulation dynamics, but robust comparative estimates of dispersal remain scarce for threatened butterflies. Limited quantitative data on movement constrain effective conservation network design.

2. We focused on three threatened butterfly species-Euphydryas aurinia, Parnassius apollo and Phengaris arion-co-occurring on Gotland, Sweden, which together provide an ideal model for comparative dispersal analysis.

3. To quantify and compare movement patterns among species using extended capture-mark-recapture (CMR) data, test for random-walk behaviour, and identify the best-fitting dispersal kernels.

4. CMR datasets for E. aurinia and P. apollo were extended to 2024 and combined with comprehensive data for P. arion, yielding 9670 net-displacement observations (the distance between the first and last captures) collected from 2017 to 2024. One movement value per individual was used to avoid pseudoreplication, detection probabilities were estimated with Cormack-Jolly-Seber models, and four dispersal kernels were evaluated.

5. Median net displacement differed significantly among species (chi(2) = 450.14, p < 0.001): E. aurinia showed the lowest value (0.135 km), while P. apollo (0.253 km) and P. arion (0.252 km) were similar. Movements deviated from random-walk expectations (log-log slope = 0.431 versus 0.5 expected, p < 0.001), indicating area-restricted movement. Lognormal kernels best described E. aurinia and P. apollo, whereas an exponential distribution fitted P. arion best, with maximum displacements of 8.19, 10.69 and 4.31 km, respectively.

6. Even butterflies traditionally regarded as sedentary exhibit substantial dispersal capacity. Species-specific movement strategies influence metapopulation connectivity, and the derived parameters provide essential inputs for designing habitat networks within each species's dispersal range.

Conservation planning typically relies on species occurrence records, yet where individuals concentrate within occupied landscapes may matter more for population persistence than where species are merely detected. We applied Lorenz curves and Gini coefficients, metrics originally developed to quantify wealth inequality, to spatially referenced abundance data for three threatened grassland butterflies protected under the EU Habitats Directive: Euphydryas aurinia, Parnassius apollo, and Phengaris arion. Across a 60 km2 landscape on Gotland, Sweden, subdivided into 1-ha grid cells (795 species & times; grid observations), peak densities were strongly right-skewed and spatially concentrated for all three species. The top 19% of occupied grids accounted for 44-53% of the summed peak counts, and capturing 50% of the summed peak counts required only 20-25% of occupied grids. Phengaris arion was the most spatially concentrated species, and P. apollo the least, a ranking that was robust across Gini coefficients and concentration metrics. Habitat associations were species-specific: ground moisture and grassland proportion were positively associated with peak density for E. aurinia and grassland proportion for P. apollo, whereas Ph. arion peak densities declined with increasing grassland proportion and were poorly predicted by the landscape-scale habitat metrics examined. These findings demonstrate that high-density supersites can be difficult to identify from occurrence-based assessments, yet may represent important targets for population-focused conservation. The Lorenz-Gini framework provides a transferable, metric-based approach for identifying such sites across taxa, complementing area-based conservation targets by explicitly considering where populations actually persist.

Nature reserves are expected to recreate old-growth conditions through passive succession, yet long-term, stand-wide tests are rare. We studied a hemiboreal forest in south-eastern Sweden, which has been protected since 1923. We resurveyed every tree within 10.4 ha. 88 years after the initial 1937 census, to investigate whether protection alone restores structure without eroding compositional diversity. All stems present in 1937 (≥ 8.89 cm diameter at breast height, DBH) were relocated and matched to their 2025 status; all living recruits ≥ 1 cm DBH were mapped. Living stem density declined 22 % to 412 ha-1, yet basal area rose 6 % to 40.7 m² ha^-1 and living volume 12 % to 451 m³ ha^-1. Deadwood accumulated to 190 m³ ha^-1, and large living trees (≥ 45 cm DBH) increased 20 % to 109 ha^-1. Tree species richness increased from five to thirteen, but stand dominance shifted towards shade-tolerant Norway spruce, now 62 % of stems; Scots pine density fell 40 %. Neighbourhood basal area strongly enhanced survival (odds ratio 7.8), especially for pine, whereas wetter microsites reduced survival. Recruitment (6276 stems) was 90 % spruce and concentrated where the 1937 stand had been densest; pine recruitment declined with increasing moisture. In this small reserve embedded in a managed forest landscape, strict protection rebuilt structural capital but shifted composition towards spruce dominance. Natural disturbances (windthrow, beetles) generate canopy gaps that maintain structural heterogeneity, yet outcomes are constrained by reserve size and landscape context rather than representing universal expectations for protected forests.

Smaller insect species are routinely described later than larger species, perpetuating the Linnean shortfall. Using 425 saproxylic beetle species from 28 veteran oak sites, we test whether body size varied with year of description and earliest georeferenced record year, and whether body length predicted detection probability. Linear models show a significant negative relationship between body length and both year of description (β = -0.00063 +/- 0.00021 SE, p = 0.003) and earliest georeferenced year (β = -0.00057 +/- 0.00024 SE, p = 0.016). Single-season occupancy models reveal only a marginal size effect on detection (β = -0.021, p = 0.055); instead, the trophic group drives variation, with sapro-xylophagous species detected most readily. Fourteen species remain undescribed, and 29% lack any georeferenced record, highlighting combined Linnean and Wallacean shortfalls. Historical taxonomic bias towards conspicuous species and methodological undersampling of specific trophic guilds together impede accurate biodiversity assessment. We recommend integrated taxonomic revision, diversified trapping protocols and trait-based occupancy models to rectify these biases and improve conservation planning.

Global forest ecosystems are experiencing increasingly frequent and severe insect outbreaks, driven by complex interactions among climate change, land-use alterations, and shifting species distributions. Species that are morphologically indistinguishable - often referred to as cryptic species - can differ significantly in distribution, host use, and susceptibility to natural enemies and might thereby differ in outbreak dynamics. Yet, the contribution of cryptic species to temporal changes in the frequency and severity of insect outbreak dynamics remains poorly understood. Motivated by recent defoliation events in northern European oak forests, we investigated an emerging leaf-miner outbreak in Sweden. Through targeted surveys, rearing from 22 sites and Malaise trapping at 34 sites (56 sites total), we identified a pronounced spatial clustering of outbreaks at higher latitudes. The newly recognised cryptic species Acrocercops andreneli was strongly associated with these outbreaks, whereas sites with only Acrocercops brongniardella never showed outbreaks. Host-parasitoid networks related to the two cryptic moth species were strikingly similar. Our findings demonstrate the importance of cryptic species for outbreak dynamics and their consequences for host plant health in ways that are easily overlooked by traditional taxonomy. Moreover, such outbreak dynamics cannot always be linked to a lack of top-down control by natural enemies.

Theory, manipulation experiments and observational studies on biodiversity and ecosystem functioning largely concur that higher intraspecific diversity may increase the overall productivity of populations, buffer against environmental change and stabilize long-term productivity. However, evidence comes primarily from small and short-lived organisms. We tested for effects of genetic diversity on variation in forest growth by combining long-term data on annual individual growth rate (basal area increment (BAI)) with estimates of intrapopulation genetic variation (based on RAD-seq SNPs) for 18 natural Quercus robur pedunculate oak populations. Higher total or adaptive genetic variability of populations was neither associated with faster average growth nor with increased temporal or spatial stability of growth nor with among-individual asynchrony in growth. However, as expected, we found that greater asynchrony of growth responses within the populations increased their temporal stability. Together, these findings point towards a negligible role of genetic variation in structuring growth patterns in natural populations of tree species. Identifying which environmental factors and phenotypic traits (and its genetic basis) contribute to asynchronous growth responses is an important next step towards a better mechanistic understanding of the causes of temporal stability in tree growth and forest productivity.

Introducing non-native tree species into forest ecosystems is a growing trend, in part as climate change may cause a decline of native species and shifts in species distributions. In European forestry, Quercus rubra (northern red oak) has increasingly been considered a candidate substitute species for native oaks. However, it remains largely unknown how this substitution affects associated biodiversity. This study compares the biodiversity supported by the native oak species Q. petraea (sessile oak) and Q. robur (pedunculate oak) and the invasive Q. rubra in southern Sweden, focusing on both oak-associated organisms and general forest biodiversity. Arthropods were sampled using Malaise traps at the site level. At the same time, vascular plants, leaf herbivory and endophytic insects (leaf miners and gallers) were recorded at the tree level in three sites per oak species. Our results reveal guild-specific effects of oak species on biodiversity. The introduced Q. rubra supported significantly fewer endophytic insects than native oak species. Vascular plant species richness was marginally lower in Q. petraea and Q. rubra sites compared to Q. robur. In contrast, the species richness, abundance, biomass and community composition of arthropods and leaf herbivory did not differ significantly between the three oak species. These findings indicate that the ecological consequences for biodiversity of introduced tree species, such as Q. rubra, are most pronounced for specialised herbivores, including leaf miners and gallers, and suggest that broader forest biodiversity measures may be less responsive to changes in tree species than to local environmental conditions.

This study evaluates five sampling methods for characterising saproxylic beetle assemblages in a recently burned (18-21 June 2021) boreal forest in southeastern Sweden. We compared species richness and community composition in samples collected using trunk traps, flight-intercept traps, pheromone traps, Malaise traps, and manual searches, deployed in the Finsjöbrännan nature reserve between 2022 and 2024. A total of 2258 beetle specimens were collected, representing 559 species, including 36 red-listed taxa. Flight-intercept traps yielded the highest species richness (331 species) and abundance, while trunk traps were most effective at detecting red-listed species (13.4% of species captured by trunk traps were red-listed). Coverage-based rarefaction indicated that deploying trunk and flight-intercept traps together could detect approximately 89% of the beetle species (by extrapolated richness); however, this two-method combination necessarily spans different succession stages. Notably, 61.1% of the red-listed species were captured exclusively by a single trap type, underscoring the complementary nature of different sampling approaches. Although non-overlapping sampling years may have confounded the comparisons between trap types, these results suggested that methodological choices, combined with temporal factors, may strongly influence biodiversity assessments in post-fire habitats. We provide evidence-based recommendations for effective sampling protocols, emphasising the need for multi-method approaches and temporal consistency in conservation-focused monitoring. Implications for insect conservation: Post-fire forests represent critical habitats for numerous saproxylic beetles, including many species of conservation concern that depend on recently burned wood. Our results demonstrate that no single sampling method adequately captures the full diversity of post-fire beetle assemblages, as red-listed species often occur exclusively in one type of trap. Therefore, conservation monitoring and biodiversity inventories in burned forests should employ a combination of complementary methods-particularly flight-intercept and trunk traps-to maximise species detection. Standardising temporal deployment and ensuring sampling continuity across years are essential for reliable comparisons among fires and regions. Incorporating such multi-method protocols into long-term monitoring will strengthen assessments of post-disturbance recovery and support evidence-based management of fire-dependent insect communities.

Fermentative yeasts play important roles in both ecological and industrial processes, but their distribution and abundance in natural environments are not well understood. We investigated the diversity of yeasts at the northern range limit of their oak tree hosts (Quercus spp.) in Sweden, and identified climatic and ecological conditions governing their distribution. Yeasts were isolated from bark samples from 28 forests and identified to the species level using DNA metabarcoding. Most communities were dominated by species in the Saccharomycetaceae family, especially by species of Saccharomyces, Kluyveromyces and Pichia. Each genus showed a distinct latitudinal and longitudinal distribution, and both temperature and precipitation metrics predicted significant variation in their abundance. Consistent with this, laboratory assays revealed significant effects of temperature on the growth of strains collected from different longitudes and latitudes. We found that older trees harbour more diverse and more balanced fermentative yeast communities with more evenly distributed species abundances. Communities across trees were more similar when sharing a common dominant species. This work provides a baseline for future studies on the impact of climate change on the fermentative yeast biodiversity of temperate forests in northern latitudes and contributes to a growing collection of wild isolates for potential biotechnological applications.