Changes in biodiversity impact atmospheric chemistry and climate through plant volatiles and particles

The study addresses how biodiversity changes influence atmospheric chemistry and climate via plant-emitted biogenic volatile organic compounds (BVOCs) and their oxidation products forming biogenic secondary organic aerosols (BSOA). The biosphere and atmosphere are tightly coupled; plants emit species-specific BVOCs that respond to abiotic (heat, drought, radiation) and biotic (herbivory, pathogens) stress and affect atmospheric processes, radiative balance, and cloud formation. The authors propose three hypotheses along tree diversity gradients: H1 (productivity): more diverse forests, with higher stand productivity and leaf area, should emit higher amounts of BVOCs; H2 (abiotic stress amelioration): diversity reduces abiotic stress via facilitation and microclimate buffering, decreasing BVOC emissions; H3 (biotic stress reduction): diversity reduces per-capita herbivory and pathogen pressure (top-down control, non-host effects), decreasing BVOC emissions, though potentially increasing compound diversity. They highlight a critical knowledge gap: integrated, concurrent measurements of BVOCs and BSOA along tree diversity gradients are lacking, despite their relevance to climate-biogeochemical feedbacks. The purpose is to present a conceptual framework and provide first empirical tests using a tree diversity experiment to evaluate biodiversity effects on BVOC emissions and BSOA formation.

The paper synthesizes literature on: (1) BVOC emissions and their drivers—species-specificity of emissions (isoprene vs monoterpenes), production pathways, responses to abiotic (heat, drought, radiation) and biotic stress, and roles in plant communication and defense; (2) biodiversity-ecosystem functioning—resource partitioning, structural complexity, and enhanced productivity and leaf area in mixtures, microclimate buffering, and reduced sensitivity to climate extremes; (3) biotic interactions—diversity effects on herbivory and pathogens through enemy control, host dilution, and trait heterogeneity that reduce herbivore performance; (4) atmospheric chemistry—oxidation of BVOCs by OH, O3, and NO3 to oxygenated organic molecules that form BSOA, and the consequences for radiation (direct effects) and cloud-precipitation processes (indirect effects), along with dependencies on oxidant regimes, temperature, kinetics, and aqueous/particle-phase chemistry; (5) climate feedbacks—warming-driven increases in BVOC and aerosol number concentrations potentially moderating climate via cloud interactions; (6) limited prior work measuring BVOCs and BSOA across diversity gradients, establishing the need for integrated, multidisciplinary studies.

Study site and design: Conducted at the MyDiv tree diversity experiment (Saxony-Anhalt, Germany). From September 21 to October 13, 2021 (13 days), ten plots were sampled among eighty 11 × 11 m plots. Treatments included monocultures (Acer pseudoplatanus L., Fraxinus excelsior L., Prunus avium (L.) L., Sorbus aucuparia L.), three two-species mixtures (A. pseudoplatanus + F. excelsior; A. pseudoplatanus + P. avium; P. avium + S. aucuparia), and one four-species mixture (all four species). Sampling scheme: per day, two monocultures, one two-species mixture, and one four-species mixture were sampled for at least four consecutive non-rainy days. Measurement setup: A custom aluminum sampling box per plot at canopy-top contained gas-phase (BVOCs) and particulate matter (BSOA) sampling systems. Gas-phase BVOCs were sampled through a 2-cm tube with glass frit onto Carbotrap 300 and Tenax cartridges using a GilAir Plus pump at 150 mL min−1. Particulate matter was collected with a PM10 PEM impactor connected to a Gillian 12 pump at 10 L min−1. Offline sampling duration was 4 hours (10:00–14:00). Cartridges were sealed, stored at 4 °C, analyzed within 3–5 days; quartz filters were stored at −18 °C after preheating (105 °C, 24 h). Field blanks were <10% of ambient values. A background high-volume sampler could not be used for direct background correction due to differing sampling characteristics. Analytical chemistry: BVOCs were quantified by thermodesorption GC/MS (Perkin Elmer TD; Agilent GC/MS; ZB-5ms column). Thermal desorption: 7 °C min−1 to 220 °C at 50 mL min−1; cryofocus at −30 °C; transfer at 220 °C (3 min) with 1 mL min−1. GC temperature program: 35 °C (5 min), ramp 5 °C min−1 to 120 °C, then 20 °C min−1 to 320 °C (12 min hold). External calibration (six levels, 0.5–20 ng per tube) with triplicates; RSD 2–10%; LOD at S/N ≥ 3. BVOCs identified: α-pinene, β-pinene, camphene, 3-carene, p-cymene, limonene, α-terpinene, isophorone, acetophenone. BSOA constituents were extracted from filter pieces with 1 mL water:acetonitrile (1:1), shaken 30 min at 1000 min−1, syringe-filtered (0.2 μm wwPTFE), and analyzed by UHPLC (C18, Waters Acquity HSST3, 2.1 × 100 mm, 1.8 μm) with eluents (A) water + 0.1% formic acid, (B) acetonitrile + 0.1% formic acid; flow 0.3 mL min−1; gradient: 5% B 0–1.0 min, to 100% B at 16.0 min, hold to 18.0 min, return to 5% B at 18.1 min, end 21.0 min. Detection by Orbitrap MS (Q Exactive Plus) with ESI− at 3.5 kV, full scan m/z 50–750, R=70k at m/z 200; water blanks and QC standards every tenth sample. External calibration range 0.024–400 μg L−1 with LOD at S/N ≥ 3. BSOA compounds quantified: diaterpenylic acid acetate (DTAA), 3-methyl-1,2,3-butanetricarboxylic acid (MBTCA), norpinonic acid, pinonic acid, terebic acid, terpenylic acid, pinic acid, adipic acid, pimelic acid, azelaic acid, suberic acid, succinic acid, glutaric acid, salicylic acid, sebacic acid. Ancillary data: Meteorology (air temperature, humidity, pressure, wind speed/direction, precipitation, UV radiation) at canopy top; overall NO3 and O3 concentrations. Tree wood volume (2015, 2021) and annual wood productivity derived from plot inventories. Statistical analyses: One-way ANOVA to compare observed BVOC and BSOA among biodiversity treatments; Tukey’s HSD for post hoc comparisons. Pearson correlations between BSOA and products of α-/β-pinene with O3 and UV, analyzed per week. Expected mixture amounts computed from monoculture observations divided by mixture richness. Two-sample t-tests compared observed vs expected in mixtures. Effect sizes computed as standardized mean difference (Cohen’s d variant): SMD = (Mean_observed − Mean_expected)/sqrt((SD_observed^2 + SD_expected^2)/2), weighted by inverse variance, pooled across compounds. Separate SMD analyses for overall mixtures, two-species mixtures, and four-species mixtures. Visualized with forest plots. All analyses performed in R v4.1.3.

- BVOCs: Nine BVOCs were quantified (α-pinene, camphene, β-pinene, 3-carene, p-cymene, limonene, α-terpinene, isophorone, acetophenone). Most compounds showed no significant differences between monocultures and mixtures, with exceptions: limonene (ANOVA p = 0.01) and acetophenone (p < 0.01) were higher in monocultures, notably in Sorbus aucuparia plots. - Observed vs expected in mixtures: In four-species mixtures, observed limonene and acetophenone were significantly lower than expected (t-test p = 0.04 and p < 0.01, respectively), indicating suppression in diverse mixtures. - Pooled BVOC effect: Standardized mean difference (SMD) analysis across compounds showed a significant overall decrease of BVOC concentrations with increasing diversity (pooled effect SMD 95% CI approximately [−0.64; −0.27], z = −4.85, p < 0.01). Limonene showed a significant negative SMD (p < 0.01); β-pinene, p-cymene, α-terpinene, and acetophenone were marginal (p = 0.09, 0.08, 0.06, and 0.06, respectively). The decreasing trend held for both two- and four-species mixtures (overall SMD p < 0.01), with limonene significant in both (p = 0.05 and 0.03), and acetophenone significant in four-species mixtures (p < 0.01). - BVOCs vs productivity: Limonene, isophorone, and acetophenone were negatively associated with annual wood productivity (significant correlations), contradicting the H1 expectation that higher productivity would increase BVOCs. - BSOA: Fifteen BSOA compounds were quantified. No significant differences among treatments; tendencies included higher succinic and sebacic acids in monocultures, and higher pinonic and terebic acids in mixtures. - Observed vs expected BSOA: Mixed, compound-specific deviations without overall significance; certain two-species mixtures produced more norpinonic and glutaric acid than expected. - Pooled BSOA effect: Overall SMD across BSOA compounds was non-significant (pooled effect 95% CI roughly [−0.22; 0.06], p = 0.26). Two-species mixtures showed a weakly significant overall effect (p = 0.05), while four-species mixtures were non-significant (p = 0.89). - Interpretation: Decreases in BVOCs with diversity support H2 (abiotic stress amelioration) and H3 (reduced biotic stress). Mixed BSOA responses likely reflect atmospheric transport, oxidant availability (OH, O3, NO3), temperature/humidity effects, and differing gas-phase mechanisms for α- vs β-pinene oxidation.

The findings provide initial empirical support for the conceptual framework linking biodiversity to atmospheric chemistry. BVOC emissions decreased with increasing tree diversity, aligning with H2 and H3 rather than H1. Mechanistically, diverse stands likely reduce abiotic stress via resource complementarity and microclimate buffering (cooler, moister canopies and diversified rooting depth) that diminish stress-induced BVOC emissions. Biotic mechanisms include reduced herbivory and pathogen pressure through increased predator presence, host dilution, and heterogeneous defense traits, all lowering stress-related emissions. The lack of a clear, consistent BSOA response across diversity levels reflects the complexity of atmospheric processing: BVOC-to-BSOA conversion depends on oxidant regimes (OH, O3, NO3), temperature, humidity, and gas-/particle-phase chemistry and timing/transport of air masses. Compound-specific differences (e.g., MBTCA vs pinic/pinonic acid) are consistent with known oxidation pathways for α- and β-pinene. Climatically, BSOA alter radiation via scattering/absorption and modulate cloud droplet number concentrations and precipitation processes; thus biodiversity-driven changes in BVOC composition and quantity could influence regional radiative forcing and cloud-precipitation dynamics. However, given mixed BSOA results and limited sampling, stronger inference requires expanded datasets and integrated atmospheric measurements and modeling.

This study introduces a conceptual framework and provides a first empirical test indicating that increasing tree diversity can reduce BVOC emissions, while effects on BSOA are mixed and overall non-significant in this initial dataset. The work underscores that biodiversity can shape biosphere-atmosphere feedbacks relevant to climate via plant volatiles and aerosol formation. The authors caution against overgeneralization from this limited case study and recommend comprehensive, multidisciplinary research: detailed microclimate monitoring; quantification and manipulation of biotic (herbivores, pathogens, soil microbiota) and abiotic stresses; accounting for phenological seasonality of photosynthesis; and integration of local/regional models with coordinated field and chamber experiments across long-term biodiversity gradients. Such efforts will clarify how biodiversity alters BVOC emissions, BSOA formation, and their climatic impacts.

- Limited sample size and duration: Only ten plots measured over 13 days in late growing season; restricted temporal coverage and statistical power. - Background characterization: High-volume background sampler differed from plot samplers, limiting discrimination of local vs background BSOA. - Spatial/transport confounding: Regional air mass influence and time needed for BVOC-to-BSOA conversion hinder clear local coupling. - Non-manipulative stress design: Abiotic and biotic stressors were not experimentally manipulated; microclimate and stress metrics were not measured in depth, constraining mechanistic attribution. - Compound coverage: BVOC measurements covered nine compounds, not full BVOC spectrum; BSOA analysis limited to 15 marker compounds. - Productivity linkage: Correlations with productivity are observational and may be influenced by unmeasured covariates. - Generalizability: Single site, specific species set and region; findings may not extrapolate to other forest types, seasons, or climates.