Long-term impacts of elevated carbon dioxide and grazing on grassland soil microbiomes

<p><strong>Grazed grassland represents the most extensive land use globally, covering approximately 26% of the Earth’s terrestrial surface. These ecosystems provide essential services and play a pivotal role in global biogeochemical cycles, particularly carbon (C) and nitrogen (N) cycling, with substantial implications for global C pools under changing environmental conditions. Increasing atmospheric carbon dioxide (CO₂), as the primary agent of global change, can alter soil microbiomes with consequences for these soil biogeochemical processes. The response of soil microbial communities to global change drivers, such as rising atmospheric CO₂ and intensified grazing practices, will determine the stability, productivity, and resilience of these landscapes. While numerous studies have explored the sensitivity of soil microbial communities to global change factors, there remains a critical gap in our understanding of how these communities respond to the interactive influences of elevated CO₂ (eCO₂) and grazing practices. Addressing this gap is vital for predicting and managing the future functioning of grasslands under global change. Here I utilised a 23-year Free Air Carbon Dioxide Enrichment (NZFACE) experiment incorporating sheep grazing to disentangle the independent and interactive impacts of eCO₂ and grazing on 1) soil biogeochemical properties and soil microbial community abundance and structure, 2) microbial co-occurrence networks, and 3) the abundance of microbial N-cycling functional genes. Bacterial (16S rRNA gene) and fungal (ITS region) sequencing revealed that both eCO₂ and grazing increased bacterial diversity at the expense of fungal diversity. Grazing without excrement drove an increase in oligotrophic taxa linked to decreases in cation exchange capacity, pH, and OlsenP. Conversely, grazing with excrement return lowered soil C:N and near 3-fold increase in ammonium, while eCO₂ significantly decreased OlsenP and C:N. Microbial phospho- and neutral lipid quantification indicated significant interactive effects of eCO₂ and the grazing with excrement return treatments, increasing the biomass of Actinobacteriota, Gram-positive and Gram-negative bacteria, total bacteria, and total microbial biomass. As expected this was driven by N inputs and the maintenance of phosphorous (P) levels under grazing with excrement return alleviating the stoichiometric constraints introduced under eCO₂ by increased labile C inputs. These results highlight the pivotal role of grazing in shaping soil properties and microbial communities in these systems. Microbial co-occurrence network topology revealed eCO₂ increases network complexity, reflected by higher edge count, connectivity, and geodesic efficiency. However, this complexity comes at the cost of reduced modularity, indicating homogenisation of microbial niches and potentially increased vulnerability to disturbance. In contrast, grazing particularly with excrement return, increased network modularity, but reduces overall network size, connectivity, and the prevalence of keystone taxa, thereby increasing functional redundancy, but fragmenting microbial networks potentially weakening ecosystem resilience. Notably, the combined eCO₂ and grazing with excrement return treatment produced an intermediate network, partially offsetting grazing induced fragmentation, and eCO₂ induced hyperconnectivity. The connectivity of nitrogen cycling associated bacteria was greatest under the interactive treatment, suggesting increased microbial reliance on N-cycling functions with the large influx of nutrients from the combination of eCO₂ and nutrient return from grazers. These results show the complex interactions of eCO₂ and grazing practices on microbial interactions, highlighting the importance of N-cycling bacteria for microbial network stability. Quantification of nitrogen cycling genes associated with archaeal and bacterial ammonia oxidisation (amoA), nitrite reduction (nirK,nirS), and nitrous oxide reduction (nosZI, nosZII) revealed that grazing with no excrement return elicited significant decreases in abundance of all bacterial functional genes archaeal amoA, while the return of excrement resulted in no significant changes in abundance. Elevated CO₂ treatments induced an increase in nirK genes and a decrease in nosZI genes, causing a denitrification ‘bottleneck’ whereby nitrogen accumulates in the form of N₂O, likely increasing N₂O emissions. Interestingly, grazing with excrement return mitigated nosZI suppression under eCO₂, likely by stabilising soil pH through NH₄⁺ influx evidenced by strong positive correlations between nosZI abundance and soil pH. This suggests that grazing activities can potentially mitigate the accumulation and emissions of N₂O under eCO₂, highlighting the critical role of grazing management in modulating functional microbial responses. These findings underscore the dual role of excrement return in mitigating CO₂ driven stoichiometric imbalances and buffering soil degradation. Under eCO₂, grazing and excrement return stabilised microbial networks and functional gene abundance, counteracting functional niche homogenisation and a denitrification bottleneck. However, all treatments induced microbial network restructuring, raising concerns for ecosystem resilience, and potential risks soil fertility and N₂O emissions. Ultimately, significant interactive effects suggest pasture management approaches must consider the complex interplay between land use and global change factors, as treatments often produced non-additive outcomes that challenge the validity of single factor-experiments as proxies for real world grassland responses.</strong></p>