Predicting emergent animal biodiversity patterns across multiple scales.

Restoring biodiversity-based resilience and ecosystem multi-functionality needs to be informed by more accurate predictions of animal biodiversity responses to environmental change. Ecological models make a substantial contribution to this understanding, especially when they encode the biological mechanisms and processes that give rise to emergent patterns (population, community, ecosystem properties and dynamics). Here, a distinction between 'mechanistic' and 'process-based' ecological models is established to review existing approaches. Mechanistic and process-based ecological models have made key advances to understanding the structure, function and dynamics of animal biodiversity, but are typically designed to account for specific levels of biological organisation and spatiotemporal scales. Cross-scale ecological models, which predict emergent co-occurring biodiversity patterns at interacting scales of space, time and biological organisation, is a critical next step in predictive ecology. A way forward is to first capitalise on existing models to systematically evaluate the ability of scale-explicit mechanisms and processes to predict emergent patterns at alternative scales. Such model intercomparisons will reveal mechanism to process transitions across fine to broad scales, overcome approach-specific barriers to model realism or tractability and identify gaps which necessitate the development of new fundamental principles. Key challenges surrounding model complexity and uncertainty would need to be addressed, and while opportunities from big data can streamline the integration of multiple scale-explicit biodiversity patterns, ambitious cross-scale field studies are also needed. Crucially, overcoming cross-scale ecological modelling challenges would unite disparate fields of ecology with the common goal of improving the evidence-base to safeguard biodiversity and ecosystems under novel environmental change.

Temperature thresholds of ecosystem respiration at a global scale.

Ecosystem respiration is a major component of the global terrestrial carbon cycle and is strongly influenced by temperature. The global extent of the temperature-ecosystem respiration relationship, however, has not been fully explored. Here, we test linear and threshold models of ecosystem respiration across 210 globally distributed eddy covariance sites over an extensive temperature range. We find thresholds to the global temperature-ecosystem respiration relationship at high and low air temperatures and mid soil temperatures, which represent transitions in the temperature dependence and sensitivity of ecosystem respiration. Annual ecosystem respiration rates show a markedly reduced temperature dependence and sensitivity compared to half-hourly rates, and a single mid-temperature threshold for both air and soil temperature. Our study indicates a distinction in the influence of environmental factors, including temperature, on ecosystem respiration between latitudinal and climate gradients at short (half-hourly) and long (annual) timescales. Such climatological differences in the temperature sensitivity of ecosystem respiration have important consequences for the terrestrial net carbon sink under ongoing climate change.

Multiple environmental controls explain global patterns in soil animal communities.

Soil animals play important roles in ecosystem functioning and stability, but the environmental controls on their communities are not fully understood. In this study, we compiled a dataset of soil animal communities for which the abundance and body mass of multiple soil animal groups were recorded. The mass-abundance scaling relationships were then used to investigate multiple environmental controls on soil animal community composition. The data reveal latitudinal shifts from high abundances of small soil animals at high latitudes to greater relative abundances of large soil animals at low latitudes. A hierarchical linear mixed effects model was applied to reveal the environmental variables shaping these latitudinal trends. The final hierarchical model identified mean annual temperature, soil pH and soil organic carbon content as key environmental controls explaining global mass-abundance scaling relationships in soil animal communities (R2c = 0.828, Ngroup = 117). Such relationships between soil biota with climate and edaphic conditions have been previously identified for soil microbial, but not soil animal, communities at a global scale. More comprehensive global soil community datasets are needed to better understand the generality of these relationships over a broader range of global ecosystems and soil animal groups.

The influence of soil communities on the temperature sensitivity of soil respiration.

Soil respiration represents a major carbon flux between terrestrial ecosystems and the atmosphere, and is expected to accelerate under climate warming. Despite its importance in climate change forecasts, however, our understanding of the effects of temperature on soil respiration (RS) is incomplete. Using a metabolic ecology approach we link soil biota metabolism, community composition and heterotrophic activity to predict RS rates across five biomes. We find that accounting for the ecological mechanisms underpinning decomposition processes predicts climatological RS variations observed in an independent dataset (n = 312). The importance of community composition is evident because without it RS is substantially underestimated. With increasing temperature, we predict a latitudinal increase in RS temperature sensitivity, with Q10 values ranging between 2.33 ± 0.01 in tropical forests to 2.72 ± 0.03 in tundra. This global trend has been widely observed, but has not previously been linked to soil communities.