In impact assessment the spatial and temporal scales are part of the system boundaries, i.e. part of the frame within which effects are analysed. More precisely, define where and when impacts need to occur in order to be assessed. It is possible for assessments to conduct analyses at multiple spatial or temporal scales, for example analysing effects at field and at regional scale or effects within the current growing season and after ten years.

The selected spatio-temporal scales determine the set of available indicators and have a strong influence on the overall assessment results, because different processes occur at different spatial and temporal scales. The scales should therefore be defined at the start of an assessment and choices should be properly documented.

As with all research projects, the resources available for conducting an impact assessments are limited. This results in a trade-off between thematic, temporal and spatial width of the assessment. The wider the spatial and temporal scales are set, i.e. the more spatial scales from field to global and the more points in time are assessed, the lower is the number of impact areas that can be covered.

A definition of both spatial and temporal scales that fits the purpose of the assessment is essential. If spatio-temporal scales are set too wide, it will result in bloated data requirements, loss of focus and may even obscure relevant findings amongst non-relevant data. If on the other hand they are set too narrow, assessment results may be misleading because relevant impacts are not addressed. For example, effects of agricultural management on the biodiversity of farmland birds may not be visible at the field scale but manifest at the the regional scale. Management for increasing soil organic carbon will hardly show an effect after a single year but may have measureable impacts after 20 years.

With regard to defining spatio-temporal scales, leakage effects and indirect land use changes pose a particular challenge because they require a very wide boundary setting to be detected (Lambin & Meyfroidt, 2011).

The spatial and temporal scales need to be chosen in a way that ensures that all relevant, potential impacts are addressed. This choice depends on the specific topic of the assessment, the purpose of the assessment and on the selected impact areas. For example, an impact assessment with the purpose to analyse policy options needs to cover impacts at the national or regional scale, while a farming system assessment would normally focus on the farm scale. However, if impact areas like habitats, aesthetic value or climate change mitigation are assessed, it may be necessary to expand the system boundaries to the landscape or even global scale.

Due to the complexity involved, expert based pre-assessment and potentially additional research is required before the start of the actual impact assessment.

Inclusion of effects up to a global scale may be required to account for effects like leakage or indirect land use change, where local management decisions result in effects emerging in other world regions. At the temporal scale, the definitions must be sufficiently wide to capture relevant improvements or deteriorations of soil functions or ecosystem service provision, which often only emerge after considerable time lags (Fremier et al., 2013). In the context of agriculture, consideration of multiple years is highly recommended in order to account for long term management effects, crop rotations, pre-crop effects and inter-annual yield variability (Preissel et al., 2015; Zhang et al., 2017).

A structured way to include effects at large spatial and temporal scales is life cycle assessment (LCA). LCA are usually conducted to analyse the use of resources (e.g. water, land, energy) and the occurrence of emissions (e.g. greenhouse gasses, acidification, etc.) along a production process or product life cycle. However, the large spatio-temporal scales covered in LCAs lead to high data requirements and a high workload for assessments. This limits the number of impact areas that can be covered and results in a trade-off between spatio-temporal and thematic detail. Where a high number of impact areas needs to be assessed and the analysis of trade-offs between impact areas is in the focus of the study, LCA may therefore not be the method of choice and standard multi criteria analysis (MCA), for which spatial and temporal scales are not formalised and can be freely adapted to the context of the research, may be preferable.

LCA focuses on environmental impacts of products and production processes (Roer et al., 2013). It seeks to integrate effects from all stages of the product cycle:

–    Production and transport of required materials, machinery and fuels
–    Production of the investigated product
–    Distribution and use of the product
–    Disposal or recycling of the product

This is a very complex and work intensive task and requires a substantial amount of information. It is usually facilitated by the use of LCA specific software and often draws on information from external, commercially available LCA databases (see www.nexus.openlca.org/databases for an overview). However, for a comparison between products or productions processes for which a part of the life cycle is considered to be identical, a full LCA may not be necessary. For different wheat varieties, for example, use and disposal will most likely be the same and an LCA comparing them may therefore focus on the remaining stages of the life cycle. Consequently, a number of studies use only partial LCAs, replacing the underlying “cradle to grave” or “cradle to cradle” concept with “cradle to farm-gate” or “farm-gate to farm-gate” (Brock et al., 2016; Hijazi et al., 2016).

LCAs can be used complementary to impact assessments when specific items (e.g. energy use) need to be assessed in detail along the value chain. Complex, systemic processes such as those related to ecosystem services are very difficult to be assessed with LCA.

Spatio-temporal boundary setting adequate for the respective assessment is necessary to ensure the validity of assessment results. Godinot et al. (2016) show the relevance of including production, transport and losses of inputs into the assessment of nitrogen use efficiencies. Likewise, Roer et al. (2013) demonstrate the considerable effect boundary setting has on assessment results by calculating LCAs of the same products using different system boundaries.  

Brock P M, Muir S, Herridge D F, Simmons A (2016) Cradle-to-farmgate greenhouse gas emissions for 2-year wheat monoculture and break crop–wheat sequences in south-eastern Australia. Crop and Pasture Science, 67(8), 812-822. DOI:10.1071/CP15260

Fremier A K, DeClerck F A J, Bosque-Pérez N A, Carmona N E, Hill R, Joyal T, Keesecker L, Klos P Z, Martínez-Salinas A, Niemeyer R, Sanfiorenzo A, Welsh K, Wulfhorst J D (2013) Understanding spatiotemporal lags in ecosystem services to improve incentives. BioScience, 63, 472-482. DOI:10.1525/bio.2013.63.6.9

Godinot O, Leterme P, Vertès F, Carof M (2016) Indicators to evaluate agricultural nitrogen efficiency of the27 member states of the European Union. Ecological Indicators, 66, 612–622. DOI:10.1016/j.ecolind.2016.02.007

Hijazi O, Munro S, Zerhusen B, Effenberger M (2016) Renewable and Sustainable Energy Reviews, 54, 1291-1300. DOI:10.1016/j.rser.2015.10.013

Lambin E F, Meyfroidt P (2011) Global land use change, economic globalization, and the looming land scarcity. PNAS, 108 (9), 3465–3472. DOI:10.1073/pnas.1100480108

Preissel S, Reckling M, Schläfke N, Zander P (2015) Magnitude and farm-economic value of grain legume pre-crop benefits in Europe: A review. Field Crops Research 175, 64–79. DOI:10.1016/j.fcr.2015.01.012

Roer AG, Johansen A, Bakken A K, Daugstad K, Fystro G, Strømman A H (2013) Environmental impacts of combined milk and meat production in Norway according to a lifecycle assessment with expanded system boundaries. Livestock Science, 155, 384–396. DOI:10.1016/j.livsci.2013.05.004

Zhang Y, Wang R, Wang S, Wang H, Xu Z, Jia G, Wang X, Li J (2017) Effects of different sub-soiling frequencies incorporated into no-tillage systems on soil properties and crop yield in dryland wheat-maize rotation. Field Crops Research, 209, 151-158. DOI:10.1016/j.fcr.2017.05.002