Potential updates for post-V6 versions of Holos: Agri-Environmental Indicators and ecosystem services modelling

[sarahp1980]

The move to a more spatially-explicit map-based 'Component Selection' screen in a future version of Holos could also provide the basis for modelling the inclusion of AAFC's Agri-Environmental Indicators (https://agriculture.canada.ca/en/environment/resource-management/indicators) and modelling of the provision of additional ecosystem services from Holos-simulated farms.

How exactly we could do this is still very vague, but some initial ideas are presented for discussion below:
1) Inclusion of the Agri-Environmental Indicators into Holos: we do not yet have detailed information on precisely how these indicators are modelled by the Sustainability Metrics group, but their inclusion into Holos could happen in a number of ways, for example:
a. The Particulate Matter Indicator: This indicator estimates particulate matter (PM; TSP (total suspended particulates), PM10 and PM2.5 in kilotons (kt) per year) from ag operations, with sources including wind erosion, land preparation, crop harvesting, crop residue burning, pollen emission, fertilizer and chemical application, animal feeding operations and carcass burning. To calculate the APMEI, activity data are collected for each agricultural source and a corresponding EF is applied e.g., emissions of primary PM from crop harvesting are calculated by multiplying the area of the crop concerned by an emission factor (kg of PM per ha of crop type per year). The Sustainability Metrics (SM) group calculates this indicator at the SLC polygon level, but if we had the relevant farm and activity data, we could use their EFs to estimate a farm-level PM indicator value. Estimates of PM from wind erosion are based directly on the Erosion Risk Indicator, so if we also incorporate this indicator, that could be our data source for that PM emission category.
A much simpler approach would be to just use a lookup table with SM's already calculated indicator values for the relevant SLC polygon for the farm in question.
https://agriculture.canada.ca/en/environment/resource-management/indicators/particulate-matter-indicator

b. The Soil Erosion Risk Indicator: This is an indicator for the combined risk of wind, water and tillage erosion when climate, soil, topography and climate practices are considered. the SM group calculates this at the SLC polygon scale as the sum of wind, water and tillage erosion rates. Soil erosion is calculated using landform data and the associated soil and topographic data in the National Soil Database - we will need to obtain more a more detailed methodology from the SM group to determine how we can implement these equations in Holos. We could also investigate if and how we could improve on the SM methodology, e.g., by considering the effects of erosion control practices such as grassed waterways, winter cover crops, shelterbelts, contour tillage, etc.; if we consider the erosion effects of these practices, then we should also try to consider their position in the simulated farm landscape, relative to the direction of movement of the eroded soil.
A much simpler approach would be to just use a lookup table with SM's already calculated indicator values for the relevant SLC polygon for the farm in question.
https://agriculture.canada.ca/en/environment/resource-management/indicators/soil-erosion-risk-indicator

c. The Soil Cover Indicator: This indicator summarises the effective no. of days per year that ag soils are not exposed. The indicator considers cover provided by crop canopy, crop residues, and snow on the soil surface. We could estimate this on a per farm or per field basis using information entered by the user, e.g., % residues returned to soil for cropped fields and climate data (no. days of snow cover).
One soil cover day (SCD) can be achieved with 100% cover for one day, 50% cover for two days, 10% cover for 10 days, and so on. The indicator results are expressed in the mean annual number of SCDs at SLC level.
The Soil Cover Indicator takes into account the following variables:

• the days on which significant changes in soil cover occur (for example, planting, harvesting, tillage) and the percentage of soil cover upon completion of the operation,
• the duration (number of days) and percentage of soil cover between operations,
• canopy development and decline between planting and harvest,
• the decomposition of residue,
• the total number of days of snow cover greater than 2 cm,
• multiple cuts of hay and grazing on pasture.
  https://agriculture.canada.ca/en/environment/resource-management/indicators/soil-cover-indicator

d. The Residual Soil N indicator and the Risk of Water Contamination by Nitrogen Indicator (IROWC-N): The Residual Soil N Indicator is the difference between total N inputs to ag soils (fertilizer and manure, fixation by leguminous plants, wet and dry atmospheric deposition) and total N outputs (harvested crops and gaseous N losses). Surplus N (mostly water soluble NO3-N) may remain in the soil over winter and be used by the next crop (residual soil N) or may be leached.
IROWC-N evaluates the risk of water contamination by N across ag regions. This uses the Residual Soil N Indicator (amt of N in soil after harvest) and climatic factors to determine the risk of leached NO3-N reaching ground water or tile drainage water - it does not estimate losses of N in surface runoff, which are considered to be minor.
In Holos, we already have estimates of the various N inputs and outputs, as well as of the amount of NO3-N lost from housing/storage and land application of fertilizer/manure. From the model outputs already available, we could calculate the amount of N that remains in the soil after harvest and the amount of NO3-N lost per hectare of farmland?
https://agriculture.canada.ca/en/agricultural-production/water/nitrogen-indicator

e. The (Potential) Wildlife Habitat Capacity on Farmland Indicator: This indicator provides a multi-species assessment of broad-scale trends in the capacity of the Canadian ag landscape to provide potential reproductive and feeding habitat for populations of terrestrial vertebrates. Potential Wildlife Habitat Capacity (PWHC) was determined for the ag extent of Canada at 5-year intervals from 2000-2015 at the SLC level. A habitat association matrix was constructed for 545 terrestrial vertebrates (332 birds, 134 mammals, 41 amphibians and 38 reptiles) that use land cover within the agricultural extent of Canada for reproduction and/or feeding. Each cover type was identified as Primary (always used, critical or strongly preferred habitat), Secondary (often used, important habitat) or Tertiary (occasionally used, low value habitat) with values of 1.0, 0.75 and 0.25, respectively. Habitat association matrices were spatially linked to land cover information within the agricultural extent by rectifying species breeding distribution data to the SLC level.
We would need to determine if and how we can use the SLC-level calculations at the farm scale, and if and how we can incorporate measures of landscape heterogeneity and habitat fragmentation into our estimates.
https://agriculture.canada.ca/en/environment/resource-management/indicators/wildlife-habitat-capacity-farmland-indicator
Other potential collaborators could include:
Jess Vickruck (AAFC RES Fredricton NB): doing work on pollinators (particularly bees, incl. ground-nesting bees) in the NB Living Labs project and on the effects of landscape structure and composition in agricultural systems as well as ag management practices on pollinators, and vice versa;
Etienne Lord (AAFC, Saint-Jean-sur-Richelieu Research and Development Centre, QC) focusses on biodiversity monitoring in ag landscapes using 'soundscapes'; particularly interested in birds and if they can be good indicators of the agricultural landscape and the use of AI in processing/analysing the recorded data. Will be part of new LL in QC that will run until 2028, doing this work.
John Wilmshurst (Native Grassland Conservation manager, Canadian Wildlife Federation) works on LL Central Prairies project; working on annual crop and perennial cover ecosystems to examine impacts on carbon/GHGs and biodiversity;
Tom Forge (AAFC Summerland, BC) works on soil biodiversity and is part of the LL BC and LL Peace Region projects.

2) Modelling of additional ecosystem services (ES):
a. Using the AEIs to estimate ES provision: The inclusion of some of the Agri-Environmental Indicators could facilitate the modelling of certain ecosystem services, e.g., via more direct estimation of the level of a service provided by the modelled farm using indicators such as Residual Soil N Indicator (soil quality regulation/nutrient cycling) or via more indirect measurements whereby changes in the indicator can be used to derive estimates of service provision, e.g., a change in the Particulate Matter Indicator could be used to signal an increase/decrease in the air quality regulation capacity of the farm social-ecological system;

b. Inclusion or restoration of landscape features that provide multiple ES within and beyond the boundaries of the simulated farm, for example:
i) shelterbelts and lineal tree plantings, already modelled in Holos in relation to their C sequestration services, also provide a range of other ES, e.g., air quality regulation, water quality regulation, water flow/supply regulation, local climate regulation for people and animals (via shade provision, protection from wind, rain and snow), aesthetic value, habitat provision/wildlife habitat capacity, etc. As a first step, we could see if and how the existing AEIs could be used to model these ES in relation to trees/shelterbelts, and if not, we could start with a simplified approach based on the scientific literature (where available), e.g., try to derive lookup values for the nutrient/pollutant retention capacity of a specified width/depth/density of tree planting, taking into account the location of the trees relative to the slope of the land, the location of the nutrient/pollutant source and nearby water bodies and soil/climate factors that affect nutrient movement throughout the landscape;
ii) (perennial) buffer strips, similar to shelterbelts in that they can provide multiple ES, including pollination services, and the level of ES provision depends on the type of ES, the location of the buffer strip(s) in relation to other activities on the farm, e.g., distance/slope from pollutant sources and sinks, distance to cultivated fields (for pollination services), soil/climatic factors, etc.;
iii) wetlands, which we plan to implement in a future version of Holos in relation to GHG emissions and soil C fluxes. These landscape features could also be considered in terms of their impact on the provision of other ES, in a manner similar to that for shelterbelts and (perennial) buffer strips. The value of wetland habitats for biodiversity/wildlife is of particular interest;

c. Using Holos to explore the impacts of a wider range of agricultural production systems on ES provision: It would be interesting, if we can include a wider range of ES (including habitat quality/biodiversity/wildlife indicators) into Holos, to then use the model to explore the more holistic effects of different farming practices on multiple ES, e.g., conventional vs organic farms, large-scale industrial farming vs small-scale integrated crop-animal farms, etc.