This document describes the development of a reactive nitrogen model for Canadian agricultural lands. The model accounts for nitrogen inputs like fertilizer and biological fixation by legumes, and outputs like food and feed crops harvested, nitrogen lost through leaching and greenhouse gases, and residual soil nitrogen. The model was applied at various geographic scales within Canada from 1981-2011. Results show nitrogen inputs and outputs are balanced each year, and nitrogen leaching increased from 1981-2001 before declining from 2001-2011. The model provides insights into nitrogen flows and trends over time on Canadian farmlands.

1. Agriculture and
Agri-Food Canada
Agriculture et
Agroalimentaire Canada
J.Y. Yang and C.F. Drury
Harrow Research and Development Centre
Agriculture and Agri-Food Canada
Harrow Ontario, N0R 1G0, Canada
Developing a Reactive Nitrogen
Model for Canadian Agricultural
Land
Canada
The 72th SWCS International Annual Conference, Madison, Wisconsin,
July 30-Aug 2, 2017

2. Agriculture and Agri-Food Canada
Agriculture and
Agri-Food Canada
Agriculture et
Agroalimentaire Canada
Canada
Harrow Research and Development Centre (42.03N;82.92W)

3. 1 Introduction
2 Reactive N model
N inputs
N outputs
Data sources
Scale-up Methodology
3 Results and discussion
4 Summary
Table of contents

4. Canada has ≈ 60 million ha of farmland, about 1.6
Mt fertilizer N is used, and 1.3 Mt biological N
fixation by legumes, and 1.1 Mt manure N is
produced each year (2000 - 2011).
About 30-60% of N input was lost via
- Greenhouse gas (N2O, NO)
- Ammonia emission (NH3)
- N leaching and runoff (NO3)
Introduction

5. Reactive N model
Reactive N model and assumption
At given region, N input = N output in a yearly cycle,
assuming soil organic N is stable.
Reactive N model is the updated version of the
Canadian Agricultural Nitrogen Budget (CANB) v4.0
model
The CANB v4.0 model was used to estimate two Agro-
Environmental Indicators at Canadian Farmland 1981-
2011
Reference: Yang et al., 2013. Ecological Modelling 267, 26-38

6. N food = N removal by food and fiber crops by humans
N feed = N removal by animal feed
N loss = N2O + NH3 + NO3 (leaching + runoff)
RSN = Residual soil Inorganic Nitrogen present in the soil
after harvest
N input = N fert + N fix + N depo + Nfix_min
N output = N food + N feed + N loss + RSN
N fert = N input from fertilizer
N fix = Biological N fixation and residual N mineralization
N depo = Dry and wet N deposition from atmosphere
N fix_min = N mineralization from legume crop residue
Reactive N model

7. Reactive N model
N feed = N meat + MN total
N manure = MN inorganic + MN organic + MN_NH3 + MN_N2O
Internal cycles of manure N

8. Fig. 3 Canadian Ecoregion map (1:1M)
Canadian
Farmland has
Ecozone 8
Ecoregion 68
Ecodistrict 377
SLC 3300
Province10
Base calculation unit: Soil Landscapes of Canada
(SLC) 1:1 M soil polygon
Nested in ecodistrict, ecoregion, ecozone
The CANB-Reactive N model

9. National
Soil – Great Group (28)
Province (10)
Eco-Zone (8)
Eco-Region (68)
Eco-District (377)
Soil Polygon (3300, 1:1M)
1981 – 2011 - Current study at SLC
20161981, 1982, ……, 2011
2016 – Census data will be available in 2017
Output and scale-up methodology

10. Data sources
At Soil Landscape of Canada scale
At Eco-district scale
At Provincial scale
At National (Canadian) scale
At Soil Great Group scale

11. Examples of N data
at SLC scale
Fertilizer N
Biological N fixation
Crop yields

12. Fertilizer N
recommendation
24 crop types
28 soil types
1981 to 2011
Inputs of
CANB model
Adjust fertilizer N rate to equal the total provincial N sale
Calculate total fertilizer N
application
Crop area
27 crop types
3200 SLCs
1981 to 2011
Fertilizer N sales
10 province
1981 to 2011
)p(total_Nrcmd
)p(Nsales
)j,i(Nrcmd)j,i(Nfert raterate 
Fertilizer N
Yang et al., 2011. Can J Soil Sci. 91: 107-122


24
1
),(),()(_
j
rate jiCropAreajiNfertiapplyNfert

13. Crop yield data
For 24 COA crops
At 3200 SLCs
1981 to 2006
Inputs of
CANB model
N fixation rate
adjustment
N fixation
calculation
Crop area data
For 24 COA crops
At 3200 SLCs
1981 to 2006
Rate parameters
Mean N fixation rate
Shoot root ratio
Harvest Index


24
1
),(),()(
j
ratetotal jiAreajiNfixiNfix
)(
),(
)(),(
jyield
jiyield
jNfixjiNfix
mean
meanrate 
Biological N fixation
Yang et al., 2010. Agri. Ecosyst. Environ 137, 192-201

14. Statics Canada
Annual yield
At 65 Canadian
Agricultural Region
12 COA crops
At 65 CAR scale
Re-class
Downscale to
SLC
Statics Canada
Annual yield at 10
province
71crop+veg+fruit
21 COA crops
At 10 province
Re-class
Missing value
24 COA crop yield data
At 3200 SLC (1981 to 2006)
Yield from literature
Alfalfa
Improved pasture
Unimproved pasture
Missing value
3 COA crops
At 10 Province
Crop yield & resolution
Missing
values?

15. Result and discussions
N inputs & outputs at various scales
NO3 lost map
NO3 lost trend (1981-2011)
Reactive N balance in 2011

16. N input at Canada
kg-N per farmland area

17. Total N output at Canada
N loss N food, N feed & RSN

18. N outputs in kg N/ha farmland
kg N per farmland area

19. N input at a provincial scale (kg N/ha)

20. N output at a provincial scale (kg N/ha)

21. NO3 loss map across Canadian Farmland
NO3 leaching in 1981

22. NO3 loss map across Canadian Farmland
NO3 leaching in 2011

23. NO3 loss (kg N ha-1 farmland)
1981
1991
2001
2011
Western Canada Eastern Canada

24. NO3 loss at a provincial scale
Total NO3 loss
NO3 loss per
farmland area

25. NO3 loss trend
Average NO3 loss increased from 1981 to 2001, then
declined linearly from 2001-2011

26. Reactive N balance in 2011
Reactive N balance (1000 Tonne N)
N input = Nfert + Nfix + Ndepo + Nfix_min = 2009 + 1296 + 306 + 340 = 3951
N output = Nfood + Nfeed + Nloss + Nsoil = 1735 + 1290 + 416 + 510 = 3951
Internal cycle (1000 Tonne N)
MN total = MN inorganic + MN_NH3 + MN org = 333 + 227+ 352 = 912
N feed = MN total + N meat = 912 + 378 = 1290
N loss = NO3 + N2O + FN_NH3 = 214 + 72 + 130 = 416

27. Reactive N balance in 2011
N input 3951
N fert 2009
N fix 1296
N depo 306
N fixmin 340
Farmland
Livestock
N loss (water) RSN in soil
NO3 214 51
N input = N output = 3951 (1000 tonne N)
N feed
1290
N food
1735
N Meat
378
N manure
total 912
N loss (N2O + FN-NH3) = 72 + 130 MN_NH3 227
MN org 352
Carryover to next year
MN avail
333

28. Reactive N model estimated the N input and N output at
various scales across Canadian farmland annually
Fertilizer N and N fixation are the main N inputs
N crop removal are the main N outputs
Conclusions
NO3 loss had an increasing trend from 1981-2001 and then
decreased from 2001-2011
N balance showed negative or lower values from 1981-1995
and then became a surplus from 1996 to 2011
Temporal trend

29. Conclusions
NO3 loss was <5 kg N/ha in the Prairie provinces; it ranged
10-40 kg N/ha in Eastern Canada province due to high
precipitation in these provinces
Spatial variation
In 2011 balance, N input = 3951 (1000 tonne N)
N food+N feed = 76.6%
NO3 loss = 5.4%
N2O+NH3 = 5.1%
RSN in soil = 12.9%
Summary of N output and fate in 2011

30. Agriculture and
Agri-Food Canada
Agriculture et
Agroalimentaire Canada
Canada
AAFC Sustainability Metrics
Funding Support
Acknowledgements
Thank you