correlating soil microbial properties with crop yields in the

Correlating Soil Microbial Properties with

Crop Yields in the Canadian Prairies

Newton Z. Lupwayi1, K. Neil Harker2, Francis J. Larney1, Robert E. Blackshaw1 and John T. O’Donovan2

1Agriculture & Agri-Food Canada, Lethbridge Research Centre, Lethbridge, Alberta, Canada 2Agriculture & Agri-Food Canada, Lacombe Research Centre, Lacombe, Alberta, Canada

Introduction

• Soil microorganisms mediate many important biological processes for sustainable agriculture.

• However, correlations between soil microbial properties and crop productivity cannot always be demonstrated.

• Two case studies on the Canadian prairies were used to correlate soil microbial biomass with wheat and canola yields.

1. Spring Wheat in Irrigated Crop Rotations: 2002

to 2011 at One Site

3

Vauxhall, Alberta

4

Rotations Code Length Rotation Management

1CONV 1 Yr W Conventional

3CONV 3 Yr P-B-W Conventional

3CONS 3 Yr P-B-W Conservation

4CONV 4 Yr SB-B-P-W Conventional

4CONS 4 Yr SB-B-P-W Conservation

5CONS 5 Yr P-W-SB-W-B Conservation

W = wheat; P = potatoes; B = beans; SB = sugar beet

Outline of Conservation Practices used over 12 yr

1. Composted beef cattle manure as a substitute for inorganic fertilizer

2. Reduced tillage or direct seeding where possible

3. Beans: Narrow-row (20 cm) straight cut vs. wide-row (60 cm) undercut

Rotation

Cont Wht 3-CONV 3-CONS 4-CONV 4-CONS 5-CONS1 5-CONS2 6-CONS

Mic

rob

ial B

iom

as

s C

arb

on

, m

g/k

g s

oil

0

100

200

300

400

500

600

700

bc

c

ab

bc

ab

bc

a

bc

Microbial Biomass Carbon (MBC) in 2011

(but used 2002-2011 data for regression analysis)

8

Rhizosphere

0 200 400 600 800 1000 1200

0

2000

4000

6000

8000

10000

MBC (mg/kg soil)

Wh

ea

t g

rain

yie

ld (

kg

/ha

)

y = 3526 + 4.39x (adjusted r2 = 0.166**)

2. High-Yield No-Till Canola: Seven Sites

9

Seven Sites in the Canadian Prairies

10

Beaverlodge

Lacombe

Lethbridge Swift Current Indian Head

Brandon

Edmonton

Treatments

11

Treatment Seeding rate (seeds m-2) N rate (x recommended rate) N form

1 75 1 Uncoated

2 75 1 Uncoated + Fungicidea

3 75 1 Coated + Fungicide

4 75 1.5 Uncoated

5 75 1.5 Uncoated + Fungicide

6 75 1.5 Coated + Fungicide

7 150 1 Uncoated

8 150 1 Uncoated + Fungicide

9 150 1 Coated + Fungicide

10 150 1.5 Uncoated

11 150 1.5 Uncoated + Fungicide

Soil Characteristics at The Sites

12

Site Soil type pH OM (%) Clay (%)

Beaverlodge, AB Dark Gray Luvisola 6.5 5.4 22.0

Brandon, MB Black Chernozemb 8.1 5.0 33.0

Edmonton, AB Black Chernozem 7.6 13.0 36.5

Indian Head, SK Black Chernozem 7.6 3.2 21.4

Lacombe, AB Black Chernozem 6.4 8.3 21.0

Lethbridge, AB Dark Brown Chernozemc 8.0 3.0 33.0

Swift Current, SK Brown Chernozemd 6.5 3.4 18.2

Summary of Soil Microbial Results

• Treatment effects on soil MBC, β-glucosidase enzyme

activity or bacterial diversity were usually not significant.

• Where significance occurred:-

– Doubling the seeding rate from 75 to 150 seeds m-2 usually

increased these microbial properties.

– Increasing N rate to 1.5x the recommended rate had mostly

positive effects in canola rhizosphere.

– The effects of N form (including addition of fungicide) were

inconsistent.

13

14 MBC in rhizosphere (mg kg-1 soil)

0 500 1000 1500 2000

Gra

in y

ield

(kg

ha

-1)

0

1000

2000

3000

4000

5000

6000

7000

y = 1939 + 0.82x (r2

= 0.307**)

NS

2008: Microbial Biomass - All Sites 2 sites with

high org. C

Lacombe MBC

data - unusual

15

2008: Enzyme Activity - All Sites

Enzyme activity in rhizosphere

0 200 400 600 800 1000

Gra

in y

ield

(kg

ha

-1)

0

1000

2000

3000

4000

5000

6000

7000

y = 1413 + 2.55x (r2

= 0.457**)

y = 5630 - 0.698x (r2

= 0.041*) 2 sites with

high org. C

Lacombe enzyme

data - unusual

16 MBC in rhizosphere (mg kg-1 soil)

0 500 1000 1500 2000 2500

Gra

in y

ield

(kg

ha

-1)

0

1000

2000

3000

4000

5000

6000

2010 (3 sites): Microbial Biomass - Without Lacombe Data

y = 1361 + 1.84x (r2

= 0.744**)

No correlations within

sites.

Large, regional datasets

- important.

Swift

Current, SK

Beaverlodge, AB

Edmonton, AB

17

2010 (3 sites): Enzyme Activity - Without Lacombe Data

Enzyme activity in rhizosphere

0 200 400 600 800 1000

Gra

in y

ield

(kg

ha

-1)

0

1000

2000

3000

4000

5000

6000

y = 1217 + 5.44x (r2

= 0.610**)

No correlations within

sites.

Large, regional datasets

- important.

Root Maggot Damage

18 MBC in rhizosphere (mg kg-1 soil)

0 500 1000 1500 2000 2500

Ro

ot

ma

gg

ot

da

ma

ge

(1

to

5)

0.5

1.0

1.5

2.0

2.5

y = 1.55 - 0.00014x (r2

= 0.036*)

Biological pest control

Less root maggot damage

where MBC (and enzyme

activity) - high.

Conclusion

• Crop/soil management practices that increased soil

microbial biomass and enzyme activity were associated

with improved crop health and yields.

– Feedback: the crop affects rhizosphere soil microbials, which

in turn affect the crop.

• Multi-site experiments show the relationships better

than single-site, single-year experiments.

Acknowledgements