Advancing Cover Crop Knowledge: Assessing the Role of Plant Diversity on Soil Change

The first objective of this study was to investigate how biomass production compares among the ten cover crop treatments included in this study and incidential weeds in the fallow control (Figure 1). Legume treatments (pea and nitrogen fixers) yielded best at the Amsterdam site. This indicates that nitrogen was limiting plant growth. At the Amsterdam site, there was 2.5 times more biomass for the nitrogen fixer treatment than for the minus nitrogen fixer treatment (p-value < 0.0001). At the Conrad site, there was almost 2 times more biomass for the nitrogen fixer treatment than for the minus nitrogen fixer treatment (p-value = 0.01). This supports the previous claim that the system was nitrogen limited.

A contrast was run between the 2-species treatments (functional groups) and the 6-species treatments (minus functional groups) for each site to test if biomass was associated positively with a higher number of species in the mix. The Amsterdam site revealed no statistical difference between the 2-and 6-species treatments (p-value = 0.33). At the Conrad site, the 6-species treatments out-yielded the 2 species treatments by about 20% (p-value = 0.01), however there was greater variation due to the higher mass of weeds at this site.

Figure 1: 2018 Cover crop biomass (tons acre^- dry basis) for all 11 treatment. Each site location was analyzed separately. Summer fallow, full mix, and pea are compared then pairwise comparisons for each of the four functional groups (BC, FR, NF, TR) and their corresponding minus treatments are compared individually. Means that differ from one another (P<0.10) are marked with different letters. Means that do not differ (P>0.10) are left blank. * = pairwise comparison at both sites differ from one another.

The second objective of this study was to measure how plant functional groups affect biological, chemical, and physical soil properties, and subsequent wheat yields and protein, differently. Many soil parameters were chosen to investigate, with a strong focus being placed on the biological soil parameters. Figure 2 shows the available soil water to a 3-foot depth after the cover crops have been terminated. By looking at figures 1 and 2, water use is proportional to the amount of biomass each treatment grew, which is to be expected.

Figure 2: Available soil water (inches) to a 3 ft depth at the time of cover crop termination in 2018. Each site location was analyzed separately. Summer fallow, full mix, and pea are compared then pairwise comparisons for each of the four functional groups (BC, FR, NF, TR) and their corresponding minus treatments are compared individually. Means that differ from one another (P<0.10) are marked with different letters. Means that do not differ (P>0.10) are left blank.

Figure 3 shows soil nitrate-N levels after cover crop termination in 2018 for the fallow, pea, and full treatments. There were no statistical differences between nitrate values for fallow, pea, and the full mix at Conrad and there were no differences between pea and the full mix at Amsterdam. At Amsterdam, the fallow treatment had 2 times as much nitrate- N in the soil than the pea and fallow treatments. The observed soil nitrate values at Conrad were highly variable due to drought experienced in the previous wheat stage in 2017.

Figure 3: Soil nitrate-N (lbs acre^-1) in the top 3 ft for the fallow, full mix and pea treatments in 2018. Each site location was analyzed separately. Means that differ from one another (P<0.10) are marked with different letters. Capitol letters represent the Amsterdam site. Means that do not differ (P>0.10) are left blank.

Potentially Mineralizable Nitrogen

Potentially mineralizable nitrogen was measured at all three nitrogen rates for the full mix, summer fallow, and sole pea treatments. To maintain a balanced design, we analyzed this data set separately from the rest of the PMN data. A mixed linear effects model was used where treatment and fertilizer rate were fixed effects and block was a random effect. Interaction terms were not significant to the model. At both the Amsterdam and Conrad sites fertilizer rate did not significantly change PMN values. After accounting for differences due to block and fertilizer rate, a treatment affect was present at both Conrad (F = 4.37, p = 0.02) and Amsterdam (F = 21.1, p < 0.0001). Figure 5 shows the average PMN values for the fallow, full mix, and pea treatments at three N rates (high, medium, low) at both site locations. At Conrad, Pea had higher PMN values than summer fallow (p < 0.01) after accounting for block and fertilizer rate. There was little to no evidence showing differences among any other pairwise comparison. At Amsterdam, both pea and the full mix have higher PMN values than summer fallow (p < 0.01 for both), after accounting for fertilizer

PMN was measured for all 11 cover crop/control treatments at the medium nitrogen rate (60 lbs N-ac^-1) (Figure 4). At the Amsterdam site, both pea and the full mix have higher PMN values than summer fallow at the medium nitrogen rate (p <0.01), while there were no differences between the full, pea, and summer fallow treatments at Conrad. Pre-planned orthogonal contrasts were used to compare the presence vs absence of each functional group (functional group vs minus functional group) at the medium nitrogen rate. There were no differences between any of the pairs at either site. A two-species (the four single-functional group treatments) verse six-species (four minus treatments) contrast was analyzed at both field sites for the medium nitrogen rate (60 lbs N-ac^-1). At both sites, the six species mixes had higher PMN values than the two species mixes (Amsterdam: p = 0.0037, Conrad: p = 0.0222).

Figure 4. Potentially mineralizable nitrogen (mg-N Kg^-1) for 11 treatments at the medium nitrogen rate, following four rotations of cover crops at two field sites (2019). Summer fallow, full mix, and pea are compared then pairwise comparisons for each of the four functional groups (BC, FR, NF, TR) and their corresponding minus treatments are compared individually. Means that differ from one another (P<0.10) are marked with different letters. Means that do not differ (P>0.10) are left blank.

Figure 5. Potentially mineralizable nitrogen (mg-N Kg^-1) for summer fallow, full 8-spp mix, and sole pea treatments at three nitrogen rate (high: 120 lbs-N ac^-1, medium: 60 lbs-N ac^-1, low: 0 lbs-N ac^-1), following four rotations of cover crops at two field sites (2019).

Mycorrhizal Colonization

The four functional groups, minus brassica, full mix, and both control (sole pea and summer fallow) treatment plots were measured for mycorrhizal colonization during the subsequent wheat phase (Figure 6). At Conrad, arbuscular mycorrhizal colonization of wheat was 9% higher in the fibrous root plot than the full mix (p = 0.0015, t-value = 3.64) and 10% higher in the fibrous plot than the brassica plots (p = 0.0006, t-value = 4.04). At the Amsterdam location no differences between treatments were seen. Many studies have shown that reducing or eliminating the fallow periods (bare soil) with the use of cover crops increases arbuscular mycorrhizal fungus formation in the soil (Garcia-Gonzalez et al., 2016; De Vries and Bardgett 2012; Schipanski et al. 2014). However, none of these experiments took place in a semi-arid environment with such limiting annual precipitation.

Figure 6. Arbuscular mycorrhizal colonization (%) for 8 treatments at the medium nitrogen rate, following four rotations of cover crops at two field sites (2019). Summer fallow, full mix, and pea are compared then the four functional groups (BC, FR, NF, TR) are compared separately. Means that differ from one another (P<0.10) are marked with different letters. Minus BC was included in this analysis and compared to the functional groups. Means that do not differ (P>0.10) are left blank.

Soil Enzymes

There were no differences in soil enzymatic geometric mean when comparing treatment effect at either site. After two rotations of cover crops, higher soil enzyme geometric means were observed in cover crop treatments compared to summer fallow (Housman et al., 2012). However, after 4 rotations of cover crops (8 years) distinct differences were not seen (Figure 7).

Figure 7. Geometric mean enzymatic activity (mg PNP g soil^-1 hr^-1) of five soil enzymes (ß- glucosidase, ß -glucosaminidase, acid and alkaline phosphatases, and arylsulfatase) for 7 treatments at the medium nitrogen rate, following four rotations of cover crops at two field sites. Means that do not differ (P>0.10) are left blank (without letters).

Microbial Biomass

Mixed linear regression models were used to assess SIR-induced respiration rate as an index of microbial biomass at each site separately. Figure 8 shows the average SIR-induced respiration rate for the fallow, full mix, pea and four functional group mixes. Soil that had pea growing in it prior, had a higher microbial respiration rate than soil following summer fallow at the Conrad site (p = 0.0353, t-value = 2.28). The same pattern was seen at the Amsterdam site but was less significant (p = 0.06, t-value = 1.98). At the Conrad site, some evidence showed the full mix treatment plot had higher microbial biomass than the summer fallow treatment (p = 0.07, t-value = 1.92). No statistical differences were seen between functional groups at either field site.

Figure 8. Microbial biomass (uL CO2 g-soil^-1 hr^-1) for 7 treatments at the medium nitrogen rate, following four rotations of cover crops at two field sites (2019). Summer fallow, full mix, and pea are compared then the four functional groups (BC, FR, NF, TR) are compared separately. Means that differ from one another (P<0.10) are marked with different letters. Means that do not differ (P>0.10) are left blank.

There were no observed differences in soil biological parameters between cover crop plant functional groups in this study. In semi-arid environments, where the system is most likely limited by biomass or carbon inputs, the soil microbial organisms are processing the cover crop biomass quickly enough where any soil biological changes that may have been present during cover crop growth, does not persist in the soil long enough for the changes to be seen by the cash crop (soils taken at wheat seeding). Although observed differences in enzyme activity was not seen, Lehman et al. (2015) points out an important limitation on current soil enzyme interpretation and how its real-world application is to optimize ecosystem functions and services from soil enzymes, the processes that control soil enzymatic communities (structure and function) must be understood. Enzyme geometric means were used in this study because it is considered unreliable to assume a single enzyme can represent the cycling of nutrients in the soil. This is because the total microbial activity serves much more complex functions within the soil which many biochemical reactions contribute to (Gil-Sortes, 2005). Enzyme production should increase when simple nutrients are scarce and complex nutrients are abundant because the enzymes have the ability to break down these complex nutrients (Allison and Vitousek, 2005), but in semi-arid environment such as this cover crop biomass may be the ultimate limiting factor.

Wheat Yield

Wheat yield for both site locations at three N rates are shown in Table 4. At the medium nitrogen rate, wheat following summer fallow had a higher yield than following the full mix or pea at the Amsterdam site, while no differences were seen for this same comparison at the Conrad site. Wheat following N fixers, at the medium N rate, had a higher yield than minus N fixers at the Conrad site. Brassica had a higher yield than minus brassica at the Amsterdam site at the medium N rate and this brassica response was observed at both sites for the previous wheat cycle (Housman et al., 2012).

Wheat Protein

Wheat protein content for both site locations at three N rates are shown in Table 5. At the medium N rate, wheat following pea had higher protein levels than following summer fallow or the full mix at both sites. For presence vs absence comparisons at the medium N rate, wheat following any mix containing N fixers had higher protein than the mix with no N fixers at Amsterdam.

The third objective of this project was to increase local producer knowledge of the value, or lack there-of, for cover crop mixtures compared with less diverse cover crops. The field day presentations at both the Montana State University Research Post Farm and the Northern Agricutural Research Center (NARC) brought the findings of this study directly to local Montana producers. Oral presentations were given and an abundant amount of time was left for producers and agricultural professionals to ask specific questions about the research. This on farm presentation style attracted local producers specifically and seemed to both increase knowledge of potential benefits of cover crops and warn them of potential downfalls of growing cover crops in such a dry environment.

The final objective (objective 4) of this research project was to provide producers with information regarding the potential benefits of alternative cover crops through scientific publications, extension publications, popular press articles, social media, public presentations, and research summaries and factsheets. Several public presentations were given on the research project and factsheets were handed out at these presentations to anyone who was interested. A masters thesis will be completed and published on the project as well in summer of 2020.

References

Allison, S.D., Vitousek, P,. 2005. Responses of extracellular enzymes to simple and complex nutrient inputs. Soil Biology and Biochemistry 37, 937-944.

de Vries, F. T., & Bardgett, R. D. (2012). Plant–microbial linkages and ecosystem nitrogen. retention: lessons for sustainable agriculture. Frontiers in Ecology and the Environment, 10(8), 425-432.

Fierer, N., “Soil Microbial Biomass Determination.” Schimel Laboratory Protocols, UC Santa Barbara, 10 April 2003. Web. 1 June 2014.

García González, Irene & Quemada, M. & Gabriel, Jose & Hontoria Fernández, Chiquinquirá. (2016). Arbuscular mycorrhizal fungal activity responses to winter cover crops in a sunflower and maize cropping system. Applied Soil Ecology. 102. 10-18. 10.1016/j.apsoil.2016.02.006.

García-Ruiz, R., Ochoa, V., Hinojosa, M.B., Carreira, J.A., 2008. Suitability of enzyme activities for the monitoring of soil quality improvement in organic agricultural systems. Soil Biology and Biochemistry 40, 2137-2145.

Gil-Sotres, F., Trasar-Cepeda, C., Leirós, M. C., & Seoane, S. (2005). Different approaches to evaluating soil quality using biochemical properties. Soil Biology and Biochemistry, 37(5), 877-887.

Hausman, M. 2016. Multi-species cover crops in the northern great plains: an ecological perspective on biodiversity and soil health. Master’s Thesis, Montana State University, Bozeman, MT.

Keeney, D. 1982. Nitrogen – availability indices. In: Methods of Soil Analysis, Part 2: Chemical and Microbiological Properties (ed. S. Segou), pp. 711–733. SSSA, Madison, WI.

Lehman, R., Cambardella, C., Stott, D., Acosta-Martinez, V., Manter, D., Buyer, J., Karlen, D. (2015). Understanding and Enhancing Soil Biological Health: The Solution for Reversing Soil Degradation. Sustainability, 7(1), 988–1027. doi:10.3390/su7010988

McGonigle, T.P., Miller, M.H., Evans, D.G, Fairchild, G.L., Swan, J.A., 1990. A new method which gives an objective measure of colonization of roots by vesicular- arbuscular mycorrhizal fungi. New Phytologist 115, 495-501.

Parham, J.A. and Deng, S.P. 2000. Detection, quantification and characterization of b-glucosaminidase activity in soil. Soil Biol. Biochem. 32, 1183–1190.

Schipanski,M.E., Barbercheck, M., Douglas, M.R., Finney, D.M., Haider, K., Kaye, J.P., Kemanian, A.R., Mortensen, D.A., Ryan, M.R., Tooker, J., White, C., 2014. A framework for evaluating ecosystem services provided by cover crops in agroecosystems. Agricultural Systems 125, 12-22.