Organic Broadcaster

Experiment explores impact of amendments on soil microbial communities

By Adria Fernandez

Organic farming is based on the understanding that soils are not just a place for plants to sit—they are complex living systems, home to an enormous diversity of organisms from the tiniest bacteria to earthworms and insects. In the last few years, there has been an explosion of research into soil microbial communities using DNA sequencing technology. Our team at the University of Minnesota, supported by a grant from the Ceres Trust, has been exploring how to make these new sequencing methods useful to organic farmers.

There are two basic things that we want to know about the microbes in soil: who’s there, and what are they doing? The method that we used in this project lets us answer the first question by sequencing a region of DNA that serves as a “fingerprint” for bacteria. This “fingerprint” gives us no direct information about their functions or activities in the soil—it doesn’t even tell us whether they are alive or dead, active or dormant. All it tells us is who’s there, and who they are related to. Most of the bacteria we detect are organisms that have never been isolated, cultured, or described.

Since soil microbes carry out most of the transformation and release of nutrients in crop-available forms, we designed an experiment that would:

1. measure the effects of cover crops and fertilizers on soil nutrient cycling functions;
2. measure the effects of these amendments on soil bacterial communities (“who’s there”); and,
3. explore the relationships between changes in the bacterial communities and changes in soil function.

We planted this experiment at three certified organic sites in Minnesota: at Carmen Fernholz’s farm in Madison, Scott Johnson’s farm in Farmington, and the University’s Southwest Research and Outreach Center in Lamberton. At each site, we established a replicated experiment with eight organic amendment treatments: four cover crops, three OMRI-approved fertilizers, and a no-amendment control. Our cover crops were hairy vetch, winter rye, oilseed radish, and buckwheat. Our fertilizers were solid beef manure, pelleted poultry manure (Chickety Doo-Doos), and Sustane 8-2-4. The cover crops were planted and fertilizers were applied and incorporated in fall of 2012. Overwintering cover crops were terminated in spring 2013, and the study fields were planted to corn.

Cover crops grow at the university’s research center to test impact of soil amendments on nutrient cycling. Photo by Adria Fernandez

We performed routine soil tests (moisture, pH, OM, and nutrients) before and after cover crop and fertilizer application, and measured soil respiration, net N mineralization, and the activity of carbon-, nitrogen-, and phosphorus-cycling enzymes. We also extracted DNA from these samples and sequenced it to generate a “census”
of which bacteria were present and how abundant they were.

Four broad conclusions emerged from this data:

1. Cover crops and fertilizers do affect the microbial transformation of nutrients in soil into forms that are available to plants.
2. Although we all know that biodiversity is valuable as a broad goal, the specifics of soil microbial community diversity, both what practices will promote it and what role it plays in soil function, are complicated.
3. Knowing what bacteria are living in the soil does give us new information about soil function— information that we couldn’t have gotten from routine soil tests. Translating that raw information into predictions that can be used for making management decisions will be an ongoing process that will require extensive research and mathematical modeling.
4. Every soil is different! It is unlikely that this sort of research will result in uniform recommendations; instead, we hope to learn how microbial community profiling can be used to make site-specific predictions about best practices for particular farms and fields.

Soil is such a complex environment that we often measured vastly different levels of soil activity and community composition, even in two subsamples from the same plot and treatment. Also, samples from different locations differed much more than samples from different treatments. Nonetheless, we could see effects of our cover crop and fertilizer treatments. Pelleted poultry manure and Sustane produced large, short-lived spikes in nitrate in the spring, and reduced soil pH. At Farmington, a later-season increase in nitrate in our cover crop treatments suggested that, at that location, the cover crops were doing what they are supposed to: scavenging N during the off-season, then releasing it in plant-available form when the corn crop is ready to take it up.

In some cases, cover crops and fertilizers increased soil nutrient cycling—but not always. Winter rye and pelleted poultry manure were the ones that most frequently increased C-, N-, and P-cycling enzyme activities, and also tended to boost soil respiration, which is a marker of total microbial activity. At Madison, however, we actually saw decreases in phosphatase activity in our oilseed radish, beef manure, and Sustane plots.

Our bacterial community profiles gave further hints about soil function. Although these differences were not consistent across locations and sampling times, we saw increases in Actinobacteria, a group associated with disease suppression, in oilseed radish and winter rye plots, and decreases in Rhizobiales (the family including symbiotic Rhizobium) in poultry manure plots.

We often assume that adding organic matter will always increase diversity in soils. But, it turned out that cover crops and fertilizers also had mixed effects on bacterial diversity. In spring samples, diversity was often lower in cover crop and fertilizer plots than in control plots; though, in July samples, diversity was sometimes higher in plots that had been cover cropped. We also found that most of our nutrient-cycling functions did not increase with higher bacterial diversity.

Both microbial community composition and soil function were closely related to soil test values. In other words, the moisture, pH, OM, and nutrient levels in the soil strongly determined who was living there and what they were doing.

So now we wanted to know, does having all this data about the microbial community actually improve our ability to predict how the soil will function, or are differences in community composition simply a response to differences in soil properties? To address this, we divided our data about each soil sample into three profiles: 1) soil test values; 2) soil functions (nutrient cycling and corn yield); and, 3) census of bacterial groups.

We found that both bacterial community and soil test profiles helped explain the variation in soil functional profiles. In other words, bacterial community composition was not just responding to soil properties, but also providing information about C-, N-, and P-cycling activities in soil that we could not have gotten out of routine soil tests alone.

The price of DNA sequencing continues to drop, and microbial community profiling is becoming more routine. I think it is possible that, in the not-too-distant future, growers might be able to get a microbial community “census” of their soil alongside ordinary soil tests. With more research and modeling, it’s possible that such a census could be used to make predictions about things like whether crops are likely to respond to the addition of specific nutrients or organic materials, or even whether inoculants are likely to be effective. The Ceres Trust is continuing to support our research, including exploring how microbial community differences relate to weed seed degradation, adding fungal sequencing to our analyses, and gathering data from a greater range of organic farm soils. We are excited to discover how sequencing-based microbial community profiling can serve the organic farming community.

Adria Fernandez is with the University of Minnesota Department of Agronomy/Plant Genetics.

From the March | April 2016 Issue

Comments are closed.