Microbial Processes Underlying the Natural Weed Suppressiveness of Soils

Microbial Processes Underlying the Natural Weed Suppressiveness of Soils

Summary

In the first year of the project, we have focused upon developing our understanding of the dynamics of microbial communities in the rhizosphere of weeds. We have shown that weeds “train” microbial communities in their rhizospheres in a species-specific way, and that these “trained” communities can have dramatically different impacts upon different weed species. We have performed extensive validations of the PCR-DGGE technology demonstrating that we can effectively compare the soil microbial communities between different soil samples. These findings have been presented at three conferences (NCWSS, Louisville; WSSA Kansas City; IWSS, Durban), and at a field day in 2004 (Ames, IA).

Objectives/Performance Targets

PROPOSED OUTCOMES:

Long-term:
• Systemic changes in the way farmers manage soils through a clear understanding of the ways in which microbial communities and key soil microbe species can be manipulated.
• Systemic changes in the purposes for which farmers manage soil, including weed management.

Intermediate-term:
• Deepening of the understanding farmers and extension personnel have of the complexity, composition, and dynamics of soil microbial communities.
• Provide needed information to farmers and extension personnel regarding the impact of management regimes upon soil microbial communities.

Progress on intermediate-term outcomes:
Our conference presentations and farmer field days (listed below) contribute to this outcome. These presentations have reported our preliminary findings which begin to unravel some of the fundamental processes that determine the dynamics of soil microbial communities and their impacts upon weeds. The data required to prepare our first article for a refereed journal is complete, and I envisage submission of a manuscript in the next few months.

Short-term:
• Develop microbial community profiles from soils under different management.
• Quantify the relationship between microbial community structure and key microbial species with soil management regimes.
• Correlate key microbial taxa with weed management outcomes.

Progress on short-term outcomes:
These outcomes have been the primary focus of our efforts to date. We have performed validations of PCR-DGGE (denaturing gradient gel electophoresis of PCR-amplified ribosomal RNA genes) for this system (Accomplishments I), examined the dynamics of microbial communities in the rhizosphere of weeds (Accomplishments II) and measured the impact of microbial communities from the rhizospheres of different weeds (Accomplishments III).

Accomplishments/Milestones

I. Preliminary validations of PCR-DGGE for analysis of soilborne microbial communities.

Methods:
Soil cores (10 cm depth) were taken from experiments near Ames, IA evaluating the effects of various tillage and soil amendment practices on soil characteristics and weed and crop performance in fall 2002 and fall 2003. Plots under the same management regimes (no-till, no compost amendment) but at different points in the rotation (corn vs. wheat vs. soybean) were selected.

PCR-DGGE: DNA was extracted (FastDNA SPIN kit, QBiogene), and bacterial 16S or fungal 18S rDNA amplified using universal bacterial primers (Muyzer et al. 1993) or universal fungal primers (Vaino & Hantula, 2000) with a GC-clamp. PCR-products were separated on DGGE gels (Bacteria: 10% T, 19:1 [5%C] acrylamide/bis-acrylamide, 160V, 18 h, 600C, 35-75% denaturant gradient; Fungi: 8% T, 37.5:1 [2.6%C] acrylamide/bis-acrylamide, 120V, 18 h, 600C, 45-70% denaturant gradient). Gels were stained with SYBR Green and photographed over a UV transilluminator. Sorensen coefficient of similarity was calculated for each band pair as Cs = 2j (a+b), where j = number of shared bands, a = number of bands in lane “a” and b = number of bands in lane “b”. Coefficients were expressed as replicate or sub-sample means (Cs). Dendograms were clustered from Dice coefficients calculated in Quantity One (BioRad).

Result and Discussion:
DGGE separations from repeated PCR reactions from the same DNA extraction and sub-samples of the same soil generated identical or very similar profiles. The similarity between DGGE profiles of replicate plots sampled in 2002 was only slightly greater than the similarity between different treatments however the same replicate plots showed a high degree of similarity in 2003.

Our understanding of the nature of interactions between weeds and soil microbes has been confounded by the diversity and complexity of soil microbial communities and the paucity of suitable techniques. Studies to date have almost exclusively investigated specific interactions in isolation using culturing techniques. The preliminary investigations reported here demonstrate that PCR-DGGE has considerable potential to be applied to studies of the complex soil microbial communities that are associated with weeds. We have shown that our sampling and PCR amplification techniques are robust. Changes in soil microbial communities under different crop management systems were detected by PCR-DGGE. Specifically, we have shown that the soil microbial communities developed under corn, soybean or wheat in the different phases of a rotation are different. Future research will continue to develop our understanding of these community dynamics in order to correlate community structure to crop management and weed demographic parameters.

The practical implications of these findings are that we can now begin to search for microbial explanations for the effects upon weed population demographics that we observe under different cropping systems. For example, Matt Liebman’s group has shown, in these same plots, that the survival of adult giant foxtail was highest in the a 3-yr rotation (23.6 plant per m2) and lowest in the 2-yr rotation (0.3 plants per m2). We will test the hypothesis that part of the explanation for these differences is the impact of the different soil microbial communities that are found in these different cropping systems.

II. Investigation of the soil microbial communities in the rhizosphere of weeds.

Methods:
Shattercane, jimsonweed, velvetleaf, common lambsquarters, barnyard grass, redroot pigweed and ivyleaf morningglory were grown from surface-sterilized seed in identical field soil in the greenhouse for 21-28 d in small containers (Cone-tainers® – to ensure good root colonization of the soil with the expectation of a strong rhizosphere effect). Plants were excavated, the roots shaken to remove excess soil and then the soil clinging to the roots (designated rhizosphere soil) was collected by washing in sterile water. Control modules were maintained without plants. Rhizosphere microbial communities were analyzed at this point, and plants removed and discarded, and new individuals planted. Plants were repeatedly grown to approximately 3 inches and then removed and discarded for a further 70-80 d in order to maximize any “training” of the rhizosphere microbial communities by the plant roots (total approx. 100 d growth of seedlings of a given species in each soil sample). After 100 d rhizosphere microbial communities were analyzed in the same way as before.

PCR-DGGE: As above (I).
PCA analysis by PRINCOMP PROC in SAS and plots prepared from first two principal components in Sigmaplot (SPSS).

Results and Discussion:
The DGGE profiles from the rhizospheres of replicate samples of the same species were very similar, demonstrating the reproducibility of similar microbial communities in the rhizosphere of a weed species. The profiles from the rhizospheres of different weed species grown in the same soil were very different, demonstrating that weeds “trained” microbes in their rhizospheres in a species-specific way. After subsequent re-plantings, the microbial communities in the rhizospheres of weeds continued to change in composition. The rhizospheres of different weed species “trained” different microbes from a single field soil, suggesting that different weed communities and different weed management systems may have far-reaching impacts upon soil microbial communities with concomitant indirect effects upon other cropping system processes.

These findings are important since they show that the weed community in a field can have an impact upon the composition of microbial communities in the soil. The are numerous possible outcomes of this, and we hope that once we understand the dynamics of these changes in sufficient detail we may be able to design crop management strategies that promote the development of soils that are naturally suppressive towards weeds. We explore some of these dynamics below (Section III).

III. Measuring the growth of weeds in “trained” soils.

Methods:
The “trained” soils generated by 100 d of growth of each weed species was collected. Each of these soils was then planted with each of our weed species in an 8 x 8 factorial experiment to determine the effect of soil training by each weed species upon its own growth and upon the growth of each other weed. Plants were grown for 21-28 d and aboveground dry biomass was measured by cutting at the soil surface and drying at 65C.

Results and Discussion:
The microbial communities established to be “trained” differently by different weed species (above: II) were shown to have different impacts. Specifically, we found that some weed species exhibited negative feedback when repeatedly grown in the same soil, whereas other weed species exhibited positive feedback. The most striking result was with velvetleaf. In this case, the plant grew strongly for two re-plantings, but then suddenly crashed, and all plants were killed at the cotyledon to two-leaf stage. We believe that this crash was caused by the gradual training of the soil with a fungal plant pathogen (possibly Verticillium dahliae). No other weed species were affected by the velvetleaf-trained soils. The strongest positive feedback was observed with shattercane where growth of plants was strongest in soil that had been trained by previous plantings of shattercane. We hypothesize that this positive feedback in the rhizosphere of shattercane is due to the training of the soil microbial communities in its rhizosphere with an increased impact of beneficial microbes (nitrogen fixers? mycorrhizal fungi?) rather than deleterious/pathogenic microbes. Further experiments will be designed to unravel the mechanisms of these interactions.

These results are particularly exciting. They demonstrate that relationships between weeds and soil microbes are involved in determining the ecological fitness of weed populations. These findings are similar to the seminal work of Klironomos recently published in Nature (Vol. 417:67-70) (Klironomos 2002) that suggested that the role of microbes may be central in the determination of invasiveness and rarity of plants in natural systems. Here, we are beginning to compile evidence that microbes may be central to the long-term outcomes of weed populations in agricultural systems. We hope that deepening our understanding of these interactions will lead to the development of strategies that can exploit their impacts.

References:
Klironomos, J.N. 2002. Feedback with soil biota contributes to plant rarity and invasiveness in communities. Nature 417:67-70.

Muyzer, G, EC deWaal & AG Uitterlinden. 1993. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Applied and Environmental Microbiology. 59:695-700.

Vaino, E.J. and J. Hantula. 2000. Direct analysis of wood-inhabiting fungi using denaturing gradient gel electrophoresis of amplified ribosomal DNA. Mycol. Res. 104:927-936.

Impacts and Contributions/Outcomes

We have made considerable progress with this project in the first 18 months. We have refined and validated our protocols, and have carefully gathered a large array of samples that will enable a very large analysis of soil microbial community dynamics through time, correlated with weed management and weed demographic data. We have shown that weeds train microbial communities in their rhizospheres in a species-specific way, and that these microbial communities can have different impacts upon weeds after training. Our most important discovery to date is that different weed species appear to be in different dynamics relationships with their rhizosphere microbial communities. While some species appear to become restricted (exist in negative feedback) by their own presence in the soil (e.g. velvetleaf), other species appear to promote their own proliferation (exist in positive feedback) by their impacts upon soil microbial communities.

Extension and outreach for this project have been active, with the presentation of preliminary findings at farmer field days, and with the following conference presentations (abstracts available):

Anderson KI, M Liebman & SG Hallett. 2003. Analysis of soil microbial communities associated with weeds using denaturing gradient gel electrophoresis of PCR-amplified ribosomal RNA genes (PCR-DGGE). Proc. North Central Weed Science Society (NCWSS) annual meeting, Louisville, KY, 12/03.

Anderson KI & SG Hallett. 2004. Application of denaturing gradient gel electrophoresis of PCR-amplified ribosomal RNA genes (PCR-DGGE) for the analysis of soil microbial communities found in different crop and weed management systems. Annual meeting Weed Science Society of America (WSSA), Kansas City, MO, 2/04.

Anderson KI & SG Hallett. 2004. Development of PCR-DGGE for the investigation of soilborne natural enemies of weeds. 4th International Weed Science Society (IWSS) Congress, Durban, Republic of South Africa, 20-24 June, 2004.

We expect to submit the current findings of this project as a refereed journal article in the near future.

Current research is focusing on the large-scale analysis of field soils from the weed management experiments managed by Matt Liebman in Iowa, the results of which, to date, support the hypothesis that diverse rotations that exploit multiple stress and mortality factors contribute to weed suppression. Several DGGE gels will be analyzed to look for correlations between soil microbial community composition with weed management practices and with weed population demographics. Future research will also introduce sequencing to identify key organisms that may be implicated in the responses we observe.

Collaborators: