Progress report for OW20-357
The producers of vermicompost have long acknowledged the benefit of having a diverse and active soil microbial community. However, this diversity has yet to be properly quantified and is often still seen to be a black box. Next-generation DNA sequencing has allowed researchers to shed light into these black boxes and have led to large projects such as the Human Microbiome (Turnbaugh et al. 2007) and Earth Microbiome Projects (Gilbert et al. 2014). However, one of our most valuable resources in restoring soil and plant health, the microbial community of vermicompost, has been neglected and its producers left in the dark. The goals of this project are 1) develop a baseline microbial community analysis of vermicompost and teas from producers across the nation and 2) help this community better understand and quantify diversity in order to improve and increase adoption of their products. Questions such as “do inputs affect diversity,” “does vermicompost have a higher biological diversity than big box compost”, and “does education of producers and their clients help sales”? are among the many questions that can be addressed with this data. Physical and chemical analyses will also be used to evaluate vermicompost qualities to develop some understanding of how these familiar measurements relate to microbial communities. Communication through email and website will be crucial. We plan to present our results and conclusions at the 21st NC State Vermicomposting conference as well as publish in an open source journal. Education and outreach materials will be created and disseminated to producers.
The overall objective of this project is to collect and analyze a large and diverse amount of vermicompost and vermicompost tea in order to enhance the knowledge of sustainable vermicomposting practices and to bring a new technology and understanding to compost producers. The first objective will be to collect samples nationally from vermicompost producers. We aim to collect 150-250 samples from as diverse group of producers as possible. Our second key objective is to successfully sequence received samples and have the vermicomposts analyzed conventionally using Seal of Testing Assurance Program (STA) methods. DNA extraction and sequencing preparation will be performed at the Colorado School of Mines saving significant sample analysis costs over a private lab. Compost STA properties will be analyzed by Midwest Labs.
Our third objective is to use bioinformatic tools to process the sequencing data for quality control and taxonomic assignment. By comparing the communities found in samples, the metadata from the vermicompost STA analysis, and producer answered questionnaires we can begin to answer key deliverable questions. These questions include: Is there a core microbial community found in all vermicomposting samples and why might they be universally present? Do composting inputs and techniques have an effect on overall diversity? Can a metric be developed based on this survey of samples to distinguish between different vermicomposts? How do the microbial communities of vermicompost change overtime? Can we also use this technology for pathogen tracking from manure inputs? We also want to compare vermicomposts to their tea derivatives and to big box vermicomposts and compost brands.
Our fourth objective is to educate the producers and the public about microbial soil diversity and it’s benefits to sustainable agriculture. This will be accomplished by virtual interaction, face to face discussion (virtual conference), publishing in an open source journal, and providing producers with a public outreach package.
- - Technical Advisor (Researcher)
Sample Collection: In order to collect samples from producers nationally, Zack and Kathy will attend the NC State vermicomposting conference (Oct 2019) to recruit producers. Beginning in April of 2020, sampling kit, instructions, and a questionnaire will be prepared and mailed to each participant. The questionnaire will ask producers detailed questions about their vermicomposting methods which will greatly enhance our data set. Producers with a more refined or unique process will be asked to provide unique samples or even take microscopy pictures of tea to help with the analysis. The sampling kit will include screw top tubes with ice packs for approximately 2 grams of soil. Shipping supplies and separate bag for a 200 grams sample is provided by Midwest Laboratory. Product sampling will follow a protocol as defined by Midwest Laboratory.
DNA Extraction and Sequencing: Once all samples have been received, DNA will be extracted with a MoBio Power Soil extraction kit. DNA from each sample will be quantified and then amplified with “universal” primers (Parada, Needham, and Fuhrman 2016) that amplify the 16S and 18S genes in order to identify bacterial, fungal, and eukaryotic organisms. Amplicon will be sent out for sequencing at Duke university using an Illumina MiSeq. While the taxonomic resolution of this type of sequencing reaches to the genus level it still allows for an adequate profile of diversity and general shifts in populations which can be correlated to specific variables.
Soil Analysis: The larger 200g sample of vermicompost collected from each producer will be sent to Midwest Laboratories in Omaha, NE for analysis. They will perform the “compost plus package” analysis which includes moisture, total nitrogen, phosphate, pH, total carbon, as well as other elements and nitrogen species.
Data analysis: Once the DNA sequencing and soil analyses are performed, an in-depth analysis of the microbial and soil testing data will be undertaken. The sequencing data will be processed for demultiplexing, quality control, and taxonomic identification using open source software R, QIIME2 (Bolyen et al. 2019, 2) with the SILVA database (Pruesse et al. 2007). Processing power will be provided by the systems already owned by the University and lead investigator. The final output from this program will yield an operational taxonomic unit (OTU) table in which each sample has its microbial community broken down by relative abundance as well as taxonomic tree file. This OTU table and tree file, will be merged with the soil analysis data using a sample mapping file which will allow for each sample to have both soil data and microbial community data associated with each other which is known as a biome file.
From this point, the statistical analysis can begin primarily using a combination of free and open source software including Tableau, (“Tableau” 2016) and R. A basic alpha diversity analysis will be performed using the Faith’s phylogenic diversity analysis which yield a numerical diversity measurement for each sample. This allows for sample diversity to be directly compared to one another with a logical scale. Additionally, a beta diversity analysis will be performed which allows for the visualization of how similar the microbial communities are to one another on a coordinate system. Colors and shapes can be used to highlight similar community clustering based on experimental variables in order to see which soil properties correlate to certain microbial taxa a canonical correspondence analysis will also be performed using the taxonomy data and vermicompost analysis performed by Midwest Labs. In order to better understand microbial community similarities, a core microbiome analysis will be performed. This analysis will help us understand if communities are similar between vermicomposting methods and materials. In contrast, a differential abundance analysis can be performed in order to see which taxonomies are different between different experimental conditions.
Our first, and likely most difficult project objective, was sample collection. Dr. Jones began by calling all (~30) vermicompost producers as described in the Education Section to obtain and clarify DNA sample collection and MidWest Lab sample collection. All producers were to send triplicate samples of at least 1 vermicompost for DNA analysis and MidWest Lab phys/chem analysis. Additional samples for DNA analysis were to be collected of precompost, manure, soils, etc on an individual basis depending on composting processes. Producer questionnaires were prepared on an individual basis to collect more specific data about the samples sent to us. All questionaires had the same questions especially for vermicompost in order to collect consistent information across the sample set. These were included with the sampling kits a long with sampling instructions. In the future I would recommend digital questionnaire data collection to increase data clarity and consistency. It would also drastically reduce time spent translating data to a digital medium later.
With the help of Ms. Doesken, nearly 30 sample collection kits with instructions. Samples were sent on ice because covid did not allow for in lab testing of a preservative. This change in plans caused problems, especially with the USPS issues, and summer temperatures. Several samples had to be recollected to ensure DNA sample viability. In the future I would use a preservative that allows DNA to be stable at room temperature for months. Examples of the sample collection kits and collection kit preparation are below.
Sampling kits were sent with a pre-paid return label for overnight shipping. producers were also told to use dry ice if available. All kits were mailed back to Dr. Jones who then stored DNA samples frozen until DNA extraction. Midwest Lab sample bags were forwarded for analysis. A total of 20 producers submitted tripicate samples for MidWest Labs analysis and over 25 producers have samples for DNA testing. Producer questions about their samples were also collected with the sampling. All metadata from each sample has been recorded in a large spreadsheet. This completes objective one of sample collection.
Objective 2: A total of ~250 samples are being processed for DNA microbial community analysis. DNA extraction of samples has been completed. Because of covid, we lost university laboratory access for the time being. I was still able to perform DNA extractions by loading a high throughput DNA extractions kit with all the samples . The extraction was then performed at an outside lab with high throughput extraction instrumentation. This was slightly more expensive as I had to pay for a 384 sample kit and only had 250 samples to run. However, because of the already large delay in getting this data it was a worth while trade off. (Pictures below)
The extracted DNA has also been amplified with the sequencing primers. Amplification success is tested by gel electrophoresis as evidenced below. A dark band indicates a successful amplification.
The DNA amplicon just needs to undergo barcode and adapter ligation in order to load it on the sequencing instrument. I am hoping to have sequencing data in the coming weeks would would complete this objective.
Objective 3: All Midwest samples have been processed and the data received. Due to the delay in sequencing these results have undergone an initial exploratory analysis which has been very positively received. Producers are already starting to see the value in comparing their vermicompost results and associated composting techniques. Each producer was told their respective number so the results could remain anonymous yet they could still compare to one another. Objective 3 is also completed.
Objective 4: Data analysis
Because of the delay in obtaining sequencing data, only a preliminary analysis has been done with the physical and chemical data. Initially, data was presented unsorted and minimally processed (below)
Producers already found this valuable, because it allowed vermicompost producers to get a sense of the range of values out their and how their own compared to others. There was quite a range of values despite nearly every producer being proficient in vermicomposting. We aim at exploring these differences more in relation to compostings inputs and techniques.
Hypotheses surrounding mainly manure content and treatment were explored as we thought this variable would be most apparent with this data type. Total manure was sorted left to right (Below)
Correlations were explored in regards to phosphorus and conductivity.
We noticed that some manures that were aged outside first had more nutrients generally but less salts likely due to weathering. Additionally Fish poop from pond demonstrates this as well (#8).
There was also a correlation in regards to pH and calcium(below).
While it is difficult to drawn any firm conclusions because calcium doesn’t directly affect pH, we think it could have to do with the amount and maybe type of soil getting into the vermicompost via compost inputs being explored to soil. These soils like have calcium carbonates which could buffer the pH of these systems above neutral (max 8.2). The amount of degradation and aeration may also play a roles as the more organic matter degrades, especially under anaerobic conditions, more organic acids are formed. This data will be more formally explored once the DNA sequencing data is ready.
We should get initial bioinformation results shortly after sequencing occurs, assuming the sequencing is successful. A more in depth analysis could take a few months. Figures will be shared on the blog as they are made and hopefully at least 1 virtual conference, and in an open source peer reviewed journal. Each produced will get a custom report and E&O material generated from the dataset.
Educational & Outreach Activities
At the beginning of the project, each producer was contacted over the phone to discuss their vermicompost operation and which samples for DNA sequencing would be most beneficial for research as well as their operation/business. About 0.5-1 hr was spent with each producer discussing how what to sample. While time consuming given their was about 30 producers, we learned a lot and got to understand them and their operation on a more personal level. It has been a great experience and it has been fascinating how versatile vermicomposting can be.
I have also had personal consultations over the phone, blog, and email ranging from topics of DNA sequencing, microbiology, new technology for vermicomposting, and best testing practices and labs. These producers have a lot of more ideas than I originally thought and I aim to help futher their own research endevors. So far 1 letter of support has been written to this end for an NSF grant.
Currently, most of the education and outreach is being done via the project blog: vermimicrobiomeproject.com. I am being a series of lessons on microbial ecology while we wait for DNA sequencing data. Project progress as well as an introduction to academic literature has also be logged there. There are currently 77 subscribers to the blog. In the last month there have been 110 post views. We aim on growing the audience now that we will have data soon.
While the initial part of the project was mean to be decentralized, covid has caused the NC State vermicomposting conference to be cancelled for 2020. I am currently trying to put together and help organize a virtual NC State conference with our travel money. We expect all of the education and outreach activities can be completed this calendar year including a conference presentation(s), a publication submitted, and educational materials distributed.