report

Abstract

The transition from the Digital Botanical Garden Initiative (DBGI) pilot project to the main Earth Metabolome Initiative (EMI) is a very important step towards adapting our collection and extraction methods to a wider spectrum of organisms or groups of specimens. This transitional project, undertaken as a Bachelor's thesis at the University of Fribourg, involves a comparative study of the metabolomic content of crops, weeds and invertebrates present in two different fields, one cultivated according to a biodynamic farming system and the other using conventional agriculture and only mineral fertilizers. The main objective of this study is not the metabolomic results themselves, although they are very interesting, but rather the adaptation of methods initially designed for data collection in a botanical garden to less controlled environments, with the aim of eventually setting up collection and extraction methods feasible in wild environments. This methodological adaptation is crucial to the successful scaling-up of the EMI, ensuring that robust and reliable data can be collected in diverse and variable natural environments, thereby facilitating the achievement of the initiative's overall objectives.

Background

Earth metabolome initiative

The Earth Metabolome Initivative (EMI) is a large scale open science project whose aim is to profile the metabolomic content of all organisms currently known on the planet Earth. This initiative have three main objectives : The first one is to protect biodiversity by creating molecular and metabolomic arguments, the second is to understand the chemical functioning of the biosphere and the last one is to benefit human society, notably through drug discovery or food sustainability. Understanding the chemical functioning of the biosphere is fundamental to developing effective conservation strategies. This understanding enables us to better protect biological diversity and guarantee the sustainability of the ecosystem services on which the survival of all species, including our own, and our well-being depend. Establishing a global molecular map of chemical content could help us to deep into new concepts such as "chemical endemism" or create some argument based on the chemodiversity level of certain zones. Ultimately, this will create a library accessible to all, enabling targeted research and protection of certain organisms and metabolites. These data will enable us to observe the chemodiversity of specific locations and thus add chemical arguments to environmental protection.

https://doi.org/10.5281/zenodo.10663625 https://doi.org/10.5281/zenodo.7486984

Digital botanical gardens initiative

Background

The digital botanical garden Initiative (DBGI) is an EMI pilot project whose aim is to extract, catalogue, analyze and create a digital library of metabolomic data from plants found in botanical gardens. This enables the methods to be tested and the bioinformatics resources needed for data processing to be created.

Why choose the botanical gardens for this pilot project? Botanical gardens, although human-made, are hotspots of biodiversity and the ideal place to test collection, extraction and analysis methods. They display a considerable number of plants in a small, accessible area near to university and laboratories supplies. Botanical gardens provide an initial approach and the methods needed to eventually collect different organisms from wild ecosystems (see Earth Metabolome Initiative). Their biodiversity also makes it possible to set up a library of chemical extracts, enabling the development of the IT tools needed to analyze these data.

Fribourg is privileged to have a university botanical garden where over 5,000 species are cultivated and conserved. This represents one of the largest systemic collections in Switzerland. The botanical garden and its research group play an active role in the conservation of rare and endangered species at cantonal and national level. At this point, around 2400 plants have already been collected, 1500 have been extracted and 1300 have been profiled with liquid chromatography mass spectrometry.

Tropical greenhouse

The tropical greenhouse (number 1 on the map) was sampled during my bachelor work. This greenhouse features decorative green plants, climbers, epiphytes and aquatic plants, all of which come from the rainforest. In all, some 280 plants are present with no specific taxonomic or provenance links.

Map of Fribourg botanical garden. The tropical green house is The tropical greenhouse is located at point 1

DBGI Methodology

Collection

To collect plants in this greenhouse the DBGI methodology was applied. The collection method consists of cutting one or several leaves (depending on the size), wrap them in a coffee filter and put them in a falcon tube with a perforated cap. Each falcon tube is pre-labelled with a unique code assigned to the extract and use as an identifier throughout the experiments. The tube are then store in liquid nitrogen and finally put in the freezer (-80°C) if they are not freeze-dried immediately. Each collected plant, will be report as a point on a qfield map, with several additional data such as the precise location, identification of the plant if possible and minimum five photos. These photos show :

1) The plant identification panel
2) The plant identification panel with the labelled falcon tube
3) A general photo of the plant
4) A detail photo of the plant
5) The cutting area with the scalpel
6) An additional photo can be taken if necessary

Extraction

All plant extracts are freeze-dried and store in falcon tube with non perforated cap and store in a labelled container.

The first step of the process is to weigh out 50 micrograms of each extract and place it in an Eppendorf 2ml with rounded bottom. We accepted an 5% of the weight as an error (±2.5 micrograms). Then add thee metal beads (4mm) inside the tube which will be used to crush the extracts. The plant remains are returned to the falcon and stored in the container.

Dry plant ready to be weighed

The sample are then passed in the Retsch machine (shaker) for 2.5 minutes at 25Hz. In this process, the plants are ground to a fine powder.

Eppendorf tubes with grounded plants

Once the plants have been ground, the tubes are carefully reopened and 1.7 ml of the extraction solution is pipetted into each tube. The extraction solution mix is composed of 80% methanol, 20% distilled water and 0.1% formic acid. The tubes are then returned to the Restch machine to be shaken again for 2.5 minutes at 25Hz.They are then centrifuged for 2 minutes at 13'000RPM to separate the supernatant from the plant deposits.

Finally, the supernatant is collected and pipetted into a vials tube with hermetically sealed caps. Approximately 1.4 ml of supernatant can be recovered. These tubes are then stored in a labelled container at -80°C.

Vials tube containing extracts of various plants from the tropical greenhouse

Analysis

120 microliters of plant extracts are taken from the vials to form aliquots into new vials containing an insert. Those new vials are closed with a slipped cap. These aliquots are analyzed by liquid chromatography mass spectrometry. Spectra are then analyzed using bioinformatics tools.

Aliquots ready for the mass spectrometry

Inaturalist resources

All photos taken at the time of collection are then upload on iNaturalist platform. This help us to check the identification of the plant by external experts and enables us to correct our identifications if necessary. It also benefits iNaturalist, as some plants not included in the database have been added by our request

Here a link to the list of the plants collected in the tropical greenhouse https://www.inaturalist.org/observations?nelat=46.793262993305206&nelng=7.155081027961212&page=2&place_id=any&subview=map&swlat=46.79316199515332&swlng=7.154816830373245&user_id=dbgi&verifiable=any

Sample tracking

Because of the large number of samples, it is important to have reliable sample tracking. As said before, as soon as the samples are collected, each plant is assigned a unique code. This code will follow the plant through each stage of extraction, analysis and storage. This unique code will be report on each different tube during the extraction process and all data relating to this sample are added to the database. By consulting this database, we can quickly find out which plant has already been extracted, how many times it has been extracted, the exact weight that have been extracted and in which container the various tubes are stored. The whole process is automated, and the database automatically updated.

Transitioning from DBGI to EMI

Background

The problematic

In this section, we will take a closer look at a transition project between DBGI and EMI. Until now, the methods implemented were adapted to the collection, extraction and profiling of plants from the botanical garden. However, it is important to adapt this methodology in order to broaden our possibilities and move closer to a solid methodology for the EMI project. There are several aspects that complicate the method in uncontrolled environments. Firstly, unlike in a botanical garden, plants and other organisms are not labelled in a wild environment. This makes the use of the Inaturalist platform as a control by others all the more important. Sufficient good-quality photos should be taken of each sample. On the other hand, the collection of new types of samples implies the implementation of new collection and extraction methodologies. And finally, data processing methods also need to be updated, as we take samples containing many different organisms. The links between these need to be established in the data processing.

A cultivated field is an ideal place to start the transition. The composition of the crops is known, and a certain biodiversity is present within the field. Other organisms such as insects and soil macrofauna can also be collected quite easily.

The DOK trial

Historical background

In the 1970s, discussions on food production in relation to the environment and yield had already been generating debate for several decades. Some pioneers of organic farming practice and scientific from the Research Institute of Organic Agriculture (FiBL) and from the Swiss federal research center for agricultural chemistry and environmental hygiene founded the DOK trial in 1978.

The DOK trial is a long term experiment comparing three cropping systems. DOK stands for BioDynamic, bioOrganic and Konventionell. These three farming system have been studied for 45 years. The DOK trial mimics organic and conventional certified farming methods that are practiced in Switzerland. Since the start of the trial, a large number of scientific publications have been produced thanks to this long-term experiment. Important topics such as climate adaptation, biodiversity, soil quality and yield have been studied on this site.

The site

The DOK site is located in the canton of Baselland, halfway between Biel-Benken and Therwil. The area is located in the Rhine Valley. The floor of this Valley are filled with a layer of gravel and covered by loess material (fine Calcareous silt). This large layer of1 to 1.3 m deep created a fertile soil, that allows water to rise through capillary transport during the summer. The climate is favorable for agriculture, with temperatures averaging 9.7° C until the end of the last century and current annual rainfall of 872 mm. However, average temperatures have been rising steadily in recent decades, with an average of 11.2°C between 2010 and 2020. Climate change and crop resilience are also being studied at DOK, with a view to future extreme weather events (drought, heavy rain, hail, etc.).

Experimental design

The field is divided into 96 plots of 5x20 m. A total of 8 different treatments are applied. Both biodynamic (BIODYN) and bioorganic (BIOORG) systems receive slurry and farmyard manure. The conventional system (CONFYM) also receive slurry and farmyard manure but additionaly it is also treated with mineral fertilizers. This three treatment are applied with two different intensities : 0.7 and 1.4 LU per hectare. A second conventional treatment (CONMIN) system is only treated with mineral fertilizer. In addition, there is control treatment (NOFERT) without any fertilizer. Manure treatment differs between systems, and consequently organic carbon and nitrogen inputs also differ. The conventional system (CONFYM) uses stacked manure, the bioorganic system (BIOORG) uses rotted manure and the biodynamic system (BIODYN) uses compost manure. These are aerated more intensively and for longer, respectively. Mineral fertilizers comply with Swiss Federation regulations for concentrations of calcium-ammonium-nitrate, urea, potassium chloride and triple super phosphate.

Plant protection in both conventional systems consists of the application of herbicides, fungicides, insecticides, and molluscicides. In BIOORG and BIODYN the plant protection follows the standards (Lampking 1990). In addition, the BIOORG system treats potatoes with copper sulfate (CuSO4). The Bacillus thuringiensis bacterium is sprayed against potato beetles in BIOORG and BIODYN systems. The NOFERT system uses the same protections as the organic systems.

3 crops are planted simultaneously (columns A, B and C in the schema) and a 7-year crop rotation is carried out in the plot following this sequence since 2013 : Maize, Soybeans, Vinter wheat 1 + catch crop, Potatoes, Winter Wheat 2, Grass-clover 1, Grass-clover 2. Grass-clover mixture is composed of Trifolium pratense 6%, Trifolium repens 12%, Dactylis glomerata 17%, Festuca pratensis 36%, Lolium perenne 21% and Phleum pratense 12%.

Field setup of the DOK trial with spatial orientation of blocks, columns, rows, and subplots. Each year three crops are cultivated in the subplots A, B, and C. N, NOFERT, unfertilized; D, BIODYN, biodynamic; O, BIOORG, bioorganic; K, CONFYM, conventional with farmyard manure; and M, CONMIN, conventional purely mineral fertilization. Organic fertilization: 0.7 and 1.4 correspond to organic fertilization at 0.7 and 1.4 livestock units per hectare. (H-M Krause et al. 2022)

Previous experiments and results obtained

Around 200 studies based on the DOK trial have been published to date. Some interesting results from different major studies conducted on the DOK trial site :

On average, organic crop yields are about 20% lower (https://doi.org/10.7717/peerj.15000) (although yields of specific crops vary) compared to the conventional system receiving mineral fertilizers, while nutrient inputs in the conventional system are 60% higher for mineral nitrogen, 45% for potassium and 40% for phosphorus (https://www.cabidigitallibrary.org/doi/full/10.5555/20063115035).

Organic systems reduce fertilizer and energy inputs by 34 to 53% and pesticides by 97%. An improvement in soil fertility and biodiversity is observed in organic system plots. These plots are less dependent on external inputs. (https://doi.org/10.7717/peerj.15000)

The soil organic matter (humus) content remained stable only in the biodynamic system during the first 21 years of the trial. All other systems showed a decline in humus. This difference is significant between the biodynamic and the purely mineral conventional systems. The soil microbial biomass showed 25% lower values in the BIODYN and CONFYM while system without manure application (CONMIN) showed 34% lower values. (https://doi.org/10.1016/j.agee.2006.05.022)

In term of nutrition quality, a broad spectrum of molecules relevant to human nutrition have been identified in wheat including analysis of : seed weight, protein content, phosphate levels, antioxidative capacity of phenols, fiber, fructan, oxalate and phytic acid. A clear difference in grain weights from the different farming systems was observed. In fact, thousand-seed weight and protein were significantly lower in unfertilized wheat. On the contrary, total oxalate was significantly higher in these plots. For the rest, and for the majority of substances of nutritional interest, no significant differences were observed between biodynamic, organic and conventional farming systems. (https://core.ac.uk/outputs/235697697/?utm_source=pdf&utm_medium=banner&utm_campaign=pdf-decoration-v1)

In terms of energy efficiency, organic crops require 30-50% less energy to produce per unit area. This includes the energy needed to produce fertilizers and pesticides. Taking into account the higher yield per unit area in conventional systems, the energy used per unit yield is still 19% lower for organic farming system. 



Long-term measurements show a nitrogen deficit in all cultural systems. This means that the cultivated plants used more nitrogen than what was provided to them by the fertilizer applied. This additional nitrogen comes from soil mineralization, nitrogen fixation by legumes and their symbiosis, and aerial deposition. Regarding the phosphorus and the potassium, the deficits are much greater in biological and biodynamic systems than in conventional systems.

https://www.fibl.org/en/locations/switzerland/departments/soil-sciences/bw-projekte/dok-trial

DOK project experiments

Global experimental design

For our experiments, we decide to compare the BIODYN and the CONMIN systems which are probably the two most different systems in terms of cultivation methods. 4 experiments were design, that have been replied 4 times (one in each replicate plots). The aim was to see if there is any metabolomic differences in the crops (Wheat) and the other organisms present around the culture such as weeds and invertebrates (insects, mollusks, annelids and other soil macrofauna). Crop A was used in all 4 experiment, which this year was the montalbano variety of wheat. Our experiments were carried out in plots with the following numbers: 6, 40, 56 and 96 for CONMIN plots and 12, 46, 50 and 90 for BIODYN plots.

Collection review

The collection took place on April 18 from 1:30 p.m. to around 7 p.m. at the DOK trial (Localization : 47.502578, 7.539575 (Therwil, Basel canton)). The collection went smoothly, with cool weather. It didn't rain in the afternoon, but conditions were damp because it had rained that very morning. We didn't find many insects on site, probably because the weather was too cold for them, but we did collect a few, along with slugs and soil macrofauna. All the wheat fields (with no difference between cropping systems) had been infected by a fungal disease, which may have an influence on the metabolomic results.

Collection methodology

Equipment list for one afternoon collection

80 Falcon tubes with label
80 Falcon tubes (reserve without label)
80 Label (reserve)
16 big zip lock bag
1/pers tablet / smartphone with Qfield application and the project synchronized
1/pers scalpel + a spare in case
Gloves : 5 pairs/pers (each size)
Liquid nitrogen
Thermos for the liquid nitrogen
Shovels
Measuring tape
Big boxes
1/pers Ethanol 70% spray
Sardines
Rope / string
External battery with cables for each type of device
Indicator arrows
160 Coffee filter
Scissors
Sharp knife
Soil corer

The following paragraphs presents the 4 experiments carried out on the site, with the aim of testing different collection and analysis methods.

20x20 crushed : chemodiversity

Experimental concept

We randomly select a 20cm x 20cm square in the plot, if possible avoiding edge effects. We cut everything in this square above ground level. The cultivated plant as well as any weeds growing in between and possibly insects, invertebrate or mushroom that were present in the area. All this samples are store in a Ziplock with an unique code. They will then be extracted as one entities to rate and qualify the chemodiversity of the parcelle.

Methodology

We applied the methodology once in each of the 8 plots. Each type of agriculture is replicated four times on the DOK trial. That is allowing us to have four replications of above-ground-biodynamic specimens and four replication of above-ground-conventional specimens.