Valorizing Andean microbial diversity through sustainable intensification of potato-based farming systems

Final Report Summary - VALORAM (Valorizing Andean microbial diversity through sustainable intensification of potato-based farming systems)

Executive Summary:
VALORAM aims at exploring and valorizing Andean soil microbial diversity for the development of alternative, efficient technologies and crop management practices to improve the sustainability and productivity of Andean cropping systems benefiting rural farming households. The project will focus on potato because of its vital importance for small-scale farmers in the central Andean highlands. The participants will use genomic, metagenomic, proteomic and metabolomic analysis to identify novel traits of microorganisms and to characterize beneficial soil microbial communities, to achieve the objective. The project specific aims are to (1) explore the agro-ecosystem functions of soil microbes in potato-based cropping systems and preserve the components of this microflora in international culture collections, (2) elucidate the role of rhizosphere microbial communities in promoting plant growth, suppressing soil borne diseases and priming plant biotic defenses, developing eco-efficient technologies/products for sustainable crop production systems, (3) develop applied technologies and knowledge-based systems to improve the sustainability and resilience of potato based cropping systems for the benefit of the indigenous farmers and (4) promote the exchange of scientific knowledge and technologies among partners and the LA scientific community to stimulate research in this area and support the continuous development of crop production technologies. The strategy for VALORAM implementation is to engage LA and EU partners in developing and further strengthening collaborative research activities in order to sustainably improve potato-based systems. This is supported by a multidisciplinary team of experts with highly complementary skills and based on a robust management structure with an efficient workshop and communication program. The results will directly benefit the local partners and may also contribute to improve organic potato production in the EU.
The main results achieved are as follows:

- Establishment of metagenomics libraries;
- Development of isolation protocols of indigenous potato endophytes and rhizosphere fungi (principally AMF) and bacteria;
- Cultivation-independent diversity assessment of indigenous potato rhizosphere fungi and bacteria;
- Establishment of long term preservation of isolated AMF and bacteria in reputable EU germplasm collections;
- Development of protocols for the preservation of complexes of microorganisms;
- Isolation, screening and characterization of bacterial and AMF isolates for beneficial traits such as plant growth promotion and disease alleviation under in vitro and greenhouse controlled conditions with the establishment (1) of protocols using GC-MS for the measurement of bacterial volatiles in pure culture (2) of gnotobiotic cultures for screening the priming activity of bacterial and AMF isolates which had been characterized for antagonism and plant growth promotion in vitro using a transcriptomic and proteomic approach (3) of protocols for screening for plant growth promotion and disease suppression using a microhydroponic system in vitro;
- Establishment of protocols for metagenomic analysis;
- Development of methods for measurement of glycoalkaloid content in potato tubers;
- Establishment of a database with edapho-climatic conditions and main abiotic and biotic factors characterizing the experimental pilot sites;
- Determination of microbial characteristics appropriate for product formulation;
- Production of stock cultures of microbial inoculants for transfer to SMEs for mass-propagation and release;
- The formulation and quality control of bacterial and AMF cultures for field tests;
- Publication of field experimental results on the performance of bacterial and fungal strains;
- Molecular tracing of AMF within field and determination of plant, environmental and agricultural practices impacting the inoculated organisms and the indigenous microflora;
- Dissemination of knowledge and information on potato-based systems using microbial inoculants by (1) set-up and development of a web-site, (2) organizing workshops/trainings with organizations and institutions of the Andes on potato-based systems using microbial inoculants;
- Documenting and publication of research results and knowledge.

Project Context and Objectives:
Project context:

The Central Andean Highlands are the center of origin of the potato (Solanum spp.) and one of the centers of diversity with more than 3800 different native potato varieties grown by farmers (CIP, 2007). For centuries, indigenous subsistence farmers have selected landraces adapted to harsh and variable agro-ecological conditions and managed the land with minimal inputs to achieve sustainable yields. Hence, potato is grown and adapted to a huge variety of agro-ecosystems, ranging from the coastal valleys of Peru to the Altiplano 4,000 m asl in Bolivia with variable climatic conditions, soils and land use systems. Nowadays farmers employ a wide range of cropping systems, comprising high input systems with fertilizer applications of up to 500 kg NPK ha-1 and frequent pesticide use to low-input subsistence systems with limited organic manure application and almost no pesticides. Yield levels are medium to low, ranging between 15 and 20 t ha-1 for high input and 5 to 8 t ha-1 for low input systems (Davies Jr. et al., 2005; MINAG, 2007).

Potato is Peru’s most important crop. It covers approx. 270,000 ha (mainly in the Andean highlands) with a value of US$ 347 million, contributing an annual 7.7% to Peru’s GDP (FAOSTAT, 2008). Also in Bolivia it is the largest source of agricultural gross value with a total value of US$ 148 million and 130,000 ha of cultivated land; it is Bolivia’s third largest crop by area planted, after soybean and corn. In Ecuador about 50,000 ha are cultivated with potato, which produce around 450,000 tons of tubers with a total value of US$ 120 million. Average consumption in these countries range between 30 and 70 kg per person and year, however, in rural districts people consume up to 250 kg per person and year. Hence potato is the most important subsistence and commercial crop in the Andean region and hundreds of thousands of farming households depend on it for their survival.

Demographic pressure and demand for potato is increasing, especially with high commodity prices for cereal crops. Therefore more land is taken into cultivation and fallow times as well as rotation periods are reduced. As a result pests and diseases increasingly affect potato crop reducing tuber yields but above all tuber quality, rendering them often unfit for commercial use. In these conditions also soil fertility management has a prominent role for crop production. An appropriate soil fertility management guarantees a certain level of yield stability even in difficult agro-ecological conditions and increases the overall resilience of the cropping system to adverse biotic and/or abiotic stresses (Fliessbach et al., 2007, Karlen et al., 1997, Mallory and Porter, 2007). Furthermore the enhancement of food production that is required to feed the growing population would have to come from “land saving” technologies oriented to intensify the production of the system, converting marginal lands into productive areas and restoration of degraded areas (Lal, 2000). These types of technologies include, for example, increasing the efficiency of nutrient use by the crop, control of biotic stresses and soil-water conservation practices, particularly under rain fed conditions such as in the Andes. The supply of new, improved varieties, accessible through the genetic resources contained in the totality of potato landraces in the Andean area may, in this context, represent convenient options to produce yield surpluses to generate income that enables farming community to move out of - and stay out of - poverty. In support of this objective, small-scale farmers need sustainable yield improvement and input-saving technologies.

Soil microbes are key-components of any agricultural system and exert multiple functions, from detrimental (as pathogens) to beneficial (in particular in low input systems, e.g. plant growth promoters and pathogen antagonists), impacting yield and quality of food. Nowadays, increasing attention is devoted to rhizosphere and endophytic microbes which play a central role in promoting plant growth and health via: (1) acquisition and recycling of nutrients important to plant growth, (2) modulation of plant hormonal balance, (3) direct or indirect protection of the plant from detrimental organisms (e.g. pathogens), (4) protection against abiotic stress (e.g. drought, heavy metals), and (5) improvement of soil structure. It has been stated that “the ultimate agricultural goal in studies of the biology of the soil-root interface, must be the manipulation of microorganisms in this zone to increase plant health and growth” (Rovira, 1979). Research should aim at improving our knowledge of the interactions between plants and microbes and of sustained management of these microbes to benefit the plant-food-consumer chain. This knowledge could help to reduce excessive use of agrochemicals alleviating hazardous effects of agricultural production on the environment.

The approach combining suitable potato ‘genotypes’, proper land management and inoculation with appropriate beneficial microbes (e.g. arbuscular mycorrhizal fungi (AMF), plant growth promoting rhizobacteria (PGPR), pathogen antagonists) is therefore a major challenge for the sustainable intensification of potato-based farming systems in the Andean area. This combines improving soil fertility, plant resistance, and plant nutrition. This approach is in agreement with the UNESCO (2008) policy that potato-based agricultural systems need a continual supply of new, improved varieties, albeit at the same time potato diversity should be protected, conserved and exploited, and with the EU Code for “Good Agricultural Practices” that seeks to minimize environmental damage associated with agricultural practices through the reformed common agricultural policy (CAP) and with various EC directives. Within this project, we strongly support the sustainable intensification of potato based cropping systems focusing attention on solutions that will stop or reverse the loss of natural resources (UNESCO, 2008). As proposed here, a better use of native or improved potato varieties and developing eco-efficient technologies (i.e. based on beneficial microbes and encouraging beneficial communities) while combining agricultural practices more respectful of natural diversity will decrease current pressure on land use, erosion, and ecological imbalance, at the same time alleviating poverty by increasing incomes of the rural farming households.

Main objectives:

The overall objective of VALORAM is to promote the sustainable development of potato-based systems in the inter-Andean valleys and Altiplano areas to focus on cropping systems that will make use of natural microbial resources as inputs to improve production of high quality potato crops.

In more details, the objectives of VALORAM are

1. To explore the soil microbial diversity in potato-based cropping systems of the high Central Andean Highlands and preserve components of this microflora in local and international culture collections.

Tactics to reach the objective:

- Selection of appropriate field sites for sampling and field trials according to established agro-ecological factors.
- Metadiversity and metagenomic approaches to analyse the soil microbial flora at the different sites during the crop growth season and mining of the data for markers of beneficial soil communities.
- Use of genomic and classical techniques to identify fungi and rhizosphere bacteria.
- Use of in vitro root organ culture to maintain AMF isolates.
- Develop methodologies for ultra-preservation of microbial diversity

Key achievements:

- A general and fundamental improved knowledge on microbial diversity via assessment and comparison of the microbial flora (using metagenomic techniques) in potato roots and rhizosphere in traditional potato fields in the presence of landrace(s) and improved potato variety(ies)
- Isolation and identification of indigenous potato endophytes and rhizosphere fungi (especially AMF) and bacteria in traditional potato fields.
- Long-term maintenance of isolated fungi (especially AMF) and bacteria within BCCM/MUCL and BCCM/LMG international germplasm repositories and mirrored in the local repositories of the three Andean partners.

2. To elucidate the role of rhizosphere microorganisms and communities in promoting plant growth, suppressing soil borne disease and priming plant biotic defenses, developing eco-efficient technologies/products for sustainable crop production.

Tactics to reach the objective:

- Isolation, screening and characterization of microbial isolates for beneficial traits such as plant growth promotion and disease alleviation under ex-situ controlled conditions.
- Using gnotobiotic cultures, genomics, proteomics and metabolomics (incl. volatilomics) to characterise the growth promoting and pest/disease suppressive traits of selective isolated rhizosphere bacteria
- Metagenomics and novel screening approaches for the identification of novel, desirable traits.
- Testing and evaluation of selected isolates for their plant growth promoting and protecting properties in field trials, studying their interactions with potato varieties, native rhizosphere microorganisms and agro-ecologies.

Key achievements:

- Adding value to microbial inoculants as agricultural inputs to improve sustainable intensification of potato-based farming systems.
- Enhanced appreciation among farmers and consumers of the principles of sustainable development in agricultural practices and the necessity of protecting the natural resource base on which mankind depends, which becomes a global and major factor during the next 10-30 years.

3. To develop applied technologies and knowledge-based systems to improve the sustainability and resilience of potato based cropping systems for the benefit of the indigenous farming communities.

Tactics to reach the objective:

- Multiplication of fungal and bacterial inoculants for mass-production by selected local sub-contracting companies (to be subcontracted by the Latin-American partners).
- Quality-control of the mass-produced inocula.
- Establishment of field trials with the mass-produced inocula to assess beneficial fungal and bacterial inoculants.
- Testing the ecological impact of the improved soil management systems including a tracking of the survival of the inoculants in Andean field trials.

Key achievements:

- Adding value to microbial inoculants and formulations to improve production in potato-based farming systems.

4. To promote the exchange of scientific knowledge and technologies, created by the project, among partners and the Latin American scientific community to impulse research in this area and support the continuous development of crop production technologies.

Tactics to reach the objective:

- Organizing and supporting workshops/trainings with interested organizations and institutions of the Andean region (NARS, Universities NGOs etc.) on potato-based systems using microbial inoculants.
- Documenting and publication of research results and knowledge.
- Linking the consortium with local farmer associations for transfer of technology.

Key achievements:

- Dissemination of the results of the project – both beneficial microorganisms and knowledge – using the networks of the consortium and a project website established and maintained by the consortium.

References

CIP – International Potato Center (2007). Facts and figures: improvement and conservation. http://www.cipotato.org/pressroom/facts_figures/improvement_conservation.asp
Davies Jr FT, Calderon CM, Huaman Z and Gomez R (2005). Influence of a flavonoid on
mycorrhizal acitivity in the highlands of Peru. Sci. Hort. 106:318-329.
Faostat, 2008. http://faostat.fao.org
Fliessbach A, Oberholzer HR, Lucie G and Maeder P (2007). Soil organic matter and
biological soil quality indicators after 21 years of organic and conventional farming.
Agricult. Ecosys. Environ. 118:273-284.
Karlen DL, Mausbach M, Doran J, Cline R, Harris R and Schuman G (1997). Soil quality: a concept, definition and framework for evaluation. Soil Sci. Soc. Am. J. 61: 4-10.
Lal R (2000). Soil management in the developing countries. Soil Sci. 165:57-72.
Mallory E B and Porter G A (2007). Potato yield stability under contrasting soil management strategies. Agron. J. 99:501-510.
MINAG - Ministerio de Agricultura de Peru (2007). Importancia de las papas nativas.
http://www.portalagrario.gob.pe/papa_datos.shtml
Rovira AD (1979). Biology of the soil root interface. In: Harley J.L. Russell R.S. (Eds). The soil root interface. Academic Press London, pp. 145-160.
UNESCO (2008). International Year of the Potato. http://www.potato2008.org/en/aboutiyp.

Project Results:
The project was constructed in seven work packages (among which five were S&T) to reflect the complexity and interdisciplinarity of the various tasks. The overall objective was to promote the sustainable development of potato-based systems in the Central Andean Highlands and to focus on cropping systems that will make use of natural microbial resources as inputs to improve production of high quality potato crops. This research objective was met by detailed soil microbial flora analysis (using metagenomic techniques) of various soil management systems (WP3). Culture collections were established both in reputed international collections and in the countries of origin (WP4) to preserve the local diversity. Microorganisms with new, desirable properties relevant for a sustainable crop production in the Andes were characterized (WP5). Promising management systems as well as biofertilizers and pest control agents were selected and further tested in field trials to elaborate their applicability and benefits in comparison to conventional practices (WP6). Field testing was accompanied by a comprehensive ecological impact assessment in order to warrant that novel and recommended practices have either a benefit on biodiversity and soil fertility or a lower impact than conventional practices (WP7). The major results obtained in these 5 WPs (WP3 to WP7) are described below. Graphs, Figures, Tables are listed in an attached document under the heading “supplementary material” for Annex I (for WP3) to Annex V for WP7.

WP3 – Exploring microbial biodiversity in contrasting Andean agroecosytems

Graphs, Figures, Tables are in supplementary material under heading Annex I.

• Main result 1: Diversity of rhizosphere bacteria in Andean agroecosystems assessed by cultivation-independent analysis

A. Prior to sampling, storage and analysis protocols were established taking into consideration that samples had to be shipped from LA countries to Europe. The elaborated DNA isolation protocol for the analysis is shown below.

In: Pfeiffer S, B Mitter, A Oswald, B Schloter-Hai, M Schloter, S Declerck, A Sessitsch. Core, dynamic and opportunistic rhizosphere microbiome components of potato cultivated in the high Andes of Peru. Submitted to ISME J.

DNA extraction from rhizosphere soil for the analysis of diversity and functionality of bacterial communities

Material Solutions provided from FAST DNA soil kit (MP Biomedicals, Solon, OH)

• Lysing Matrix
• Filter columns
• Sodium Phosphate Buffer
• MT Buffer
• Protein Precipitation solution
• SEWS
• Pyrogenase free water

Procedure

1. Weigh in rhizosphere soil (approx. 0.5 g)
2. Add: 678 µl SodiumPhiosphateBuffer(Kit), 120 µl MT buffer (KIT), 300 µl PCI (Phenol-Chloroform-Isoamylalcohol; fridge, pH adjusted). Work under Fume hood!
3. bead beating: 30 seconds 5.0; Centrifuge at 14000 x g; 5 min (4°C); transfer upper phase into a clean 2 ml Eppendorf tube (put on 4°C)
4. repeat bead beating and centrifugation step twice
For the second and third bead beating, add amounts are 200 µl Sodium Phosphate Buffer (kit), 50 µl MT buffer (kit) and 200 µl PCI. After the third bead beating, centrifuge for 10 min
5. Pool supernatants of bead beatings
6. Add max. 350µl PCI; Invert Eppendorf tubes for five minutes; centrifuge at 14000 x g 10min (4°C)
7. Transfer upper phase; add 580µl PCI; shake for 5 min; centrifuge at 10000 x 5 min; 4°C
8. Transfer supernatant; add 580µl CHISAM (24:1) ; invert for 5min; centrifuge at 10000 x 5 min; 4°C (Be careful to transfer ONLY upper phase)
9. Proceed with the manufacturers’ protocol from FAST DNA soil kit (MP)
a. Add 250µl PPS solution, shake 10 times per hand
b. Centrifuge for 5min at 14000xg
c. Transfer supernatant into 15 ml Greiner tube, add 1ml DNA binding Matrix (MP) and invert strongly by hand for 2min; wait for 3 min for settlement of binding matrix
d. Discard 600-700 µl of supernatant (be sure not to touch the DNA binding matrix (MP)
e. Resuspend binding matrix and transfer 700 µl (800 µl if more than 2 ml in Greiner tube) of the solution into filter columns; centrifuge for 1min at 14000xg; discard flow through; Transfer rest of binding matrix and centrifuge again at 14000xg for 1 min; discard flow through
f. Resuspend DNA binding matrix in the filter column with 500 µl of SEWS solution using the pipette tip
g. Centrifuge at 14000xg /1min; discard flow through; Without any addition of liquid, centrifuge at 14000xg for 2 min, discard the collection tube and transfer the filter column into a new clean collection tube; wait for 5 minutes
h. Resuspend binding matrix into 100 µl of DES (MP) using the pipette tip
i. Incubate suspended DNA binding matrix at 55°C for 5 min on a heat block
j. Centrifuge at 14000xg for 5min; DNA is now in the flow through; discard filter column

To clean the sample from low molecular PCR inhibiting substances

10.Agarose plug digestion
a. Low-melting point (LMP) agarose 1.6%: 0.32 g in 20 ml; melt in microwave
b. Immediately pour 50 µl DNA + 50 µl 1.6% warm LMP agarose into 100 µl plug molds (Bio-Rad)
c. Put the plug molds into the fridge and wait for at least 15 min until the agarose is solidified
d. Transfer the solid agarose plug into 2 ml Eppendorf tube containing 2ml TE buffer (200 mM EDTA; 10mM Tris) and wait at RT until low molecular contaminants are diffused into TE buffer (~3h);
e. Exchange buffer until coloration deriving from humic acids has disappeared
f. Melt agarose plugs at 70°C for 5 min, and dilute 1:20 or 1:40 into UV treated H2Obidist.

B. To determine the key bacterial groups to be analyzed, initially clone libraries of pooled rhizosphere DNA from each LA country were constructed and sequenced. Five dominating groups (Alpha-, Betaproteobacteria, Actinobacteria, Firmicutes, Bacteroidetes) were selected for further analysis. As a next step the specificity of published group specific primers (planned to be used for T-RFLP and qPCR analysis and targeting the above mentioned taxa) was tested and shown to lack specificity. Therefore, we designed new and improved primer sets and validated their specificity and sensitivity particularly considering their suitability for the rhizosphere soils sent by LA partners. These primer pairs are now published by Pfeiffer et al. (2014). Furthermore, different primer pairs were evaluated regarding their suitability to be applied in the analysis of bacterial endophyte community structures (Ghyselinck et al., 2013).

In: Pfeiffer S, M Pastar, B Mitter, K Lippert, E Hackl, P Lojan, A Oswald, A Sessitsch. 2014. Improved group-specific primers based on the full SILVA 16S rRNA gene reference database. In Press in Environmental Microbiology.
In: Ghyselinck J, S Pfeiffer, K Heylen, A Sessitsch, P De Vos. 2013. The effect of primer choice and short read sequences on the outcome of 16S rRNA gene based diversity studies. DOI: 10.1371/journal.pone.0071360. PLOS One.

C. These new primer pairs were then applied to quantify 16S rRNA genes derived from Alpha-, Betaproteobacteria, Actinobacteria, Firmicutes and Bacteroidetes as well as for community fingerprinting by T-RFLP. We performed qPCR (in triplicates) and T-RFLP analysis on 175 rhizosphere samples, fully covering the 3 countries, 4 altitudes and 3 plant development stages. We observed a significant correlation between Betaproteobacteria abundance and soil organic matter (P<0.05). Firmicutes abundance was correlated to the amount of Phosphorus (P<0.05) and to the altitude (P<0.001). Alphaproteobacteria and Bacteroidetes were positively correlated (P<0.05) to Potassium. Statistical T-RFLP analysis of the five taxonomic groups showed no significant influence of the vegetation stage or sampling sites on Alpha-, Betaproteobacteria, Actinobacteria, Firmicutes. The reason for this might be that these groups account for a greater proportion of the core microbiome, which was revealed by 454-sequencing (see the next point of bacterial community analysis). The cultivar had a significant influence on the bacterial diversity (P<0.05) only for Bacteroidetes (See Figure 1 in supplementary material Annex I).

In: Pfeiffer S, et al. Manuscript in preparation.

D. To obtain more in-depth information on alpha- and beta-diversity and an improved understanding on how rhizosphere microbial communities are shaped 454-sequencing of 16S rRNA gene amplicons derived from 45 Peruvian rhizosphere samples was performed, distributed over 4 runs. Samples included 5 different rhizosphere samples respectively at 3 vegetation stages (emergence, flowering, senescence) at 3 sites - Pazos (4075 m.a.s.l.) Sincos (3245 m.a.s.l.) and Sicaya (2750 m.a.s.l.). Sequencing was performed at the Helmholtz Centre in Munich. In total, 436 863 sequences were received after quality filter processing. We performed sequence processing using UPARSE (http://drive5.com/uparse/) yielding a total of 5264 distinct, non-singleton, non-chimeric operational taxonomic units (OTUs). Statistical analysis revealed that the bacterial diversity of the potato rhizosphere was stronger influenced by plant development than by edapho-climatic conditions (See Figure 2 in supplementary material Annex I). Through application of least discriminant analysis we could determine an increased relative abundance of Gemmatimonadetes and Sphingobacteriales at the Pazos field site, which is more acidic, and has higher levels of nutrients and soil organic matter than the other field sites. Abundance of Acidobacteria was also significantly higher at the Pazos field site, with the exception of subdivision 6 / Holophaga, which were significantly more abundant at the Sincos field site. We defined a core microbiome, which consisted of bacteria which were consistently found across field sites and followed the same dynamics through plant development (See Figure 3 in supplementary material Annex I). The most dominating bacterial species of the core microbiome were found to be evenly distributed during plant growth, despite significant community changes between the vegetative and the germinative phases of plant development. Furthermore, some taxa of the core microbiome were constantly found in rhizosphere communities associated with potatoes cultivated in Europe, indicating a close association with S. tuberosum.

In: Pfeiffer S, B Mitter, A Oswald, B Schloter-Hai, M Schloter, S Declerck, A Sessitsch. Core, dynamic and opportunistic rhizosphere microbiome components of potato cultivated in the high Andes of Peru. Submitted to ISME J.

• Main result 2: Abundance and diversity of rhizosphere bacteria conferring N2-fixation potential

To obtain functional information on functional groups the diversity of nifH genes responsible for N2-fixation was assessed by qPCR and sequence analysis. Different abundances and communities of nifH genes were found at different plant growth stages, however, patterns followed different trends in different countries and sites. Other genes potentially involved in beneficial plant-microbe interactions such as genes encoding ACC deaminase were considered, but no suitable primers could be identified targeting the broad diversity of known genes.

• Main result 3: AMF diversity assessed by cultivation-independent analysis

Initially AMF diversity was addressed by cloning, RFLP analysis and subsequent Sanger sequencing of a rRNA operon gene fragment of >1.5kb (SSU200bp-ITS-regioncomplete-LSU800bp) described in Senés-Guerrero et al. (2013) for Peruvian root and rhizosphere soil samples. The LMU reference sequence database could be extended with 208 new and divergent sequences from the VALORAM samples. The clone library and Sanger sequencing approach revealed that the AMF communities in rhizosphere soil and roots differed significantly. Altogether, 20 species belonging to 11 genera were annotated. The most abundantly detected species were Funneliformis mosseae, an unknown Claroideoglomus sp. and Rhizophagus irregularis. Principal coordinates analyses showed that neither altitude nor plant developmental stages were influencing the AMF community composition (See Figure 4 in supplementary material Annex I). However, species accumulation curves demonstrated that the sampling density was not sufficient to represent the AMF diversity of the roots. Moreover, the Chao index indicated a possible total number of >26 AMF species for the root samples, supporting that a more in detail characterization by deep sequencing was necessary (See Figure 5 in supplementary material Annex I).
New primers were developed and tested for 454GS-FLX+ targeting a ~760 bp region in the LSU rRNA gene, which was previously shown to provide the best resolution for GS-FLX+ sequencing read lengths. A pipeline based on a high throughput phylogenetic annotation approach using a reference dataset and an evolutionary placement algorithm was developed to monitor AMF at the species level. Selected root samples covering the 3 countries, 4 altitudes and 3 plant developmental stages and the rhizosphere soil from Peru were sequenced using 454 GS-FLX+ pyrosequencing. In total, 141 samples were successfully sequenced (3 replicates of 47 samples) resulting in > 600,000 reads with an average length of 700 bp. From the root samples, 41 species from 12 AMF genera were annotated, 15 species (37 %) are unknown or previously not described in sequence databases. The most abundant colonizers were Cetraspora nodosa and two unknown Acaulospora spp. Most plants (85%, 87 plants) were colonized by one of the Acaulospora sp (Fig. 6). Altitude, plant stage and plant variety had no influence on the AMF species community composition. Most of the samples are colonized by a conserved group of AMF (67% of the plants from 12 studied sites were colonized by Cetraspora nodosa, Acaulospora spp., Rhizophagus spp. and Claroideoglomus spp. simultaneously) which are putatively main players in potato AM in the Andean region (See Figure 6 in supplementary material Annex I).

In: Senés-Guerrero C, G Torres-Cortés, S Pfeiffer, M Rojas, A Schüßler. 2013. Potato-associated arbuscular mycorrhizal fungal communities in the Peruvian Andes. Mycorrhiza DOI 10.1007/s00572-013-0549-0.
In: Senés-Guerrero C & Schüßler A. A conserved arbuscular mycorrhizal fungal species community in potato from the Andes. Submitted to New Phytologist.

• Main result 4: Construction of metagenomic libraries

Three potato-rhizosphere-soil-DNA metagenomic libraries were constructed, one library from Ecuador soils with ~2.5 mill clones and average insert DNA size of 10.0 kbp, one from Bolivia with ~2.3 million clones and average insert size of 10.5 kbp and one from Peru with 0.5 mill clones and average insert size of 28 kbp. Before library construction, for each country, rhizosphere soil samples of flowering potato plants from all four distinct altitudes were combined together with samples of juvenile and mature plants from highest altitudes. Pooled soil samples were used to isolate DNA by enzymatic digestion of soil immobilized in the agarose plugs (Ecuador, Bolivia) or by the SDS-CTAB method (Peru). Applying PFGE, DNA larger than 6 kbp (Ecuador, Bolivia) or DNA of 25-60 kbp (Peru) was isolated. DNA isolated from Ecuador and Bolivia was blunt-end cloned into the vector pCC1fos (Epicentre) and Peru DNA into the vector pJC8. Ligated DNA was electroporated (Ecuador, Bolivia) or transduced (Peru) into the E.coli cells. Forty-three and 38 glycerol stock vials were prepared from Ecuador and Bolivian soil metagenomic DNA, respectively, each containing 2x104 - 105 distinct clones. Clones obtained from Peru soil were stored as a single glycerol vial. To determine the quality of constructed libraries, polyketide synthase (PKS) and non-ribosomal peptide synthase (NRPS) gene operons were partially amplified by PCR using the metagenomic fosmid DNA isolated from the pooled glycerol stock vials as template and applying primers specific for actinobacterial KKS and NRPS synthesis genes. Sequence analyses of amplicons and derived gene bank clones revealed that numerous metagenomic clones comprise at least partially PKS and NRPS gene operons and that the majority of obtained sequences show low homology at nucleotide and protein level to the known PKS and NRPS gene operon sequences in the NCBI databank (less than 80%) (See Figure 7 in supplementary material Annex I).

In: Nikolic et al. Manuscript in preparation.

• Main result 5: Assessment of the functional potential of plant-associated bacteria by using a cultivation-based approach

Bacteria from the root environment (rhizosphere and endophytes) of field grown potatoes were isolated on R2A and characterized for their ability to antagonize the pathogens P. infestans and R. solani as well as to produce IAA, ACC deaminase and to solubilize P. Table 1 (See supplementary material Annex I) shows the functional characteristics of isolates endophytes, which showed broad phylogenetic diversity belonging to Proteobacteria, Firmicutes, Actinobacteria and Bacteroidetes. Thereby, 23 of 322 (14 Ecuador, 3 Peru, 6 Bolivia) bacterial endophytes expressed biocontrol activity against R. solani, while 3 bacterial endophyte isolates showed biocontrol activity against P. infestans. A qualitative measurement for indole-3-acetic-acid (IAA) phytohormone production was performed, whereas 43 of 237 endophytes gave a positive result in this assay. Most active endophyte strains are shown in Table 1 (See supplementary material Annex I). A high number of Rhizoctonia antagonists were encountered in Cañar-Cañar in Ecuador (3561 m a.s.l.) as well as in Villaflores in Bolivia (3945 m a.s.l.) whereas Phytophthora antagonists were generally rarely encountered. In total, six strains mostly belonging to the genera Pseudomonas showed at least 3 PGP traits.
A particularly high number of plant growth-promoting rhizobacteria was isolated from Peruvian soils (See Table 2 in supplementary material Annex I). Out of 196 strains 146 isolates showed the ability to produce IAA and 72 showed P solubilization, whereas 76 and 118 isolates showed antagonistic activity against R. solani and P. infestans, respectively. The most active isolates obtained from Bolivia are listed in Table 3 (See supplementary material Annex I). In Ecuador 46 bacterial strains were isolated from the rhizosphere of potato crops during the flowering stage. Out of 46 bacterial strains 18 were able to produce IAA, 13 solubilize P and 12 were antagonists of R. solani. The most active isolates are listed in Table 4 (See supplementary material Annex I).

In: Pfeiffer S et al. Manuscript in preparation.

WP4 – Germplasm collection of Andean microbial diversity

Graphs, Figures, Tables are in supplementary material under heading Annex II.

• Main result 1: Identification of isolated microorganisms from the Andean agro-ecosystems

In total, 435 bacterial isolates were characterized based on a two-step identification approach. The first step was dereplication with either MALDI-TOF MS or rep-PCR which resulted in a grouping of bacterial isolates at the species-to-strain level. In the second step, 308 representative isolates were selected and further identified at the genus level using 16S rRNA gene sequencing. The SINA alignment tool (www.arb-silva.de) was run with 304 high quality 16S rRNA sequences to classify the isolates. Based on this SINA classification 303 isolates belonged to four divisions (Proteobacteria, Firmicutes, Actinobacteria and Bacteroidetes) and 293 isolates could be determined at the genus level, in 32 different genera. Most of the isolates were Pseudomonas (Proteobacteria) or Bacillus (Firmicutes). Table 5 (See supplementary material Annex II) provides an overview of the number of isolates per genus.

Fifty of these isolates that showed antagonistic activity against Phytophthora infestans and/or Rhizoctonia solani (WP5), and which were identified as Pseudomonas sp. or Bacillus sp. were further identified to a finer taxonomic level. For the Pseudomonas spp. this was done using using Multi Locus Sequence Analysis (MLSA) by sequencing the atpA (n=75), glnA (n=75), rpoB (n=80) and rpoD (n=75) genes. To improve identification, a reference database with Pseudomonas type strains was created. A total of 120 well defined Pseudomonas type strains were analyzed with for these four genes. This way isolates were assigned to P. corrugata (6 isolates), P. veronii (1), P. palleroniana (7), P. grimontii (4), P. koreensis or very close to this species (7) and P. viridiflava (3). Furthermore, about 20 novel Pseudomonas species were discovered.

For a selection of 36 Pseudomonas strains a whole-genome sequencing analysis was done in order to obtain a better phylogeny and identification for these isolates and to further the phylogenetic knowledge of the entire Pseudomonas genus. The obtained reads were assembled with CLCbio (Aarhus, Denmark) and annotated (RAST). The genomes were analyzed with JSpecies and an in house script for average nucleotide identity (ANI). ANI values above 95-96% indicate identical species. This comparison allowed the identification of isolates as P. kilonensis/P. palleroniana (6), P. koreensis (2), P. corrugata/P. salmonii (1), P. mediterranea (1), P. viridiflava (1), P. beatica (1), P. fluorescens (1) and P. grimontii (1) and revealed five potentially new species among the isolates.

For the Bacillus sp. finer identification was attempted by sequencing the gyrB housekeeping gene. Despite the testing of several PCR conditions with specific primers, but only three amplicons were obtained. New broad and robust primers need to be developed.

For the six Flavobacterium isolates a further characterization by rep-PCR and MALDI-TOF MS was performed. One is related to F. hibernum while the others represent one or two novel species.

In: Ghyselinck J, Pfeiffer S, Heylen K, Sessitsch A, De Vos P. The effect of primer choice and short read sequences on the outcome of 16S rRNA gene based diversity studies. PLoS One. 2013 Aug 19;8(8):e71360. doi: 10.1371/journal.pone.0071360. eCollection 2013. PubMed PMID: 23977026; PubMed Central PMCID: PMC3747265.
In: Ghyselinck J, Coorevits A, Van Landschoot A, Samyn E, Heylen K, De Vos P. An rpoD gene sequence based evaluation of cultured Pseudomonas diversity on different growth media. Microbiology. 2013 Oct;159(Pt 10):2097-108. doi: 10.1099/mic.0.068031-0. Epub 2013 Aug 6. PubMed PMID: 23920133.
In: Ghyselinck J, Velivelli SL, Heylen K, O'Herlihy E, Franco J, Rojas M, De Vos P, Prestwich BD. Bioprospecting in potato fields in the Central Andean Highlands: screening of rhizobacteria for plant growth-promoting properties. Syst Appl Microbiol. 2013 Mar;36(2):116-27. doi: 10.1016/j.syapm.2012.11.007. Epub 2013 Jan 16. PubMed PMID: 23333025.
In: Ghyselinck J, Van Hoorde K, Hoste B, Heylen K, De Vos P. Evaluation of MALDI-TOF MS as a tool for high-throughput dereplication. J Microbiol Methods. 2011 Sep;86(3):327-36. doi: 10.1016/j.mimet.2011.06.004. Epub 2011 Jun 13. PubMed PMID: 21699925.

In total, 27 AMF were isolated in trap plant culture and classified using morphological approach base on spores characters (size, shape, color…). 15 AMF were only classified at the order level, 2 to the genera level and one fungus was not classified.

For 10 AMF used as starter culture, a morphological identification was realized base on spores size, shape, colour but also wall structure, subtending hyphae and melzer’s reaction), a portion of the rDNA (SSU-ITS-LSU fragment, ~1500 bp.) sequence was also used to identify them to the species level. Some strains were more specifically identified/re-identified.

In: Potten V., Senés Guerrero C., Torres Cortes G., Schussler A., Walker C., Declerck S., Cranenbrouck S. Epitypification and DNA barcoding of Rhizophagus invermaius (=Glomus invermaium) with a re-description from in vivo and in vitro cultures. Submitted to Mycologia.

• Main result 2: Preservation of isolated microorganisms in international culture (BCCM/LMG and BCCM/MUCL) collections and duplication within the respective countries of origin.

All the isolates described above are maintained, for bacteria, in BCCM/LMG and for AMF in BCCM/MUCL. For the bacteria, the strains were maintained under cryopreservation, while for AMF they were all maintained in pot cultures, some in root organ culture and some under cryopreservation.

Most of the cultures were duplicated in the respective countries of origin, i.e. Peru, Ecuador and Bolivia.

• Main result 3: Development of preservation protocols for delicate strains and for complexes of isolates (multi-organism preservation).

o Development of a protocol for the long-term cryopreservation of AMF (See Figure 8 in supplementary material Annex II).

In: Lalaymia I., Cranenbrouck S., Declerck S. (2014). Maintenance and preservation of ectomycorrhizal and arbuscular mycorrhizal fungi. Mycorrhiza 24: 323-337.
In: Lalaymia I., Declerck S., Naveau F., Cranenbrouck S. (2013a). Cryopreservation of arbuscular mycorrhizal fungi from root-organ and plant cultures. Mycorrhiza 24:233-237.
In: Lalaymia I., Declerck S., Cranenbrouck S. (2013b). Cryopreservation of in vitro produced arbuscular mycorrhiza fungi has minor effects on their morphology, physiology and genetic stability. Mycorrhiza 23: 675-682.
In: Lalaymia I., Cranenbrouck S., Draye X., Declerck S. (2012). Preservation at ultra-low temperature of in vitro cultured arbuscular mycorrhizal fungi via entrapment-drying. Fungal Biology 116:1032-1041.

Protocol:
(1) Gelling medium, containing spores and roots of a 5-month old in vitro culture, is extracted from the Petri dishes, poured into 100 ml of sterilized (121 °C for 15 min) deionized water and subsequently blended twice for 30 s at 20,000 rpm in a sterilized (121 C for 15 min) mixer
(2) The mixture is filtered on a sterilized (121 °C for 15min) nylon filter (40 mm).
(3) The supernatant (spores and mycorrhizal/nonmycorrhizal root pieces) is encapsulated in 2 % (w/v) solution of sodium alginate (50±5 propagules per bead).
(4) The encapsulated propagules are incubated overnight in 0.5Mtrehalose and
(5) dried at 27 °C for 48 h (bead water content of approximately 8.1±4.6 %).
(6) Beads are transferred into 2-ml cryovials.
(7) The cryovials are cryopreserved in a freezer at −130 °C following a two-step decrease in temperature: a fast decrease (∼12 °C min−1) from room temperature (+20 °C) to −110 °C followed by a slow decrease in temperature (∼1 °C min−1) from −110 to −130 °C.
(8) For revival, the encapsulated AMF propagules are directly plunged in a water bath at +35 °C.
(9) The beads are dropped into sterilized (121 °C for 15 min) MSR medium, cooled in a water bath to 40 °C, and incubated at 27 °C for germination.
(10) After 4-weeks incubation, beads containing germinated propagules are associated with an excised root under in vitro culture conditions to reinitiate the fungal life cycle.

Lalaymia et al. (2013a) adapted this cryopreservation protocol to in vivo produced propagules: Pot cultures, at least 5 months old, were sampled. Spores were collected by wet sieving and decanting, while roots were collected with forceps and blended in a mixer in 100 ml deionized water for 30 s at 20,000 rpm, and filtered as above. The spores and the supernatant of the blended roots were mixed together and encapsulated in alginate beads, dried, cryopreserved, and thawed as described above. After thawing, the encapsulated propagules were placed directly in contact with roots of plants in pots containing a sterilized (2×15min at 121 °C, with 12-h interval) substrate. The plants were grown for at least 8 weeks in a growth chamber before assessing the fungal viability following cryopreservation.

o Development of a protocol for the long-term cryopreservation of bacteria

1. Prepare liquid medium suited for the bacterial strain (this is the medium which the bacteria have been grown on without the agar) and add 15-30% glycerol (this is strain dependent; some will survive freezing at low glycerol percentages whilst others might need higher concentrations) to the medium.
2. Sterilize the medium at 121°C and 1 bar for 20 minutes. Cool down.
3. Prepare the Nunc® CryoTubes® (Sigma) by inoculating the vials with the medium + glycerol. Use a sterile pipette and work under sterile conditions! Every time you open or close a vial, pass the upper part through a flame.
4. Inoculate the medium in the Nunc vial with the bacterial culture you want to preserve. This culture was prepared earlier as follows:
a. Pick up the colony from the plate with a sterile swab.
b. Make a massive culture: take a new plate of suitable growth medium and swab the whole surface with the colony on the swab.
c. Incubate at right temperature for required time.
Inoculation is performed as follows:
a. After incubation, pick up a large amount of bacteria with a sterile swab.
b. Twist the swab in the Nunc vial, so enough bacteria are transferred to the vial.
5. Let the bacteria grow in the Nunc vial (containing the medium + glycerol (step 1)), preferably 24h or more if necessary.
6. Store the samples preferably at -80°C or, if not possible, at -20°C.

o Development of a protocol for the long-term cryopreservation of complexes of organisms (i.e. AMF and bacteria).

The method was based on the encapsulation technology developed for AMF (see above). In the beads AMF and bacteria were incorporated and tested for survival following cryopreservation. The entrapment process had no detrimental effect on the survival of the AMF and bacteria when entrapped singly. Even if the number of CFUs decreased for bacteria, regrowth was observed in all cases. The combination of AMF and bacteria in beads was possible. However, germination of AMF was hindered in the presence of bacteria probably because of the release of inhibitory substances. There is therefore a necessity to develop another approach for these specific bacteria. This does not preclude that the encapsulation-drying of AMF with bacteria may be effective for the combination of AMF with other groups of beneficial bacteria.

WP5 – Properties of Microbial Inoculants

Graphs, Figures, Tables are in supplementary material under heading Annex III.

• Main result 1: The role of rhizosphere bacteria and AMF in the suppression of soil borne diseases using a metagenomic and co-cultivation approach (i.e. screening of the organisms for antagonism)

UCC- Screening for antagonism: Of the 585 isolates isolate from the fields of Peru and Bolivia, 58 isolates isolated from were effective against R. solani. However, two isolates failed to inhibit growth of P. infestans and, therefore, 56 isolates were effective against both pathogens. Antagonistic activity against R. solani ranged from 24.90% to 53.41% inhibition, while antagonistic activity against P. infestans ranged from 0% to 100%. Where there was no evidence of any growth of P. infestans it was assumed that there was 100% growth inhibition by the bacterial strain. The overall inhibition, for which the size of the inhibition zones on the plates was taken into account, was largest for P. infestans, which could mean that the isolates in vitro were more effective in controlling P. infestans than R. solani. Of the 69 isolates isolated from the fields of Ecuador, seven isolates showed antagonistic activity against R. solani and P. infestans, ranging from 6.02 to 57.4% and 14.06 to 86.75% inhibition, respectively. The commercial rhizobacterial strain B. subtilis FZB24® WG inhibited growth of R. solani and P. infestans by 21.32% and 65.11%, respectively. These results have been published (some papers in preparation) in the papers listed below.

AIT- Screening of metagenomic libraries for plant growth-promoting and biocontrol activities
A. A high throughput assay was developed and established that can be used to screen metagenomic libraries for clones with antagonistic activity against pathogens. The assay enables screening of up to hundred distinct clones per each well of 96 well plates or about 10000 clones per one plate. Shortly, bacterial glycerol stocks of metagenomic libraries are first highly diluted to comprise about 100 distinct clones in 100 µl aliquots and cultured in growth medium overnight in 96-deep-well plates. The glycerol back-up plate is prepared and remaining bacteria are infected with lCE6 phage suspension. The lCE6 phages are carrying T7 RNA polymerase gene and are used to get T7 polymerase expressed in library-host cells (E.coli). The T7 polymerase binds to T7 promoter that is placed at the vector arm in the front of the cloned metagenomic DNA. The transcription driven from T7 promoters is supposed to be very intensive and to lead to the lysis of E.coli cells due to the over-expression of encoded proteins. Proteins and/or their products that cannot be actively secreted by the E.coli might be released into the medium. Produced supernatants are lyophilised to concentrate potentially active compounds and are resolved into a low amount of water to keep concentrations high.. For the wells in which test microorganisms cannot grow, screening procedure is repeated using bacteria from the same positioned wells of back-up glycerol stock plates diluted to one or few metagenomic clones per one well of the screening plate. Metagenomic libraries were screened for antagonistic activities against P. infestans and R. solani. In a high throughput screening assay 104 clones from each of 81 stored Ecuador and Bolivia library-glycerol-stock fractions (each comprising about 2x104 to 105 clones) were tested for antagonistic activity against R. solani and P. infestans. Out of screened 8 x 105 metagenomic clones no clone showed visible growth inhibition activity against tested microorganisms. Furthermore, dual culture screenings were performed. Eight x 104 clones were screened in total on 162 agar plates resulting with no clone showing antagonism to R. solani or P. infestans. Five x 104 clones from Peru metagenomic library were also screened for the antagonism against R. solani but with no success. These results might be due to the fact that many bacterially produced bioactive molecules are synthesised by complex multi-domain enzymatic systems as polyketide synthases (PKS) or non-ribosomal protein synthases (NRPS) that might span up to 120 kbp, the size that lays far beyond average DNA insert size of tested metagenomic libraries (~10 and 28 kbp). It is also possible that library-host E. coli cells might not deal efficiently with phylogenetically distant DNAs and encoded proteins

B. Metagenomic libraries (Ecuador 2.5 million, Bolivia 2.3 million and Peru 0.5 million clones) were screened by functional screening assays for several enzymatic activities, including chitinase, cellulase, protease, phosphate solubilisation and CHN production activity. Screening was performed on modified M9 minimal medium agar plates supplemented with carboxymethylcellulose (CMC) or chitin as sole carbon sources, bovine serum albumin (BSA) as sole nitrogen source and tri-calcium phosphate (TCP) as a sole phosphor source. Two metagenomic clones were isolated on plates supplemented with CMC, one on plates with chitin (not shown) and six on each of plates that were supplemented with BSA or TCP (See Figure 9 in supplementary material Annex III). Genes responsible for the obtained enzymatic activities are not yet identified or sequenced but work will continue in the future.

Screening for HCN producing clones did not yield positive clones, instead, three interesting clones were isolated that produced a volatile compound. These clones gave a chromogenic reaction with copper(II)-ethylacetoacetate on the filter paper (See Figure 10 in supplementary material Annex III), which is placed above the colony and inside the lid of a sealed Petri dish. They have been fully sequenced and analysed for the gene content. All three clones have DNA inserts of about 38-39 kbp in size, two show homology to the Aeromonas sp. and one to Comamonas sp. DNA. Overall homology between two Aeromonas-clones is very high except few gene insertions found in one of them but is very low in comparison to the Comamonas-related insert DNA. Except genes for the transcriptional regulator that belongs to LysR family that is found in all three clones, no other genes with significant homologies were found.

C. Metagenomic libraries were screened for the presence of polyketide synthases and non-ribosomal peptide synthase encoding genes. The presence and diversity of polyketide synthase-I (PKS-I) and non-ribosomal peptide synthase (NRPS) genes in the constructed metagenomic libraries were analysed by PCR using gene specific primers and fosmid DNA prepared from several pooled library-stock vials that cover whole libraries. The obtained amplicons were cloned into the appropriate vector and derived gene-clones sequenced. The 95 PKS-I related gene-clone sequences (partial kethosynthase-methyl-malonyl-CoA transferase domain) and 71 NRPS related sequences (partial adenylation domain) were blasted against NCBI database. The 80 PKS-I sequences showed homology to PKS genes known or predicted to be involved in certain polyketide biosynthesis pathways and 16 sequences were not related to any known biosynthesis pathways (Table 1). The majority of sequences was related to Streptomyces sp. but overall they affiliated with seven additional actinobacterial species.
The obtained NRPS sequences showed homology to 24 actinobacterial species and to only few NCBI matches that are related to known non-ribosomal peptides. However, obtained sequences of partial NRPS adenylation domains showed high phylogenetic diversity and were predicted using NRPSpredictor2 to have specificity for versatile monomer substrates. Among others, four domains were predicted with a high score to use cysteine as a substrate, rarely found in NRP-es. Both, NRPS and PKS-I gene operon sequences showed rather low homology to their best matches in the NCBI databank (less than 80%), indicating Andean soil as a valuable source of microbes that potentially produce novel polyketides and non-ribosomal peptides.
The Peru metagenomic library was also analysed for the NRPS gene operon content but the obtained PCR-amplicon was sequenced directly by the 454-sequencing. The 452 partial adenylation domain sequences (NRPS) were obtained using forward and 188 using reverse sequencing primers. Neighbour-joining revealed high diversity of obtained sequences that grouped in approximately 15 phylogenetically separated groups.

In: Ghyselinck J., Velivelli S., Heylen K., O’Herlihy E., Franco J., Rojas M., De Vos P., Doyle Prestwich B. (2013). Bioprospecting in potato fields in the Central Andean Highlands: Screening of rhizobacteria for plant growth-promoting properties. Systematic and Applied Microbiology 36 (2):116-127.
In: Velivelli, S., O'Herlihy, E., Janczura, B., Doyle Prestwich, B., Ghyselinck, J. and De Vos, P. (2012). Efficacy of Rhizobacteria on plant growth promotion and disease suppression in vitro. Acta Hort. (ISHS) 961:525-532.
In: Loján P., Demortier M., Velivelli S., Doyle Prestwich B., Dupré de Boulois, H., Suarez J.P. Declerck S. (2014). Co-encapsulation of Rhizophagus irregularis MUCL41833 and bacteria and its impact on the arbuscular mycorrhizal fungal life cycle under in-vitro conditions. In preparation.
In: Nikolic et al. (2014). Screening of metagenomic libraries for plant growth promoting and biocontrol activities. In preparation.

• Main result 2: To elucidate the role of rhizosphere microorganisms and communities in plant growth promotion

UCC - Plant growth promotion and disease suppression: Strains which showed antagonistic activity were also tested in vitro for IAA production, phosphate solubilisation, ACC deaminase activity, ammonia, HCN, siderophore and lytic enzyme (chitinase, protease, glucanase) production. Fifty four isolates produced siderophores, six isolates were positive for HCN and twenty one isolates were positive for ammonia production. Twelve isolates produced IAA, thirty isolates expressed the enzyme ACC deaminase , forty eight isolates had the capacity to solubilize phosphate. The commercial rhizobacterial strain B. subtilis FZB24® WG was positive for NH3-production, and produced IAA and ACC and had the capacity to solubilize phosphate. One isolate belonging to the genus Paenibacillus and isolated from ﬁeld 2 in Peru, was able to produce three lytic enzymes (i.e. proteases, cellulases and glucanases). Almost all bacterial isolates (86.2%) produced proteases, while only a minority produced chitinase (10.3%) and glucanase (5.2%). One isolate (R-42302) produced cellulase. Eight bacterial isolates belonging to the genera Pseudomonas (5), Pedobacter (1) and Enterobacter (2) did not produce any lytic enzymes. All 58 antagonistic isolates were tested in in vitro experiments on plantlets in order to evaluate their plant growth-promoting abilities and disease suppression. During growth room experiments, 23 isolates were associated with plant growth-promotion and/or disease suppression. Ten isolates had a statistically significant impact on test parameters (i.e plant growth promotion and disease suppression) compared to the uninoculated control. Three isolates significantly promoted plant growth in healthy plantlets compared to the commercial strain (See Figure 11 in supplementary material Annex III), and seven isolates out-performed the commercial strain in in vitro R. solani diseased plantlets (See Figure 12 in supplementary material Annex III). Of seven isolates tested for the production of IAA from Ecuador, two isolates, R47065 and R49538, were found to be positive for IAA production. ACC deaminase activity was found in all isolates and five isolates were able to solubilize phosphate on plate assays. The seven strains tested in the in vitro microhydroponic system, showed growth promotion and disease suppression (R. solani) equal to that of uninoculated control and B. subtilis FZB24® WG inoculated plants. These results have been published (some papers in preparation) in the papers listed below.

At CIP, 74 selected bacterial strains previously isolated from Andean agro-ecosystems in Peru and which had been tested in vitro for their plant growth promotion capacities at CIP (antagonism to R. solani and P. infestans, production of IAA, solubilization of P), were tested in pot trials to evaluate their ability to increase plant growth and tuber yield on potato cultivar Unica. Results of the pot trials showed that among the 74 bacterial strains tested, 39 strains resulted in tuber yield increase (up to 66 %) compared to the non-inoculated control.

At PROINPA (Bolivia) seventeen bacterial isolates from VALORAM and five from early COMMINANDES were received and multiplied at PROINPA for their evaluation under greenhouse conditions as plant growth promoters and/or potato disease suppresor. AMF isolates from the MUCL collection were also received and multiplied for their evaluation.

After lab and greenhouse evaluation of 381 local bacteria isolates 98 strains were selected and evaluated in pots under greenhouse conditions for their suppressive effect on R. solani, Nacobbus aberrans and Globodera sp. on potato plants as well as of potato plant growth parameters. Out of the 98 evaluated isolates, 5 were selected and they are being evaluated under farmer field conditions in two localities for their potato growth parameters, yield, R. solani, Nacobbus aberrans and Globodera sp.

Besides, 117 endophytic bacterial strains isolated from roots of native potato cv Waycha, 29 isolayes from cvs. (Solanum tuberosum ssp andigena and S. stenotomum) and other 235 bacterial isolates maintained in our collection were tested in lab for P solubilization activity, IAA, catalasa activity, Gram reaction (stain test), and antagonistic function to Fusarium sp.and Rhizoctonia. solani.

From 98 selected isolates after lab and greenhouse evaluations 5 strains are being evaluated under field condition in two localities

In: Ghyselinck J., Velivelli S., Heylen K., O’Herlihy E., Franco J., Rojas M., De Vos P., Doyle Prestwich B. (2013). Bioprospecting in potato fields in the Central Andean Highlands: Screening of rhizobacteria for plant growth-promoting properties. Systematic and Applied Microbiology 36 (2):116-127.
In: Velivelli, S., O'Herlihy, E., Janczura, B., Doyle Prestwich, B., Ghyselinck, J. and De Vos, P. (2012). Efficacy of Rhizobacteria on plant growth promotion and disease suppression in vitro. Acta Hort. (ISHS) 961:525-532.
In: Velivelli S., De Vos P., Kromann P., Declerck S., Doyle Prestwich B. (2014). Biological Control Agents: From Field to Market, Problems and Challenges. Accepted in Trends in Biotechnology.
In: Loján P., Demortier M., Velivelli S., Doyle Prestwich B., Dupré de Boulois H., Suarez J.P. Declerck S. (2014) Co-encapsulation of Rhizophagus irregularis MUCL41833 and bacteria and its impact on the arbuscular mycorrhizal fungal life cycle under in-vitro conditions. Paper in preparation.

Volatilome of the organisms using GC/MS
The volatile organic compounds emitted by Andean bacteria were diverse in chemical nature and inculde amines, alcohols, nitrogen-containing alkylpyrazines, ketones, hydrocarbons, acids, esters, sulphur (s)-containing compounds, aldehydes, amides and phenols. Several of these volatile organic compounds, such as 2,5-dimethylpyrazine and 2-hexen-1-ol have been detected in the headspace cultures of all bacterial strains. Analysis of volatile profiles also revealed that phenol, 2,4-bis (1,1-dimethylethyl) was emitted by all bacterial isolates except R42086, R41757, R41761 and R41798. Of 27 bacterial strains, 16 isolates produced the volatile organic compound, 1-undecene. 2,3-butanediol was detected in 6 bacterial strains; R42086, R41815, R41849, R41850, R42141 and R47131. Interestingly, the compound dimethylamine was produced by all these isolates except, R42141 and R47131. Moreover this compound was also detected in R47065. Furthermore, of all bacterial isolates tested for the production of VOCs, the compound fluoroacetamide was emitted by 8 strains, followed by 2-tridecanone (7), acetic acid (5), methyl tetradecanoate, 1-decanol, 1-dodecanol (4), 1,4-undecadiene, dodecane, dodecanoic acid, methyl ester (3), 2-undecanone, methylhydrazine oxalate, dimethyl disulphide, decanoic acid, methyl ester (2), Cyclododecene, 1-tetradecene, E-1,9-hexadecadiene, 1-hexadecene, cyclopropyl carbinol, nonanoic acid, hexadecane, pentanal, acetic acid octyl ester, eicosane, oxoacetic acid, E-7-tetradecenol, 1-tetradecanol and Z-7-tetradecenol (1).

In: Velivelli S., Kromann P., Lojan P., Rojas M., Franco J., Suarez J.P. Doyle Prestwich B. (2014). Identification of mVOCs from Andean rhizobacteria and field evaluation of bacterial and mycorrhizal inoculants on growth of potato in its center of origin. Submitted to Microbial Ecology.
Velivelli S., Doyle Prestwich B. The discriminative power of microbial volatile organic compounds (mVOCs). Invited review, in Preparation for Critical Reviews in Microbiology.

• Main result 3: To elucidate the role of rhizosphere microorganisms in the priming of plant biotic defences and exploitation of their characteristics in the control of haulum (i.e stem and foliar) and storage diseases of potato.

Transcriptomics approach:

This study investigated if inoculation of potato plantlets with AMF (R. irregularis MUCL 41833) and a PGPR (Pseudomonas sp. R41805) either alone or in combination, could elicit host defence response genes in the presence or absence of R. solani. Glutathione-S-transferase 1 (GST1), 1,3-β-glucanase (PR2), and class II chitinase (PR3) for the SA pathway, lipoxygenase (Lox) for the JA pathway, phenylalanine ammonia (PAL) and basic PR1 (PR1b) for the combined JA- and ET-mediated pathway, and the ethylene response factor 3 (ERF3) for the ET pathway were chosen as pathway reporter genes. ERF3 was upregulated in the mycorrhizal potato plantlets cultivated with bacteria in absence of pathogen (p<0.05) and the mycorrhizal potato plantlets cultivated with bacteria and challenged with R. solani (p<0.05) with no significant differences in expression seen in the other treatments. This was confirmed by three-way ANOVA; AMF (p<0.001) and bacteria (p<0.01) as well as their interaction (AMF × bacteria; p<0.05) were significantly associated with ERF3 upregulation. These results suggest that the dual inoculation of PGPR and AMF can bioprime the potato plant through activation of the plant systemic defense system via ERF3. These data contribute to a better understanding of how plants modulate defence responses, particularly when co-inoculated with PGPR and AMF, thus providing potential new avenues for effective management of potato-associated diseases

In : Velivelli S., Lojan P., Cranenbrouck S., Dupré de Boulois H., Suarez J.P. Declerck S., Franco J., Doyle Prestwich B. (2014). The induction of Ethylene response factor 3 (ERF3) in potato as a result of co-inoculation with Pseudomonas sp. R41805 and Rhizophagus irregularis MUCL 41833 – a biopriming response? Submitted to Physiology and Molecular Plant Pathology.

Proteomics approach:

This study investigated the changes in expression profiles of proteins in potato plantlets inoculated with Pseudomonas sp. R41805 (B) and also in plantlets inoculated with Pseudomonas sp. R41805 and challenged with R. solani (B+R). Total proteins were extracted from the leaves, separated by a 2D-PAGE system, stained with silver, analysed by using Progenesis Samespot 2.0 software, and tentatively identified using MALDI MS/MS. Uninoculated plantlets (C) challenged with R. solani (C+R) were also set up. A total of five proteins were found to be differentially expressed at least in one treatment. The identified proteins include RuBisCo small subunit 2B and C, Oxygen-evolving enhancer protein 2 (OEE2), Plastocyanin B, Chaperonin 20, 33kDa precursor protein of oxygen-evolving complex and Chloroplast manganese stabilizing protein. The expression of Plastocyanin B was highly expressed (p<0.05) in Potato plantlet challenged with R. solani (C+R) compared to the uninoculated plantlets. The protein, RuBisCo small subunit C was significantly higher in Potato plantlet challenged with R. solani (C+R) and also potato plantlets inoculated with bacteria and R. solani (B+R). No significant difference was observed in the potato plant treated with bacteria (B). The differential expression of RuBisCo small subunit 2B was similar in all treatments compared to the control. The expression of Chloroplast manganese stabilizing protein was significantly higher in the plants treated with bacteria (B+R). The expression was statistically similar in the remaining treatments. The 33kDa precursor protein of oxygen-evolving complex protein, Oxygen-evolving enhancer protein 2 (OEE2), and Chaperonin 20 was significantly upregulated in the plantlets inoculated with bacteria (B) and also in the plantlets treated with bacteria and challenged with R. solani (B+R) compared to the uninoculated control plantlet. These results will be published in the paper listed below.

In: Velivelli S., Doyle Prestwich B. (2014). Identification of differentially expressed proteins in potato plantlets inoculated with bacteria and challenged with R. solani. In Preparation.

Metabolomic analysis of glycoalkaloids:

The total glycoalkaloid (TGA) was determined in “Unica” produced potato tubers grown in greenhouse following inoculation with Pseudomonas spp. R41805 and commercial bacterial strain B. subtilis FZB24. Potato tubers inoculated with strain R41805 had TGA value of 10.3 mg/100g FW, whereas B. subtilis FZB24 and the uninoculated plants had a TGA content of 11.9 and 10 mg/TGA 100g FW, respectively. The total glycoalkaloid content in potato “Unica” tubers was lower than 20 mg 100g/FW (acceptable safety limit). At UTPL – Ecuador (Partner 8) – total glycoalkaloid content (α-chaconine and α-solanine) from tubers collected from field trials was determined in tubers from the following treatments: Bacillus simplex isolate (E4-21) + Manure, Bacillus positive (R47065) + Manure, Non inoculated + Manure, Bacillus simplex isolate (E4-21) + Manure + NPK, Bacillus positive (R47065) + Manure + NPK and Non inoculated + Manure + NPK. The total glycoalkaloid content in potato tubers of all treatments was lower than 20 mg/100g.

In: Velivelli S, Kromann P., Lojan P., Suarez J.P. Doyle Prestwich B. (2014). The role of –omics (volatilomics, metabolomics and proteomics) in an understanding of potato health and growth. In preparation.

WP6 – Sustainable crop production and agro-ecosystems

Graphs, Figures, Tables are in supplementary material under heading Annex IV.

Certain activities conducted in WP6 were linked with WP7 as the evaluation of selected microbial inoculants in several field trials. Efforts were directed to assess the effect of management practices, on the native microbial community as well as the development of recommendations for sustainable potato cropping.

Objectives
- To evaluate the potential of Andean microorganisms on the productivity and plant health of high and low-input potato-based cropping systems;
- To evaluate the effect of microbial biodiversity on soil fertility and crop performance;
- To evaluate eco-efficient technologies and products for their ability to improve
crop performance.

• Main result 1: Evaluation of the potential of Andean microorganisms on the productivity and plant health of high and low-input potato-based cropping systems.

1. Evaluation of beneficial microorganisms in pot and researcher managed field trials.

PROINPA: Previous to field experimentations, an evaluation of AMF and bacteria formulations was performed under greenhouse conditions with the potato cv. W’aycha (Solanum tuberosum subsp andigena).

In pots containing one kilogram of substrate, the following treatments were applied (See Table 6 in supplementary material Annex IV):
- Commercial non-native AMF, Rhizophagus irregularis from SYMPLANTA, Germany: 3.2 g inoculum/pot
- Commercial non-native AMF, R. irregularis from ASP-A liquid, Canada: 3 ml inoculum/pot
- Commercial non-native AMF R. irregularis from ASP-A solid Canada: 3.2 g inoculum/pot
- Commercial non-native AMF (Funneliformis mosseae) from BioTop, Cochabamba, Bolivia: 100 g inoculum/pot
- Commercial native AMF from BioTop, Cochabamba, Bolivia: 100 g inoculum/pot
- AMF isolate R. clarus 46258 from GINCO: (100 g inoculum/pot)
- Bacillus amyloquefaciens from BioTop: (3 g inoculum/pot)
- B. subtilis from BioTop: (3 g inoculum/pot
- A control treatment which received no inoculum.

In Table 7 (See supplementary material Annex IV), the results of this experiment are presented. Statistical differences were observed only for root volume. Higher tuber weight (42.6 g), number of tubers (69) and number of stems were observed in the treatment with R. irregularis ASP-A Canada (powder 3.2 g inoculum /2 l pot) although not significant. Less galls (20%) caused by the nematode Nacobbus aberrans were also noted in the treatment with with B. amyloquefaciens (3 g inoculum / 2 l pot).

Conclusions: No significant differences were obtained as a clear effect of AMF and bacteria formulations under greenhouse conditions with the potato cv. W’aycha (Solanum tuberosum subsp andigena) in 1kg pot, although nematode attack was lower with the application of B. amyloquefaciens from BioTop. Similarly no clear effect of treatment on the yield was evident although the highest yield was obtained with R. irregularis ASP-A (Solid formulation) and R. clarus 4625 from GINCO.

CIP: Microorganisms selected in in vitro tests for their plant growth promoting capacities were evaluated for these traits in pot and field trials. Results were used to select suitable microorganisms for participatory evaluations with farmers at on-farm sites.

Bacterial strains from CIP collection (See Table 6 in supplementary material Annex IV) were tested under greenhouse conditions in aeroponic and seedbed systems and in field experiments.

In the aeroponic system, the treatments with bacteria applications were statistically significantly different from the control without bacteria application (p< 0.05). Regarding the seedbed system, results were quite similar.

The results from both production systems showed that inoculation with some bacterial strains increases potato tuber seed production, although the effect differed among varieties.

UTPL:

Experiment I: From March to September 2012, one experiment was carried out in two localities in Ecuador: (1) Estación Zamora Huayco - Loja (2160 m asl., 04°0’1.59’’S, 79°10’48.46’’W) and Estación Experimental Santa Catalina - Instituto Nacional de Investigaciones Agropecuarias - Quito (3058 m asl., 0°22’S, 78°33’W)

The objective was to evaluate the effect of different AMF commercial formulations (liquid, powder or granular) - based on the AMF Glomus irregulare DAOM 197198 - on plant performance and tuber production of potato (Solanum tuberosum x andigena var. I-Fripapa 95) a common potato cultivar in Ecuador. Two concentrations of each mycorrhizal product were applied underneath the tubers at sowing time (See Table 8 in supplementary material Annex IV).

The experiment was established in a completely randomized design with five replicates and 10 treatments (See Table 8 in supplementary material Annex IV). Each replicate consisted in one furrow 4.2m length containing 14 potato tubers spaced at 30cm. The distance among furrows was 1m. Guard furrows were left among treatments but neither AMF inoculation nor chemical fertilization was applied to the tubers.
A soil analysis was made at the beginning of the experiment to determine the chemical properties of the soil (See Table 9 in supplementary material Annex IV).

Organic amendment with chicken manure 6 t/ha-1 was applied in all the furrows regardless the treatment. For the fertilized treatment a recommended dosage of fertilizer was 50-120-100kg NPK.ha-1 applied.

Due to the high heterogeneity in the results between sites the analyses were carried out separately.

The results of plant emergence did not show any significant difference among treatments for Loja (p=0,052) or for Quito 2 (p=0,058). But the height of plants measured at 60DAS (days after sowing) showed a statistical significant difference among treatments for both sites (p<0,0001) (See Table 10 in supplementary material Annex IV).

A microscopic analysis to check the intraradical colonization of roots by AM fungi showed some variations in the mycorrhizal colonization of potato roots in Loja (p=0,000) and Quito (0,010). A tendency could be observed in most of the cases. For the lower amount of spores/tuber inoculated a lower intensity of mycorrhizal colonization corresponded (See Table 11 in supplementary material Annex IV). It has to be noted that the non-inoculated treatment also displayed mycorrhizal colonization because of the presence of native AM fungi already present in the plots.
Regarding the yield (kg.ha-1) there were not statistical differences in Loja (p=0,491) or Quito (p=0,126). These results could be derived from the high heterogeneity of soil in the plots that caused high variations within the same treatments. However a positive tendency was observed in favor of the fertilized treatment for both sites (See Figure 13 in supplementary material Annex IV).

The results demonstrate that there was no statistically important effect of DAOM 197198 on yield for both locations. This result could be due to the failure of establishment of symbiosis or survival of the inoculated AM fungi (MUCL43194) under the field conditions or the fungi is not effective to promote yield under the experimental climatic and soil conditions. A molecular analysis of the roots is in progress in LMU to test whether MUCL 43194 was present or not in the inoculated treatments.

A molecular analysis of the roots to corroborate the presence of DAOM 197198 in the plots was carried out in LMU-Germany (partner 4). The results demonstrated a fail of R. irregularis to establish symbiosis with the potato plants under the local conditions during the time frame of the experiment (5 months), this could be caused by several factors, among them the harsh local conditions and high AMF diversity and found in both experimental plots which could compete with the introduced AMF strain. The results will be part of a conjoint publication between UTPL, LMU and UCL.

Experiment II

From March to September 2012, five PGPR bacteria isolated from potato cropping systems (in Bolivia and Ecuador) in the frame of the project COMMINANDES (See Table 12 in supplementary material Annex IV) were provided by LMU-Ghent (Belgium). This group of bacteria has previously displayed potentially positive characteristics to improve plant growth and antagonism against soil borne pathogens like Rhizoctonia solani and P. infestans.

The objective of the experiment was to test if the inoculated bacteria were able to improve potato performance and tuber yield under the experimental conditions. Four control treatments: T6 (B. amyloliquefaciens) obtained through the PROINPA collection; T7 (commercial B. subtilis strain FZB24), T8 (addition of growth media of bacteria, without bacteria) and T9 (conventional fertilization) were also tested.

The results in Loja regarding displayed statistical significant difference for plant height at 60DAS, number of tubers and yield (t.ha-1) (See Table 13 in supplementary material Annex IV).

The bacterial inoculation was made by diluting 80 ml of bacterial inoculant (108 CFU.ml-1) in 2 L of water. The diluted inoculant was evenly applied to the tubers placed in the furrows (0.693l/10 tubers in a row) and immediately covered with soil. Tubers were only inoculated at planting.

Clear differences can be seen among treatments in Loja being the chemical-fertilized treatment (T9) the one which displayed the best results. The results with the lowest yields were the control (T7) with the commercial FZB24 Bacillus subtilis and the non-inoculated control (T8).

No evident differences among treatments were found in Quito for any evaluated parameter.

The results suggest that the effectiveness of bacterial inoculants could rely on the edapho-climatic characteristics of the site. On the other hand, bacterial strains, which displayed the best results, were LMG24415 (B. amyloliquefaciens) and LMG24418 (B. subtilis) in Loja. These bacteria could be proposed to farmers to decrease actual fertilization practices carried out by local farmers.

PROINPA: Five COMMINANDES selected isolates sent from European partners and five locally selected isolates from Bolivia were used to develop a pilot-scale production system under as powder and/or liquid. These inocula were then evaluated under greenhouse and farmer field conditions at four localities.

2. Pilot-scale production of AMF and PGPR inoculums

CIP: In 2011- 2012, ten selected bacterial strains (Bacillus subtilis, Bacillus amyloliquefaciens, Bacillus pumilus, Pseudomonas putida, Achromobacter spp.) belonging to CIP´s collection in Peru were locally formulated at pilot scale liquid and were applied at planting and hilling on potato cultivars Amarilla del Centro, Yungay and Huamantanga under farmer field conditions in four different fields located in the farmer communities San Jose de Aymara, Huancani, Yacta and Cassapata, Perú. Table 14 and 15 (See supplementary material Annex IV) and c show data from the trials in San Jose de Aymara and Yacta, including significant yield effects in bold of P1-20/08
(Pseudomonas putida), B1-22/06 (Bacillus subtilis) and A2-18/08 (Bacillus pumilus).

In 2012-2013, four VALORAM Pseudomonas palleroniana strains (R-42357, R-43582, R-43631, R-43628) and one COMMINANDES bacterial strain (B. subtilis LMG 24422) provided by Partner 5 were locally formulated at pilot scale liquid to be applied in two field trials located in the Huancani and San Jose de Aymara communities, Peru.

Three AMF strains (MUCL43194, Rhizophagus irregularis; MUCL 46238, R. clarus; MUCL 43204 R. irregularis) provided by partner 1 were multiplied in vivo for establishing protocols for pilot production,

CIP was looking for private companies to initiate pilot scale production for field trialling, but the efforts failed due to lack of companies with infrastructure adapted to pilot scale.

Selected bacterial strains belonging to CIP´s collection and VALORAM were formulated in alginate microbeads for pilot-scale production (See Table 16 in supplementary material Annex IV).

PROINPA: Selected bacteria isolates were evaluated in farmer participatory field trials in farmers’ fields in the growing season 2012-2013 in Bolivia (See data in Main Result 3). Isolates were formulated by Bio-Top, a key stakeholder with a formulation plant and technical personnel. This is a private company for the formulation of biological products in Cochabamba and it is a commercial branch of PROINPA Foundation.

UTPL: In Ecuador a total of 7 bacterial strains (4 obtained in the frame of VALORAM and 3 from the former EC COMMINANDES project) that displayed the best results under in vitro tests for their plant growth promoting capacities (antagonism against P. infestans and R. solani, P solubilisation, IAA production and ACC deaminase) were co-encapsulated at a concentration of 3x105CFU/bead in alginate (2%) together with individual spores of R. irregularis (MUCL41833) to determine their effect on spore germination of R. irregularis (MUCL41833) (See Table 17 in supplementary material Annex IV) .

Results:

The results showed a positive effect of the bacterial strain R47065 (Paenibacilus sp) in the germination of R. irregularis (MUCL41833) spores, when compared with the control treatment (without bacterial addition). The remaining bacterial strains had inhibitory effects (See Figure 14 in supplementary material Annex IV).

A second experiment was addressed in-vitro using only the bacterial strain, which displayed the best results in the former experiment (R47065, Paenibacillus sp.) together with propagules of R. irregularis (MUCL41833). Briefly, five alginate beads (containing each 3x105CFU and ±25 R. irregularis propagules) were placed in Petri plates (90mm, diameter) and refilled with warm (40C) MSR medium lacking sugar and vitamins and incubated during 2 weeks. After that, autotrophic in-vitro systems with potato plantlets from three different varieties (Bintje, Unica and I-Fripapa) were built following the protocol described by Voets et al., (2005) with slight modifications (See Table 18 in supplementary material Annex IV). The systems were be kept for 7 additional weeks in a growth chamber at a temperature of 21C day - 18C night and illuminated for 16 h d-1 under a photosynthetic photon flux of 225 Tmol m2 s-1.

The results showed a positive effect of Paenibacillus sp (R47065) on R. irregularis (MUCL41833) ability to establish symbiosis with all the three potato varieties evaluated in-vitro. The number of spores, hyphal growth and percentage of colonization of roots of R. irrregularis was increased by the presence of the Paenibacillus sp (R47065) in the systems, although in not all the cases the differences were statistically significant an increase is easily recognized (See Table 19, 20, 21 in supplementary material Annex IV).

• Main result 2: Effect of microbial biodiversity on soil fertility and crop performance.

Site selection for establishment of participatory field trials and identification of interested farming communities and organizations

Below are described localities where participatory activities were conducted by partners 6, 7 and 8 in Perú, Bolivia and Ecuador with farming communities and organizations in relation to the use and effects of beneficial microorganisms).

PROINPA: Four localities were selected in Bolivia (Colomi, Tiraque, Morochata and Anzaldo). Participatory trials were conducted with farming communities and organizations in these four localities with potato varieties Desiree (S. tuberosum ssp. tuberosa) a very common variety to most of the Andean countries and Waycha (S. tuberosum ssp. andigena) a variety of local importance in Bolivia.

CIP: Identification of stakeholder such as NGOs (DIACONIA, YANAPAI) and NARS (INIA-Peru) were done and reports from participatory evaluation trials (flowering and late period) with farmers was prepared. In 2011-2012, two fields were selected for participatory field trials (Yacta and Cassapata). Participatory evaluation trials were established with the help of the NGO “YANAPAI Group” involving 18 farmers from Yacta and 43 farmers from Cassapata. The farmers evaluated the trials at flowering and harvest in order to select the most accepted technology on the use of beneficial microorganisms by farmers in each community.
In 2012-2013, participatory evaluation field trials were conducted with farming communities from Huancani.

UTPL: Several fields with interested communities of famers were identified in Southern Ecuador in the provinces of Cañar, Azuay and Loja. The sites were selected according to their characteristics like climate, elevation, soil, etc. Capacitation in the use and effects of the beneficial microorganism was offered to the farmers in each preselected site. However the field experiments were carried out only in experimental plots in Santa Catalina (Quito) and Estación Zamora Huayco (Loja).

• Main result 3: Eco-efficient technologies and products to improve crop performance

1. Evaluation of novel plant-beneficial microbial-based products and evaluation of crop management methods on the effectiveness of the beneficial soil microorganisms

Two tasks were grouped together as the evaluation of the novel plant-beneficial microbial-based products was often carried out in experiments intended to evaluate crop management methods on the effectiveness of these beneficial microbial-based products.

Based on the results of WP5 the properties of selected AMF and PGPR isolates on the productivity (yield) and quality (health) of potato cultivars / landraces, major potato soil-borne pathogens (nematodes and fungi) and abiotic constraints (P-level, soil acidity, drought) were participatory evaluated under field conditions.

The most promising products were evaluated in demonstration trials in collaboration with NARS and other interested parties. Other field trials were also conducted to establish technical and agronomic procedures for application and best efficacy of microorganisms, concerning concentrations, timing and quantity applied, support medium, interactions with organic and inorganic fertilizers, etc.

CIP: Four field experiments were performed in Peru from December 2012 to July 2013. Two experiments were replicated at two sites; one at a site in a farmer’s field in Huancani in the Department of Junin and one at Aymara in the Department of Huancavelica. In Huancani, the experiments were done with participation of farmers. The results are presented below.

Participatory Farmer Field Experiment I and II

During 2012-2013, five selected bacterial strains belonging to the VALORAM project, a commercial bacterial strain and two commercial AMF (See Table 22 in supplementary material Annex IV) were locally formulated at pilot scale and were applied at planting to potato cultivar Yungay at two different field locations in Huancani at 3882 masl (Experiment I) and San Jose de Aymara at 3966 masl (Experiment II). The impact of inoculation of the microorganisms in combination was tested under two fertilization rates (00-00-00 and 75-75-50 NPK kg ha-1). Growth and yield of potato was evaluated.

At sowing, bacterial inoculations were prepared by suspending 125 ml of bacterial suspensions at the concentration of 1010 CFU ml-1 in 2 L of river water. Bacterial applications were performed using dipping method in which sprouted tubers were dipped in the mixture for about 20 min prior to planting. Then, tubers were placed in the furrows; additionally a similar mixture was applied directly on the tubers previously planted. For AMF inoculation, the commercial products Myke® Pro PS3 and ©SYMPLANTA were inoculated directly to tuber seeds at a concentration of 200 propagules/tuber and subsequently covered with soil. All experimental plots received organic fertilizer (5 t/ha sheep manure) before application of bacterial and AMF inoculations. For selected treatments, mineral fertilizers 75-75-50 NPK kg ha-1 were applied (50% rate of dosage recommended based on soil analysis). For the non-inoculated control treatments, the seeds were dipped into river water with sterilized TSB. Two control treatments were included: 1) no chemical fertilization, 2) low chemical fertilization (75-75-50 NPK kg ha-1). Each plot had five rows with twelve potato plants (60 tubers /plot). The harvest area used for data analysis was 21 m2, comprising the 5 rows and 12 plants per row. The potato crop was protected against diseases by applying pesticides according to local practice and at low label recommendations and to foliage only. Agricultural practices are shown in Table 23 (See supplementary material Annex IV). Both experiments consisted of factorial designs 9x2 with nine inoculant types and two fertilizer rates as the factors. The experiment was established in a completely randomized design with a total of 18 treatments and five replications. Each replicate consisted of 60 planted tubers per plot. The results were analyzed by an analysis of variance (ANOVA) and means were separated according Tukey’s test at p<0,05 level of significance to analyze statistical differences and to discriminate between means. When no significant value was detected, a non-parametric Kruskal-Wallis test at p<0.05 level of significance was used.

Results: For experiment I Huancaní 2012/2013, results indicated that treatment T3 (5 t manure ha-1 + inoculation with R 43631) resulted in a significantly higher tuber yield (tons ha-1) and number of tubers than the non-inoculated treatment at the same fertilizer level (See Table 24 supplementary material Annex IV) with increases of 26% and 28 % respectively. The lowest yields resulted from inoculation with the commercial AM fungi inoculant (Myke® Pro PS3).

Results of experiment II Aymara 2012/2013 showed no statistical differences between treatments (See Table 25 supplementary material Annex IV). The data were not of good quality for statistical analysis because experimental plots were affected by extreme climatic conditions as frost and hailstorm.

Participatory Farmer Field Experiment III and IV

During 2012-2013, 10 selected bacterial strains belonging to CIP´s collection (See Table 26 supplementary material Annex IV) were locally formulated at pilot scale and were applied at planting and hilling to potato cultivar Peruanita at two different field locations in Huancani at 3834 masl (Experiment III) and San Jose de Aymara at 4009 masl (Experiment IV), in order to evaluate inoculation of microorganisms in combination with chemical fertilization rate (50-50-50 NPK kg ha-1) on growth and yield of potato under field conditions.

Treatments were replicated four times in completely randomized designs. Each replicate consisted of 90 planted tubers per plot. At planting, bacterial inoculations were prepared by suspending 125 ml of bacterial suspensions at the concentration of 1010 CFU ml-1 in 2 L of river water. Bacterial applications were performed using the dipping method in which sprouted tubers were dipped in this mixture for about 20 min prior to planting. Then, tubers were placed in the furrows; additionally a similar mixture was applied directly on the tubers previously planted. At hilling (50 days after planting), similar mixtures made of 125 ml of bacteria suspension (1010 CFU/ml) diluted in 2 L of water were applied to each replicate.

At planting, all experimental plots received an equivalent to 5 t of sheep manure ha-1 and mineral fertilizer at 50-50-50 NPK kg ha-1 before application of bacterial inoculation. Three control treatments were included without bacteria inoculation; 1) without chemical fertilization 2) low chemical fertilization (50-50-50 NPK kg ha-1), and 2) high chemical fertilization (180-100-100 NPK kg ha-1). The experimental design consisted of a completely randomized design. The results were analyzed by ANOVA and means were separated according Tukey’s test at p<0.05 level of significance to analyze statistical differences and to discriminate between means. When no significant value was detected, a non-parametric Kruskal-Wallis test was used.

Results: The results of experiment III showed a statistically significant difference between T4 (inoculation of P. putida strain P1-20/08 ) and the non-inoculated control at the equivalent fertilizer level (p< 0.05) with an increase in tuber yield of 26 % (See Table 27 supplementary material Annex IV). The positive effect of inoculation with P1-20/08 was consistent with those from the 2011-2012 season where the treatments inoculated with P1-20/08 on potato cultivar Peruanita resulted in significantly higher tuber yields (36.47 tons ha-1) than the non-inoculated treatment at the equivalent fertilizer level. Therefore, the strain P1-20/08 should be considered for pilot formulation for further testing.

The results of experiment IV showed no statistical differences between treatments (See Table 28 supplementary material Annex IV). The data were not of good quality for statistical analysis because experimental plots were affected by extreme climatic conditions as frost and hailstorm.

Conclusions: The results from the participatory field trial evaluation of experiment I in Huancani indicated that the most accepted treatments evaluated at flowering and harvest stage by 40 farmers (12 women and 28 men) were: 1) SYMPLANTA® + manure + NPK; 2) LMG 24422 + manure + NPK 3) R43631 + manure + NPK and 4) R43628 + manure + NPK. These four farmer selected technologies will be promoted with partner organizations and farmers in order to continue the evaluation of suitable technologies over time and among agro ecology zones. The farmer participatory evaluation and ranking method facilitated discussion and prioritization of individual preferences concerning selection criteria of new technologies. The farmer participatory evaluation also strengthened the decision-making capacity of farmers with active participation of all participants including the women.

PROINPA: Results of 4 different field trials carried out in Bolivia are presented below.

Participatory Farmer Field experiment I

The evaluation of four VALORAM bacteria (see below) isolates and 3 AMF (see below) with and without fertilization on the potato cvs. W’aycha and Única was performed in two localities (Colomi at the Liruni community (3262 masl) and Tiraque at the community of Virvini (3227 masl)).

- Four VALORAM bacteria isolates: Curtobacterium flaccumfaciens (R42100), Bacillus Mycides (R41850), Pseudomonas jessenii (R42086) and Bacillus weihenstephanensis (R417989)
- A commercial formulated bacterial product: Bacillus subtilis from BioTop
- Three commercial formulated AMF products with R. irregularis: one from SYMPLANTA, Germany; and two from MYKE PRO in Canada, ASP-A as liquid formulations, and ASP-A as solid formulation.

Two fertilization regimes (i.e. with and without synthetic fertilizers) were applied. In all treatments, chicken manure was provided. Only in a control treatment without inoculum and without mineral fertilization, chicken manure was omitted.

Each treatment consisted of 6 plants, and each treatment was replicated in five plots.

Besides yield, tuber damage caused by Spongospora was evaluated.

The field located in Tiraque was planted and evaluated with participation of 21 local farmers and the one in Colomi with 7 local farmers.

Results from Virvine community, Tiraque
Wald test (See Table 29 in supplementary material Annex IV) shows statistical differences for emergence, total tuber yield and percentage of Spongospora free tubers.
Table 30 (See supplementary material Annex IV) shows that highest yield was obtained with Bacillus mycides (R41850) and 80-120-00 (3.870 kg/6 plants). The lowest tuber damage by Spongospora (62%) was observed with the same bacteria isolate but without chemical fertilizers

Results from Liruni community, Colomi
Wald test (See Table 31 in supplementary material Annex IV) shows statistical differences for emergence, total tuber yield and percentage of Rhizoctonia free tubers.
According results (See Table 32 in supplementary material Annex IV) the highest total tuber yield was obtained with just chicken manure incorporation (2.6 kg) although no significant different to R42100 C. flaccumfaciens, R41850 B. mycides, R42086 P. jessenii and B. subtilis Commercial from BioTop. Although not significant, the highest yield for a product was obtained with R. irregularis SYMPLANTA with N-P-K, while the same product resulted in low yield under no N-P-K application, which is difficult to explain.
The highest percentage of Rhizoctonia free tubers was obtained with absolute control (86 %) and the lowest with the isolate R42086 Pseudomonas jessenii with or without fertilizers.
Conclusions: No definitive results have been obtained after the evaluation of 21 treatments (four VALORAM bacteria isolates and 3 AMF-containing products with and without fertilization on the potato cvs. W’aycha and Única) in two localities (Virvine community, Tiraque and Liruni community, Colomi), whereas Rhizophagus irregularis SYMPLANTA has a trend showed the highest yields for both treatments, this is without and with N-P-K. The highest tuber yield was obtained with the application of Bacillus mycides (R41850) and 80-120-00 N-P-K application (3.870 kg/6pls); the highest percentage of tubers free of Spongospora (62%) was obtained also with Bacillus mycides (R41850) but without chemical fertilization. The highest percentage of healthy tubers in both cvs. was obtained with Bacillus amyloquefaciens plus the 80-30-00 fertilizer rate.

Participatory Farmer Field experiment II

A field evaluation of 17 bacteria isolates from VALORAM inoculated on potato cv. Waycha was performed in Virvine, Tiraque (3232 masl).

Three control treatments were applied: a commercial inoculum of Bacillus subtilis (from BioTop private company), a native isolate of B. amyloquefaciens (from BioTop) and a treatment with no inoculation. For B. subtilis,

Three fertilization regimes were considered: no fertilizer, 80-30-00 and 80-120-00 N-P-K. At planting time an overall incorporation of chicken manure was carried out (7 t/ha).

This evaluation was conducted with 16 local farmers who participated in evaluating the results of this experiment.
List of evaluated bacteria isolates and results obtained are presented in Table 33 (See supplementary material Annex IV).

Conclusions: As showns, no significant differences were found in tuber yield of different treatments, however in spite of irregular weather conditions in the area the highest percentages of healthy tubers (Rhizoctonia free) and tuber yield of the potato cv. W’aycha were obtained with Bacillus weihenstephanensis (R41798), Pseudomonas jessenii (R41805), P. thivervalensis (R41947), P.moraviensis (R42071) and P. thivervalensis (R42090). Considering Farmers’ opinions the best treatments were: Bacillus weihenstephanensis (R41798), Pseudomonas moraviensis (R42020) and P. marginalis (R42058) because tubers were more size and shape uniform.

Participatory Farmer Field Experiment III

The purpose of this experiment was to evaluate commercial bacteria and AMF products on potato (cv. W’aycha and Única) yield under three different fertilizer rates.
Three commercial AMF products were considered:

- Rhizophagus irregularis from SYMPLANTA, Germany,
- R. irregularis from MYKEPRO ASP-A with a liquid formulation, Canada
- Rhizophagus irregularis from MYKEPRO ASP-A with a solid formulation, Canada
Two bacteria products were considered:
- Bacillus subtilis from BioTop
- B. amyloquefaciens native bacteria from Bolivia from BioTop
- A control which received no inoculum.

Three different fertilizer regimes were considered (00-00-00, 80-30-00 and 80-120-00 N-P-K). Fertilizer was applied at planting time. Treatments were distributed as a complete Randomized Block Design with 3 replications; experimental plots were 12 m2. Regarding tuber yield statistical analysis (See Table 34 in supplementary material Annex IV) differences were found for fertilization rates, Microorganisms * fertilization; variety and the interaction variety * fertilization * microorganisms.

Statistical analysis (See Table 35 in supplementary material Annex IV) shows significant differences for occurrence of Rhizoctonia between blocks, microorganisms and fertilization as well as for the interactions between them meaning that Rhizoctonia distribution in plots was heterogeneous.
Considering results in Table 36 (See supplementary material Annex IV) the highest number of healthy tubers Rhizoctonia free of both cvs was obtained with the fertilizer rate 80-30-00 mixed with the bacteria B. amyloquefaciens.
The highest tuber yield was obtained with the potato cv. Única fertilized with rate 80-120-00 and the liquid formulation of R. irregularis ASP-A from Canada (See Table 37 in supplementary material Annex IV).

Conclusions: Statistical analysis showed significant differences for occurrence of Rhizoctonia and the highest number of healthy tubers free of disease in both cvs was obtained with the fertilizer rate 80-30-00 mixed with the bacteria B. amyloquefaciens. On the other hand, the highest tuber yield was obtained with the potato cv. Única fertilized with rate 80-120-00 and the liquid formulation of R. irregularis ASP-A from Canada.

UTPL: Field experiments were carried out only in experimental plots in Ecuador from February-July 2013.

Two localities were selected in Ecuador: Santa Catalina–INIAP (0°22’S, 78°33’W, 3058 m amsl) - Quito and La Guangora (04°0’1.59’’S, 79°10’48.46’’W, 2160 m amsl)– Taquil - Loja. The plant material used were Potato tubers (basic seeds) of var. I–Fripapa 99 (Solanum tuberosum x andigena) provided by INIAP (Instituto Nacional de Invesvestigacion Agropecuaria)
Eight different inoculants were tested: three mycorrhizal: R. irregularis DAOM197198 (Myke Pro-801-Canada), R. irregularis MUCL41833 (from GINCO-Belgium) and R. irregularis MUCL41833 (Symplanta-Germany) and six bacterial strains: Paenibacillus sp (R47065), Paenobacillus sp. (R49541), Bacillus simplex (R49538), Paenibacillus sp. (R47131) and Bacillus subtilis FZB24 (commercial strain) under two fertilization regimes (organic and organic+chemical fertilization) (See Table 38 in supplementary material Annex IV).

The evaluated parameters were: emergence, plant height, number of tubers produced by plants, total yield and incidence and severity of R.solani. A two way ANOVA, was applied to each parameter to see if there was an effect of each factor (inoculum and fertilization) and/or their interaction.

Results: Several parameters were considered and they are presented below

a) Plant height: In Locality 1 (Estación Experimental Santa Catalina-INIAP-Quito) There were no effects due to the inoculum, fertilization level, or their interaction (inoculum * fertilization).

The results in Locality 2 (Estación Zamora Huayco-UTPL-Loja), displayed a different trend. The two-way ANOVA (See Table 39 supplementary material Annex IV) shows significant differences for the inoculum, fertilization and the interaction (inoculum * fertilization).
In Loja the tallest plants were found in the treatment with the bacterial strain (R47065) Bacillus sp amended with chemical ferilizer (42,2cm) and the smallest plants in the absolute control (28,7cm). In both places there was a positive influence of the addition of chemical fertilizer in the increase of plant size, although in Quito the differences were non-significant (See Figure 15 supplementary material Annex IV).

b) Number of tubers: The results of the two-way ANOVA (See Table 40 supplementary material Annex IV) of the effects of the main treatments and their interaction on the number of tubers per plant in Quito showed a significant effect of inoculum source and fertilization level, and their interaction. The mycorrhizal treatments displayed the lowest values in comparison with the bacterial inoculants.

In Loja, there was also an effect of both factors and their interaction as shown in Table 41 (See supplementary material Annex IV).

The treatments that displayed the highest number of tubers/plant in Quito were the commercial bacteria (B. subtilis FZB24) and the bacterial strain R49541 amended with chemical fertilizer (13.53 and 12.90 respectively). In Loja, the chemically fertilized treatments produced higher number of tubers and the treatment with the bacterial strain R47065 displayed the highest number of tubers/plant (16.33) (See Figure 16 supplementary material Annex IV). Its intersting to note that although the treatment with the commercial bacteria resulted in a high number of tubers per plant in Quito, in Loja the results were just the the contrary yielding only 4.8 tubers/plant (See Figure 16 supplementary material Annex IV).

c) Yield: In Quito there was an effect on yield (tons/ha) due to the fertilization level (p=0,000) and inoculum type (p=0.000) and their interaction (p=0.049). The highest yield (43.69tn/ha-1) was reached with the bacterial strain R49541 (Paenobacillus sp B3a) amended with chemical fertilizer and the lowest yield (17.63 tn/ha-1) was obtained with the mycorrhizal inoculum MUCL41833 (R. irregulare) without chemical fertilization. The mycorrhizal treatments and the non-inoculated non-fertilized control displayed the lowest values (See Table 42 supplementary material Annex IV).

In Loja there was an effect due to the inoculum (p=0.000) and fertilization (p=0.000) separately but not for the interaction (p=0.060) (See Table 43 supplementary material Annex IV). The highest yield was reached with R47065 (20.6 tons/ha) amended with chemical fertilizer; the lowest value was obtained with the non-inoculated non-fertilized treatment (8,87 tons/ha) (See Figure 17 supplementary material Annex IV).

d) Incidence and Severity of Rhizoctonia solani in tubers: The data of severity of R. solani in Quito and Loja did not follow a normal distribution and could not be transformed. However, the no parametric test of Kolmogorov Smirnov showed significant differences among treatments at both sites (p<0.000).

The data of incidence of R. solani were transformed with the natural logarithm in order to get a normal distribution of the data. After that a two-way factorial ANOVA (See Table 44 supplementary material Annex IV) was applied to see if there was an interaction between inoculants and fertilizer level, the results showed a significant interaction (p=0.045) but when a one-way ANOVA and a post hoc analysis with Tukey, were applied the results did not show significant differences among treatments (p= 0.200). In Loja, the results showed (See Table 45 supplementary material Annex IV) a positive interaction between fertilizer and inoculum type (p=0.000). The treatment that showed the lowest incidence of R. solani was the bacterial treatment R47065+NPK.

In Quito there were no significant differences among treatments (See Figure 18a supplementary material Annex IV). On the other hand, in Loja some significant differences were shown between treatments (See Figure 18b supplementary material Annex IV).

Conclusions: The results from both experimental locations were evaluated separately. The addition of NPK had a positive effect in almost all the parameters evaluated (plant height, number of tubers/plant and yield).

In general, the bacterial inoculants yielded the best results when amended with NPK, the treatments with the highest yield were: R49541 +NPK in Quito and R47065 + NPK in Loja.
The results showed a positive effect of rhizobacteria inoculations on potato tuber growth. The yield was significantly improved following inoculation with R47065, R49541, R47131, R49538 and the commercial inoculant Bacillus subtilis FZB24® WG. The positive effect of bacterial inoculants on yield was generally consistent, although not always statistically significant and was generally easier to detect at the low fertilizer level. The experimental design used, amending all treatments with organic manure only revealed an interaction effect of the fertilizer level with the microbial inoculants at the Quito site. It would be important for future studies to include combinations of rates and types of fertilizer with each PGPB strain to optimize their use.

The mycorrhizal formulations showed a low yield maybe because of the competition with native microorganisms or the harsh conditions of the soils. We studied disease incidence on tubers from naturally occurring inoculum, but did not see significant effects of treatments on disease on tubers. In the experiment in Quito the treatments with R49541, R47131 and R49538 with organic manure alone, all resulted in higher tuber yields than the non-inoculated treatment with organic manure, and statistically similar tuber yields to the control treatment amended with chemical fertilizer. This indicates that the respective rhizobacterial strains have potential to be used in commercial formulations as bio-stimulants that enhance adsorption of nutrients, suppress pathogens and strengthen plant growth. The yields obtained with the bacterial inoculants, which were similar to the control treatment amended with chemical fertilizer, also indicate that these inoculants can be used under high-input systems to improve fertilizer-use-efficiencies with the objective of reducing application rates of agrochemicals but maintaining productivity, thereby reducing production costs and adverse environmental effects caused by agrochemicals.

It would be interesting to test the combination of bacteria and AM fungi in the field, due to the complementary effects of both microorganisms to improve plant performance. A previous experiment under strict in-vitro conditions carried out at UCL showed a positive effect of one rizobacterial strain R47065 isolated from Ecuadorian fields (in the frame of the VALORAM project) on the performance of the AM fungus Rhizophagus irregularis. The bacterium was able to improve the germination percentage of R. irregularis spores and to increase the length and number of branches of the hyphae in comparison with the control without bacteria. When the fungus was associated to a living host (S. tuberosum) the treatment with the above mentioned bacteria yielded a higher number of AMF spores and length of hyphae.

WP7 – Ecological impact of improved soil management systems

Graphs, Figures, Tables are in supplementary material under heading Annex IV.

• Main result 1: Identification of microorganisms that improved plant performance and analysis of the impact of their introduction on native microbial communities.

The results of inoculation trials showed that the following bacterial strains improved plant performance when used together with NPK: Bacillus mycoides, Bacillus amyloliquefaciens, Paenibacillus sp. B3a (R49541), Bacillus subtilis (FZB24), Bacillus positive (R47065 and R47131) and Bacillus simplex (R49538).

Regarding changes on the community patterns following microbial inoculation, results show that there was no significant effect on the abundance of selected bacterial taxonomic and functional groups (Actinobacteria, Alphaproteobacteria, Betaproteobacteria, Firmicutes, Bacteroidetes) following the inoculation of the Peruvian rhizosphere samples with Pseudomonas fluorescens R43631. However, ANOVA tables showed a significant (p<0.05) effect of the commercial AMF DAOM 197198 SYMPLANTA on the rhizosphere abundance of Actinobacteria (p<0.05) and enhanced probability (p<0.1) of an effect on Betaproteobacteria and nifH gene abundance (See Table 46 supplementary material Annex V). In Bolivia, there was a significant effect on Betaproteobacteria, Actinobacteria and nifH gene abundance following inoculation of Bolivian rhizosphere samples with the selected Bacillus mycoides strain (See Table 47 supplementary material Annex V).

For AMF, the inoculation with R. irregularis did not change the AMF community composition, as inoculated and non-inoculated sites had similar species amounts (See Figure 19 supplementary material Annex V). Moreover, in some fields, R. irregularis had very low abundance suggesting that inoculation was not leading to successful establishment of the species (See Figure 20 supplementary material Annex V), even though R. irregularis was inoculated in high amounts (250-500 spores per tuber) in field experiments performed in Ecuador (See Figure 21 supplementary material Annex V). Species belonging to Acaulospora were the dominant colonizers of potato roots, therefore these species may be suitable candidates for future experiments and inoculation schemes, where a mixture of putatively functionally complementary AMF might me much more efficient than inoculation with a single non-native strain.

In: Lojan P., Senés-Guerrero C., Kromann P., Suárez J.P. Schüßler A., Declerck S. Field inoculation and molecular tracing of Rhizophagus irregularis applied as different mycorrhizal commercial formulations on potato (Solanum tuberosum L. “I Fripapa”) cropping. In preparation.

• Main result 2: Effect of microbial inoculants on fertilizer use and cost benefit analysis of the use of beneficial soil microorganisms to decrease fertilizer input.

Decreasing the NPK dosage by half is possible if it is done in combination with the inoculation of bacteria that showed improvements in plant performance (Bacillus mycoides, Bacillus amyloliquefaciens, Paenibacillus sp. B3a (R49541), Bacillus subtilis (FZB24), Bacillus positive (R47065 and R47131) and Bacillus simplex (R49538)).

It was not possible to determine a single bacterial strain that would be suitable for all types of environments. Thus, using a combination of the before mentioned bacteria in a mixed-inoculum would probably be a reasonable strategy for application.

The results of the field trials showed that not adding fertilizers had a negative effect on yield, even when using microbial inoculants. However, reducing the NPK dosage by half and adding bacterial inoculants resulted in the same plant performance as when using the full NPK dosage, clearly indicating the potential of microorganism inoculation. Consequently, based on the previous results, a good compromise would be to use a mixed bacterial inoculum combined with a decrease in NPK dosage by half.

The costs of production of one hectare of potato depend on the country and whether is a low or high input system. In general, total costs range from US$ 1,200 to 3,300 dollars and the costs of fertilizers range from US$ 250 to 1000 dollars. Therefore, reducing the amount of fertilizers by half could represent substantial savings for farmers that could range from US$ 75 to 450 dollars, considering that the cost to produce bacterial inoculum for one hectare is US$ 50 dollars.

• Main result 3: Recommendations for sustainable potato agriculture in the Andes.

The suggestions presented here are based on the analysis of bacteria and AMF native communities obtained in WP3 and from tracing microbial inoculants during field trials in WP6.

Bacterial strains: We identified some bacterial strains, Bacillus mycoides, Bacillus amyloliquefaciens, Paenibacillus sp. B3a (R49541), Bacillus subtilis (FZB24), Bacillus positive (R47065 and R47131) and Bacillus simplex (R49538), that improved plant performance when used together with NPK. Therefore their use as microbial inocula in potato fields is recommended, preferably in a mixed-inoculum application scheme.

AMF strains: R. irregularis might be an unsuited species for potato plants or the Andean ecosystem, besides the possibility of inefficiency of a non-native isolate or putatively low vitality of inoculum. Species belonging to Acaulospora were the dominant colonizers of potato roots, beside others. Therefore these species are suggested as suitable candidates for future experiments and inoculation schemes, and it should be pointed out that a mixture of putatively functionally complementary AMF might me much more efficient than inoculation with a single non-native strain.

An easy-to-apply product is crucial. A powder formulation that would require no preparation is preferred. A liquid product is also accepted among farmers, nevertheless using liquid may not recommended due to usually shorter shelf life. This is strongly dependent on the organisms used (stability in different formulations varies greatly).

Fertilizer input: Our experiments showed that using a dosage of NPK (115-265-107), half of the recommended dosage, together with bacteria inoculation would result in the same yield as using the full dose. Not adding fertilizers had a negative impact in yield, therefore it is not recommended.

Potential Impact:
The overall objective of the project was to promote the sustainable development of potato-based systems in the Central Andean Highlands. This research objective was met by detailed soil microbial flora analysis of various soil management systems. Culture collections were established both in reputed international collections and in the countries of origin. Microorganisms with desirable properties relevant for sustainable crop production in the Andes were characterized. Promising management systems as well as biofertilizers and pest control agents were selected and further tested in field trials to elaborate their applicability and benefits in comparison to conventional practices. Field testing was accompanied by a comprehensive ecological impact assessment in order to warrant that novel and recommended practices have either a benefit on biodiversity and soil fertility or a lower impact than conventional practices.

During the project execution period several activities have been directed towards facilitating and supporting capacity strengthening activities among project collaborators and with interested researchers, organizations and other stakeholders on the use of microorganisms to improve the productivity of potato based cropping systems. Training events and training materials have been developed, and internal and external communication platforms created, developing linkages with regional initiatives and related projects on concepts of sustainable agriculture to present novel technologies from the VALORAM project and integrate results and innovation to enhance technological and socio-economic impact.

Over the past five years, we have explored microorganisms living in association with potato cultivated in the high Andean region in South America in collaboration with resource-poor farmers. All potential bacterial strains (435) and AMF strains (27) have been deposited and are being preserved in reputable germplasm collections in Europe (for bacteria in BCCM/LMG and for AMF in BCCM/MUCL) and in registered working collections in the countries of origin for the retrieval and benefit of future projects and third parties through appropriate authorization. The direct impact of valorized microbial strains is related to the best field performing bacterial strains that were selected from more than 900 field isolated strains and which have high potential to be used in biofertilizer and biocontrol applications; e.g. Paenibacillus sp. B3a (R49541), Pseudomonas palleroniana (R43631), Bacillus weihenstephanensis (R41798), Bacillus positive (R47065 and R47131) and Bacillus simplex (R49538). Concerning AMF, species belonging to Acaulospora are suggested as suitable candidates for future experiments and inoculation schemes. Following additional field trials to assess plant-growth promotion and disease suppression efficacy under multiple field environments, pilot scale production on an industrial scale using improved techniques and assessment of shelf-life and other commercial considerations, the VALORAM strains have high potential to be used in the development of commercial formulations.

A biological approach combining suitable potato genotypes, appropriate land management and inoculation with beneficial microorganisms provides a valuable opportunity for sustainable intensification of potato-based farming systems in the Andean region and a real alternative that can minimize the use of agrichemicals and protect natural resources. Preliminary cost-benefit analyses of our project results indicate that farmers may reduce their chemical fertilizer expenses to almost 50% of current conventional practices with the use of the VALORAM applied technologies with bacterial isolates.

A conservative calculation considering that the applied technologies become accessible to Andean potato farmers, and a 50% reduction in chemical fertilizer usage (average conventional cost: 100 US dollars) on 20% of the more than 400,000 ha cultivated with potato in the three Andean countries, would result in an annual production cost reduction in the range of millions of dollars, and significant impact on the livelihoods of individual rural families. In addition to economic benefits there are considerable significant environmental benefits from using microbial inoculants instead of agrochemicals, including reduced intoxication risk to farming communities related to the reduced use of pesticides.

There is an increasing demand for organic products and chemical-free, sustainably-produced food products coupled with a need to relieve pressure on natural resources. It is expected that the global biocontrol market will increase to $3 to 4 billion by 2017. Resourcing from governmental agencies, academic and industrial sectors are increasing in Europe and also in South America in order to support the development of sustainable agriculture. The global trends of the biocontrol market are continuously increasing the value of reputable microorganism collections, including the VALORAM strains and the fundamental knowledge on applied technologies developed in the project.

The wider societal implications of the scientific knowledge and technologies created by the project are so far mainly related to the strong knowledge base created by the project, which is being published in scientific journals (more than 20 papers published, in press, submitted or in preparation) coupled with other dissemination activities (> 60 posters, oral presentations…) of project results – both beneficial microorganisms and knowledge – using the networks of the consortium and the project website (http://valoram.ucc.ie) established and maintained by the consortium. Capacities have been developed and strengthened in respect to research and use of microbial soil organisms of internal and external consortium partners. The project has established links with development oriented projects being implemented by partner organizations in Bolivia, Ecuador and Peru. Protocols, scientific publications and training materials have been developed; the project documents and project website make available its research findings, expertise and knowledge resources to a wide audience.

Main result 1: Implementing a website for internal communication of the members of the consortium and external presentation of results and advances.

The Valoram project website was created by UCC (Partner 3) at the start of the project in 2009. The address for the website is http://valoram.ucc.ie. The accompanying email address (valoram@ucc.ie) is for visitors who wish to contact the project. The language of the website is English but there is also an introductory page for Spanish visitors. A mail list for communication between VALORAM partners was setup by LMU (Partner 4) using Google Groups at the project start-up. This also includes individual mail groups of the partners for each work package. This website is regularly updated with progress reports, materials and methods, conference presentations and journal publications, to allow for an exchange of information and scientific knowledge between the partners. In addition, confidential files are uploaded to a ‘restricted access’ page for project partners. Over the past 5 years (i.e month 1 to month 60), the number of visitors to the project website was 2837. The top ten countries were: Ireland (1277 visits) followed by Belgium (487), Germany (336), Austria (223), United States (202), France (186), Ecuador (180), United Kingdom (163), Peru (140) and Spain (136). Traffic sources to the website for this reporting period (i.e. month 1 to 60) comprise of 65% search traffic, 22% referral and 13% direct traffic. Six emails were received during the last five years. Two emails (2009-2011) were received from Latin America countries (Argentina, Bolivia, Ecuador, Mexico and Peru, looking for more information on the project. An email was also received from the U.S. looking for contact details of one of the project partners (2009 – 2011). One email was received to the VALORAM email account during 2012 and two were received during 2013. Both of the emails during 2013 were from television production companies (EURONEWS and ARTE) enquiring about the possibility of making a TV documentary on the project. A positive response was sent to both companies and we are now waiting to hear more. The website will be continuously updated with relevant information, publications and project final report. It is envisaged by partner 3, based in UCC that the VALORAM website will continue to be hosted on the UCC server for at least 24 months after the project finishes, in order to allow for final publications and information notices to be disseminated. Recently, the website was overhauled in order to be compatible with mobile phones.

Main result 2: Organizing workshops/trainings with Andean organizations and institutions to disseminate knowledge and information on potato-based systems using microbial inoculants.

2009 – 2010:

• A workshop/training course was organized (5-16/10/2009) at the universities of Louvain-la-Neuve (UCL) and Gent (LMU) by partners 1 and 5 with participation from all VALORAM partners and from the twinning project (Argentina). PhD students from the BIOSPAS and SOILGENE project attended;

• A seminar and training activities took place for own and/or partner staff at different laboratories (LMU on high throughput sequencing technologies; CIP on bacterial isolation for staff from PROINPA, UTPL and INIAP, Ecuador);

• Protocols were developed for diverse scientific activities (DNA extraction, sampling of microbial soil organisms, testing of bacterial strains etc.);

• During the annual 2010 meeting at CIP, Lima (15-19 February 2010), a symposium was held with approximately 90 attendees (21 from CIP, 66 from 7 different universities and 3 from private companies) from more than 10 Peruvian institutions on the ‘Utilization of beneficial microorganisms in agricultural systems: exploring microbial diversity for novel application’.

• A twinning activity was conducted with Argentina via 2 meetings in Buenos Aires (Argentina, 6-7/05/09) and in Athens (Greece, 30/06/10 to 01/07/10). In both events, VALORAM was represented and the project presented for an audience of scientists from Europe and Argentina.

2011 – 2012:

• A training workshop on soil biodiversity analysis to assess soil fertility was organized at UCC following the annual meeting in Cork in February 2011, with the participation of representatives from CIP (Peru), PROINPA Foundation (Bolivia) and UTPL (Ecuador).

• A training workshop was organized to the benefit of local Bolivian scientists and technicians at PROINPA following the annual meeting in Cochabamba in February 2012, including lectures on bacteria and on mycorrhizae, and practical trainings were given on bacteria culture and manipulations, and on mycorrhiza identification and culture.

• One training course titled “Training on the handling and use of Plant Growth Promoting Rhizobacteria under laboratory conditions” was organized at CIP in July 2012. Representatives of a private company (NOVAGRI S.A) were trained by CIP staff on techniques and use of beneficial bacteria in order to strengthen the knowledge base and facilitate information for pilot scale production.

• One training course titled “Training on the handling and use of Plant Growth Promoting Rhizobacteria under laboratory, greenhouse and field conditions” was organized at CIP in November 2012 for a total of 7 participants, representatives of the private foundation PROINPA, NGOs (DIACONIA, YANAPAI), National Agricultural Research System (INIA, SENASA), the university, (UNALM) and the private company (Beggie Peru S.A).

• One workshop titled “Introduction on uses of beneficial microorganisms and fermented biofertilizers in agriculture: Exploring biodiversity for innovative applications” was organized at CIP in November, 2012. The overall aim of this workshop was to share research results and methods developed by the VALORAM project. A total of 32 participants and representatives from NARS, NGOs, universities and the private sector attended this workshop.

• A one day workshop on the use of beneficial microorganisms in Andean Agriculture was organized with PROINPA Foundation for potato producers. 35 people from different public institutions.

2013 – 2014:

• One training course titled “Management techniques and use of plant growth promoting rhizobacteria in laboratory, greenhouse and field” was organized at CIP-Lima, Perú, December, 2013. A total of 4 participants, representatives of two private companies (Vivero Los Viñedos SAC and Ciencia para la Sanidad del Agro SAC) were trained by CIP staff on techniques and use of beneficial bacteria.

• One one-day training workshop titled “Isolation and use of disease suppressive soil microorganisms and crop yield promoters in potato” was held in Cochabamba, January, 2014. A total of 21 participants were trained representing BioTop, PROINPA, Gobernación Cochabamba, Universidad Mayor de San Simón.

• A symposium was held during the 5th annual project meeting in Ecuador organized by UTPL in January 2014, ‘Microorganisms as key component to promote sustainable agriculture’. The symposium was attended by a total of 42 participants mainly from Ecuadorian universities, national R&D institutions and private companies.

Main result 3: Establishing links with regional activities to promote the dissemination of project results.

• The VALORAM project has established links with development-oriented projects being implemented by partner organizations in Bolivia, Ecuador and Peru.

• CIP, Peru (Partner 6) - Participatory evaluations (at flowering and harvest) of field trials were established with farmers groups working with local NGOs (DIACONIA, YANAPAI) interested in the use and application of VALORAM strains.

• PROINPA, Bolivia (Partner 7) - Bio-Top, a private company and commercial branch of the PROINPA Foundation formulated biological products developed by VALORAM in Bolivia. Participatory field trials were implemented where potato farmers evaluated the effect of selected bacteria isolates and commercial AMF (including the product of Bio-Top).

• UTPL, Ecuador (Partner 8) coordinated famer participative field experiments with an EU funded and CIP coordinated regional research and development project (IssAndes) in Ecuador. The experiments were implemented with farmer communities in the province of Tungurahua with the objective to evaluate two commercially formulated VALORAM bacterial inoculants. The technology is being disseminated to local NGOs, governmental agents, universities and the national agricultural research system and farmer communities through the project IssAndes.

Main result 4: The research results of the VALORAM project are being published in different types of journals, and disseminated through different means.

More than 20 manuscripts have been published or are in press, submitted or in preparation. Among these three papers are particularly relevant owing to their large audience (review) and/or high impact factor.

• A review paper published in Mycorrhiza (Impact Factor: 2.95): Lalaymia I., Cranenbrouck S. and Declerck S. Maintenance and preservation of ectomycorrhizal and arbuscular mycorrhizal fungi (2014). 24 (5): 323-337.

• A paper accepted for publication in Trends in Biotechnology (Impact Factor 9.6): Velivelli S.L.S. De Vos P., Kromann P., Declerck S., Doyle Prestwich B. Biological Control Agents: From Field to Market, Problems and Challenges.

• A regular paper submitted for publication in ISME Journal (impact factor: 8.95). Pfeiffer S., Mitter B., Oswald A., Schloter-Hai B., Schloter M., Declerck S., Sessitsch A. Major drivers of potato-associated bacterial communities and the core rhizosphere microbiome of potatoes cultivated in their natural habitat, the Central Andean Highlands.

More than 60 posters, oral presentations… in congresses, symposia, meetings.

List of Websites:
http://www.ucc.ie/en/valoram/

Final Report Summary - VALORAM (Valorizing Andean microbial diversity through sustainable intensification of potato-based farming systems)

Share this page

Download