Characterisation of for the new annual

pasture legume muricatus targeted for medium-to-low rainfall areas of southern Australia.

Kit Alexander Burns

2019

This thesis is presented for the degree of Bachelor of Science Honours

in Molecular Biology

School of Veterinary and Life Sciences

Murdoch University

I declare that, unless otherwise stated, the work presented in this thesis is my

own and has not previously been submitted for a degree at any tertiary

education institution.

Kit Alexander Burns

ABSTRACT

Legumes play an integral role in increasing agricultural productivity, particularly in low input agricultural systems in Australia, due to their ability to form symbiotic interactions with a group of soil called rhizobia. However, in medium-to-low rainfall areas of southern Australia, there is a lack of suitable annual pasture legumes, which is limiting agricultural productivity and profitability in these farming systems. is an annual legume from the Mediterranean which possesses high nutritive value and palatability for livestock, is high yielding, capable of self-seeding and is well- adapted to hot and dry summers. As such, S. muricatus is currently being evaluated as a new pasture legume for southern Australia. Crucial to the success of introducing this legume will be the availability of a highly effective rhizobial inoculant strain. This thesis therefore sought to characterise the phylogeny, free-living and symbiotic phenotype of a range of bacteria isolated from Scorpiurus spp.

A total of 19 strains were investigated, with 16s rRNA sequencing demonstrating that 18 of these strains belonged to the genus , with the remaining strain (WSM1184) most closely related to Agrobacterium tumefaciens. Analysis of nifH and nodC symbiosis genes further showed that the characterised Mesorhizobium strains generally shared highly similar sequences for these loci, indicating a comparatively high degree of genetic similarity. In particular, WSM1343 (isolated from

Scorpiurus sulcatus growing in Morocco) and WSM1386 (isolated from S. sulcatus in Manjimup,

Western Australia) were shown to share highly similar symbiosis genes, but divergent 16S rRNA genes, suggesting the possibility that these strains may contain symbiosis genes on mobile Integrative and

Conjugative Elements (ICEs).

While the temperature tolerance and apparent optimum growth temperature of the test strains of

28°C was consistent with that commonly reported for Mesorhizobium spp., their growth rate was atypical for this genus, with 15 of the 18 strains having a growth rate on YMA at 28°C slower than that generally described for Mesorhizobium. This slower growth rate may be a common feature of rhizobia

iii from S. muricatus nodules and therefore should be considered when isolating organisms from this legume.

Symbiotic effectiveness experiments showed all Mesorhizobium strains nodulated S. muricatus and fixed N2 on this host, with the most effective strain producing 67.5% of the mean shoot dry weight of the N-fed control . Host range experiments demonstrated a subset of the Mesorhizobium strains nodulate existing Australian commercial pasture legumes Biserrula pelecinus and Lotus corniculatus, with the effectiveness data suggesting these strains fix N2 poorly on both hosts. In contrast, none of the strains tested were able to nodulate the grain legume Cicer arietinum.

While this thesis has characterised the phylogeny, free-living and symbiotic phenotype of a range of

S. muricatus microsymbionts, further work is required before a suitable commercial inoculant strain can be recommended for this pasture legume. First, all the strains tested in this thesis were isolated from S. sulcatus plants or soils with Scorpiurus spp. present, rather than S. muricatus and it is not known whether strains from either species are cross-compatible for effective N2 fixation. Future studies may therefore locate more effective N2-fixing rhizobia for S. muricatus by isolating microsymbionts from this host in the field. Second, experiments testing the ability of commercial inoculants for already-established pasture legumes B. pelecinus (WSM1497), Lotus sp. (SU343, CC829) and the grain legume C. arietinum (CC1192) to nodulate and fix N2 on S. muricatus need to be conducted to determine whether these inoculants will interact with this legume. Finally, the data strongly suggest that S. muricatus-nodulating Mesorhizobium spp. may contain symbiosis genes on mobile symbiosis ICEs. Given that the phenomenon of ICE transfer has led to the evolution of poorly effective microsymbionts for B. pelecinus, it is imperative that these S. muricatus strains be interrogated for the presence and transfer of symbiosis ICEs, in order to manage this mobility in any future commercial inoculant strain that is released for this pasture legume species.

iv

TABLE OF CONTENTS

ABSTRACT ...... iii

ACKNOWLEDGMENTS ...... vii

ABBREVIATIONS...... ix

1.0 INTRODUCTION AND LITERATURE REVIEW ...... 1

1.1 Rhizobia-legume symbioses...... 1

1.1.1 Legumes and their importance in agriculture ...... 1

1.1.2 Establishment of N2-fixing symbioses ...... 2

1.2 Rhizobial Diversity...... 5

1.3 Mesorhizobium ...... 6

1.4 Mesorhizobium spp. as inoculants in agricultural systems of Australia and New Zealand ...... 11

1.5. Scorpiurus muricatus: A potential new annual pasture legume for southern Australia ...... 13

1.6 Aims ...... 16

2.0 METHODS ...... 17

2.1 Bacterial strains and accessions ...... 17

2.2 Growth media and culturing conditions ...... 17

2.3 Amplified fragment length polymorphism (AFLP) assessment of diversity by RPO1-based PCR18

2.4 Phylogeny based on 16S rRNA gene ...... 19

2.5 Phylogeny based on nifH and nodC symbiosis genes ...... 20

2.6 Assessment of symbiotic phenotype of strains on Scorpiurus muricatus ...... 22

2.7 Assessment of nodulation and nitrogen fixation on other legume species ...... 25

3.0 RESULTS ...... 26

v

3.1 Free-living phenotype of strains ...... 26

3.1.1 Growth on ½ LA, YMA and TY ...... 26

3.1.2 Temperature tolerance ...... 26

3.2 Diversity and phylogenetic relationships ...... 29

3.2.1 Diversity based on AFLP ...... 29

3.2.2 Phylogeny based on 16S rRNA gene ...... 29

3.2.3 nifH and nodC ...... 31

3.3 Symbiotic phenotype of strains on S. muricatus ...... 36

3.3.1 Yield and nodulation of S. muricatus ...... 36

3.4 Host range of Mesorhizobium strains...... 41

4.0 DISCUSSION ...... 50

4.1 Growth characteristics and diversity of Mesorhizobium strains ...... 50

4.2 Symbiotic effectiveness and host range ...... 53

4.3 Concluding statement and future directions...... 55

5.0 BIBLIOGRAPHY ...... 58

6.0 APPENDIX ...... 73

vi

ACKNOWLEDGMENTS

I would like to extend my thanks to the many people who made it possible for me to complete this honours project.

Firstly, to my primary supervisor, Dr Jason Terpolilli, for your never-ending support, unwavering enthusiasm for the project and the amount of time you committed to answering my many questions when you heard me walking down the hallway. Even overseas conferences could not stop the valuable feedback from being delivered to my inbox. I have learnt a great deal in this short period of time and will continue to use these lessons in future endeavours.

To my secondary supervisors, Dr Graham O’Hara and Dr Ron Yates. Graham, thank you for providing me with some of your extensive knowledge, explaining how to setup robust experiments and making me see things from a point of view that I may have otherwise missed, your experience and eagle- eyed editing skills have been greatly appreciated. Ron, thank you for working through many of the agricultural aspects of the project, your feedback on glasshouse experiments and writing has provided me with an extra edge for this and future enterprises.

Many thanks to the AW Howard Memorial Trust for awarding me the honours scholarship, these funds have enabled a more in-depth investigation and allowed me to apply more time and effort to this project. This project was also supported by the Grains Research and Development Corporation,

Dryland Legume Pasture Systems, Meat and Livestock Australia and Australian Wool Innovation.

To all the Centre of Rhizobium Studies (CRS) members, thank you for extending a helping hand at every opportunity and making this year run more smoothly than I could have hoped for. Special thanks to Yvette, your expertise and assistance in the glasshouse, lab work and computer programs saved me many hours/days of hardship. To Reg, thank you for showing me the ins and outs of glasshouse experimental work, the help you provided and the weekly talks about all creatures large and small. Talitha and Georgina, your assistance in the lab has been invaluable, providing me with

vii laboratory supplies whenever I needed them and lab protocols improved through your own experiences, thank you both. Emma, thank you for your help with Rstudio, using the appropriate statistical tests took a little longer but I can be confident with the data. Brad, the process of introducing Scorpiurus muricatus as a new pasture legume is largely due to your work, thank you for providing me with the seeds and opportunity to work with this plant.

I’d like to thank all my family and friends, your support has made my extended education possible and I am very grateful. To Stu and Iain, thank you for great accommodation with endless chicken and duck stress relief during my years of study and friendship over many more.

Finally, to my parents, Sharron and Cliff. The support you have provided me from over my entire life has made it possible to dream about and do research. You’ve always believed that I could and would reach this goal, Thanks Mum and Dad.

viii

ABBREVIATIONS

°C Degrees Celcius

½ LA ½ Lupin agar

AFLP Amplified Fragment Length Polymorphism

AGRF Australian Genomics Research Facility

ANOVA Analysis of variance

BLASTN Basic Local Alignment Search Tool (nucleotide)

BNF Biological nitrogen fixation bp Base pair

Cat. Catalogue

CC Canberra rhizobium Collection

CFU Colony forming unit

CRS Centre for Rhizobium Studies

DI water Deionised water

DMSO Dimethyl sulfoxide

EtBr Ethidium bromide fix Nitrogen fixation genes

ICARDA International Centre for Agricultural Research in the Dry Areas

ICEs Integrative and conjugative elements

KNO3 Potassium nitrate

ix

LB Lysogeny broth

NaOCl Sodium Hyperchlorite nif Nitrogen fixation genes nifH Gene encoding the Fe-protein of nitrogenase nod Nodulation genes nodC Gene encoding N-acetylglucosaminyltransferase involved in Nod factor

synthesis

OD600 Optical density λ = 600 nm

P Probability

PCR Polymerase Chain Reaction reps Repetitions

RFLP Restriction Fragment Length Polymorphism rRNA: Ribosomal ribonucleic acid

TY Tryptone yeast (agar)

UV Ultraviolet

WSM Western Australian Soil Microbiology culture collection

YMA Yeast mannitol agar

x

1.0 INTRODUCTION AND LITERATURE REVIEW

1.1 Rhizobia-legume symbioses

1.1.1 Legumes and their importance in agriculture

Legumes have played important roles in agriculture for many thousands of years, being a valuable source of food for humans and livestock. Many ancient cultures included legumes in early farming systems, recognising that they improved soil fertility for subsequent non-legume crops (Heiser, 1981,

Graham & Vance, 2003). Only in recent times was this beneficial effect recognised to be the result of a symbiosis formed between soil bacteria, known as rhizobia, and legumes, in which the bacteria convert atmospheric nitrogen (N2) into plant available nitrogen, providing this element to the plant host in a process known as N2 fixation (Lewis et al., 2005). Although bioavailable nitrogen can be supplied to crops through the industrial production of nitrogen fertilisers, these processes are energy expensive, requiring high temperatures and pressures usually generated by burning fossil fuels (Drew et al., 2014). The finite supply and increasing demand of fossil fuels has led to increasing costs in fertiliser production, making N2 fixation a low-cost alternative (Drew et al., 2014).

The Legume family () is the third largest family, with 751 genera containing

19,500 species (LPWG, 2013, Lewis et al., 2013). Members of this family are globally distributed, forming ecologically important components of arid, Mediterranean, peat swamp, rainforest, savanna, seasonally dry, temperate and tropical ecosystems, with a wide diversity of plant forms within the family ranging from large rainforest trees and vines to shrubs, annuals and aquatic plants (Lewis et al.,

2005, LPWG, 2013, Yahara et al., 2013).

Globally, it has been estimated that grain and pasture legumes are cultivated on 180 million ha, representing 12-15% of the Earth’s arable surface and accounting for 27% of the world’s primary crop production (Graham & Vance, 2003). In Australian agricultural systems, rhizobia-legume symbioses fixes approximately 2.7 million tonnes of N2 annually, with an estimated value of $4 billion (Drew et

1 al., 2014). Pasture legumes have provided the meat, wool and dairy industries with a rich source of protein, fibre and energy for centuries (Graham & Vance, 2003, Lewis et al., 2005). Pasture legumes are prevalent in developing countries where the meat and dairy industry is almost solely dependent on them, and in developed countries such as the U.S. and Australia where Medicago sativa,

(alfalfa/Lucerne) is widely utilised as fodder and permanent pasture along with subterranean clover, annual medics and white clover (Graham & Vance, 2003, Nichols et al., 2012) as well as many newly introduced pasture species such as Biserrula pelecinus and Ornithopus sativus (French serradella) (Loi et al., 2005).

Nitrogen is an essential component of all life forms, however, the availability of this element is a major limiting factor to plant growth in agricultural systems (LeBauer & Treseder, 2008, Robertson &

Vitousek, 2009). Although N2 gas comprises 78% of the Earth’s atmosphere (Barry & Chorley, 2003), plants cannot use nitrogen in this form. Microorganisms known as diazotrophs are able to enzymatically reduce N2 into ammonia (NH3), subsequently making nitrogen available to plants as

+ ammonium (NH4 ), where it is ‘fixed’ into plant tissue and returned to the soil by plant decay and senescence, harvested by humans or grazed by livestock. It has been estimated that between 50 – 70

Tg of N2 is fixed by this process in agricultural systems annually (Herridge et al., 2008).

1.1.2 Establishment of N2-fixing symbioses

N2-fixing symbioses are established when legume roots are infected by rhizobia. There are three recognised modes of infection of legume roots: (1) Root hair curling, (2) Epidermal infection (e.g.

Lupinus spp.) and (3) Crack entry (e.g. Arachis sp.) (Ibanez et al., 2017). Most legumes of agricultural importance are infected by root hair curling, which will therefore be the focus in this review (Downie,

2010).

Legumes exude a wide range of molecules into the root zone (termed the rhizosphere), including flavonoids and isoflavonoids (Peters & Verma, 1990, Shimada et al., 2003). These molecules activate rhizobial NodD proteins, which in turn direct the transcription of a suite of nodulation (or nod) genes

2

(which include nod, noe and nol) (D'Haeze & Holsters, 2002) that direct the synthesis and secretion of lipochitooligosaccharide signalling molecules termed Nod factors. The combined action of the products of the nodABC genes, where NodC is an N-acetylglucosaminyltransferase, NodB a deacetylase and NodA acyltransferase, results in the synthesis of core Nod factor, with decorations and modifications of this molecule catalysed by a series of other nod gene products (Perret et al.,

2000). Mature Nod Factors are subsequently secreted by rhizobia, binding to cognate Nod Factor receptors expressed on the surface of root hair cells. Binding of Nod factor to the receptor activates nodule organogenesis which triggers root hair deformation, intracellular calcium oscillations, membrane depolarization and initiation of cell division in the root cortex, leading to the establishment of a meristem and nodule primordium (Oldroyd & Downie, 2004, Timmers et al., 1999).

To reach the developing nodule, rhizobia must first proceed from the root surface to the inner root tissue which is achieved by the formation of an infection thread. Infection thread formation occurs when rhizobia become trapped between root hair cell walls due to root hair deformation (Callaham &

Torrey, 1981). An invagination of the plant cell wall at a point adjacent to the bacterial infection locus allows rhizobia to grow and divide down the infection thread, as the thread extends through the cortical cells, to the developing nodule primordium (Gage, 2004). As infection threads grow through the root and enter the nodule primordium, they ramify, increasing the number of sites bacteria can multiply and enter nodule cells, such that a large number of cells are colonised (Gage, 2004).

Legume root nodules can be classified as either determinate or indeterminate, based on their morphology and cellular organisation: Indeterminate nodules are elongated or branched, containing one or more persistent apical meristems which give rise to new nodule cells that become infected by rhizobia and subsequently fix N2 (Udvardi & Poole, 2013, Gage, 2004). As the nodule matures, a developmental gradient is formed from the persistent meristem to the older senescent cells located adjacent to the root tissue. In contrast, determinate nodules are usually spherical, do not have

3 persistent meristems and therefore do not produce the developmental gradient found in indeterminate nodules (Gage, 2004, Udvardi & Poole, 2013).

Rhizobia are released from infection threads as infection droplets, consisting of individual bacterial cells encased in a plant-derived (i.e. originating from the infection thread) membrane. The rhizobia next differentiate into bacteroids, which is the physiological form in which they fix N2 within plant root nodule cells (Ibanez et al., 2017). Bacteroid differentiation involves a range of transcriptional changes, which include induction of N2 fixation genes (nif and fix) as well as the down-regulation of normal free- living cellular processes such as growth and division, protein synthesis and nucleic acid synthesis and repair functions (Capela et al., 2006, Karunakaran et al., 2009). One of the major signals that controls nif and fix gene expression is low O2 tension, which creates a microaerobic environment that is necessary for N2 fixation (Hwang et al., 2010, Dixon & Kahn, 2004).

The nif genes encode a suite of proteins that direct the synthesis and maturation of the nitrogenase enzyme, while the fix genes encode proteins required for bacteroid respiration and electron transfer to the nitrogenase enzyme (Poole et al., 2018). The nifHDK operon consists of three structural genes: nifH, encoding the Fe protein and nifD and nifK, encoding the Mo-Fe protein that together form the active nitrogenase enzyme complex that is responsible for the reduction of N2 to NH3 (Poole et al.,

2018, Egener et al., 2001).

The fixNOQP operon encodes a terminal oxidase with a high affinity for O2, which is critical for microaerobic respiration, mediating electron transfer and synthesis of ATP via oxidative phosphorylation at the bacteroid membrane (Black et al., 2012, Poole et al., 2018). In many species of rhizobia, the fixLJ operon and fixK sense low O2 conditions and upregulate the expression of fixNOQP genes, providing the energy required for the nifHDK operon to synthesise nitrogenase and subsequently reduce N2 to NH3 (Poole et al., 2018, Capela et al., 2006).

Knowledge of the establishment of N2-fixing symbioses has been built up from very limited agricultural symbioses, such as host specificity in Glycine max (soybean), genetic and molecular analyses in Pisum

4 sativum (pea), Glycine max and Lotus japonicus (birdsfoot trefoil) (Wang et al., 2012). There is, however, a significant amount of rhizobial diversity, both morphologically and phylogenetically, of which we know very little (Sprent et al., 2017).

1.2 Rhizobial Diversity

Rhizobia are Gram-negative soil-inhibiting bacteria, distinguished by genes required for nodulation and N2 fixation (O'Hara et al., 2016b). The term “rhizobia” refers collectively to bacteria capable of forming a symbiotic interaction with plants. Originally, all symbiotic organisms were grouped under the genus Rhizobium (Frank, 1889). Subsequently, some strains were observed to differ significantly from each other in their growth rates, leading to the division of rhizobia into fast- and slow-growing species, with the latter group comprising strains that were later recognised as a separate genus

Bradyrhizobium (Jordan, 1982). This initial division was followed subsequently by the delineation of rhizobia into Sinorhizobium (Chen et al., 1988) and then Mesorhizobium (Jarvis et al., 1997) genera.

With advances in microbiology and molecular phylogenetics, the number of rhizobia-containing genera has steadily expanded (Table 1.1). Prior to 2001, all rhizobia were classified within the α- , with most species within the Rhizobium, Mesorhizobium and Bradyrhizobium genera.

However, Moulin et al. (2001) reported Burkholderia spp. in the β-proteobacteria, to be capable of nodulating and fixing N2 on Macroptilium atropurpureum. Similarly, Chen et al. (2001) isolated another

β-proteobacterium, Ralstonia taiwanensis [later renamed Cupriavidus taiwanensis, Vandamme and

Coenye (2004)], from root nodules of Mimosa species. Therefore, there are currently 16 genera in seven families of bacteria that have been reported to form nodules on legumes, spanning both the α- and β-proteobacteria.

5

Table 1.1: Described rhizobia demonstrated to nodulate legumes and the number of species in each genus, adapted from O'Hara et al. (2016b) and LPSN (2019). Six α-Proteobacteria families containing 14 genera with 134 species of nodulating rhizobia, and one β-Proteobacteria family containing two genera with eight species of nodulating rhizobia. Family Genus Number of described species

α-Proteobacteria Bradyrhizobiaceae Bradyrhizobium 15 Brucellaceae Ochrobactrum 2 Hyphomicrobiaceae Azorhizobium 3 Devosia 1 Methylobacteriaceae Methylobacterium 1 Microvirga 3 Phyllobacterium 1 Aminobacter 1 Mesorhizobium 45 Rhizobiaceae Rhizobium 43 Allorhizobium 2 Neorhizobium 3 Sinorhizobium/Ensifer 13 Shinella 1 β-Proteobacteria Burkholderiaceae Burkholderia 6 Cupriavidus 2

1.3 Mesorhizobium

Mesorhizobium cells are Gram-negative, non-spore-forming rods, aerobic and motile usually with one polar or sub-polar flagellum (Jarvis et al., 1997). The genus Mesorhizobium was described by Jarvis et al. (1997) as primarily comprising rhizobia with a growth rate intermediate to that of the fast growing

Rhizobium and the slow growing Bradyrhizobium genera (Jarvis et al., 1997). Individual colonies of

Mesorhizobium spp. typically growing to 2 mm diameter after five to seven days on YMA, in comparison to greater than eight days for most Bradyrhizobium spp. to reach 2 mm diameter and three to five days for Rhizobium spp. (O'Hara et al., 2016b).

6

Mesorhizobium have been described as establishing N2-fixing symbioses with legumes from climactically-varied environments, including tropical, sub-tropical, temperate and arctic areas (Laranjo et al., 2014). Some Mesorhizobium species form a symbiosis with agriculturally important legumes such as Cicer arietinum (chickpea) and Biserrula pelecinus. C. arietinum is a widely grown pulse crop that is essential in many human diets, and B. pelecinus is an annual legume that provides high quality forage, can tolerate heavy grazing and is well-adapted to a wide range of pH conditions and soil types

(Laranjo et al., 2014, Lewis et al., 2005).

There are currently 50 species described in the genus Mesorhizobium (Table 1.2), 45 of these having been authenticated as rhizobia, which is where pure cultures of strains isolated from legume root nodules are inoculated back onto their host of origin, to confirm their ability to be a nodulating organism (Hungria et al., 2016). The five Mesorhizobium species that have not been authenticated are,

Mesorhizobium sanjuanii (isolated from Lotus tenuis nodules), Mesorhizobium soli and Mesorhizobium thiogangeticum (isolated from the rhizosphere of Robinia pseudoacacia and Clitoria ternatea respectively), as well as Mesorhizobium sediminum (Indian Ocean) and Mesorhizobium oceanicum

(South China Sea) that were isolated from deep sea sediments lying up to 1 km from the surface (Table

1.3).

Prior to 2004, there were only eight recognised Mesorhizobium spp.: Mesorhizobium amorphae,

Mesorhizobium chacoense, Mesorhizobium ciceri, Mesorhizobium huakuii, Mesorhizobium loti,

Mesorhizobium mediterraneum, Mesorhizobium plurifarium and Mesorhizobium tianshanense (Table

2). However, since 2004, a further 42 species have been described, with 16 of these new species isolated from plant species native to the poor, sandy soils of northern regions of China characterised, including Alhagi sparsifolia (wild shrub), Astragalus adsurgens (perennial shrub harvested for silage),

Caragana microphylla (perennial shrub used in revegetation), Lotus frondosus (wild herbaceous species), Oxytropis glabra (wild herbaceous species), Albizia kalkora (subtropical tree), the traditional green manure species Astragalus sinicus (annual) and the Australian blackwood Acacia melanoxylon

7

R. Br., which is an introduced tree species resistant to cold and drought, and used for erosion control, landscaping and timber (Zhu et al., 2015). In addition, seven Mesorhizobium sp. isolated from Sophora sp. by De Meyer et al. (2015, 2016) came from the same collection, where 48 strains were isolated in natural ecosystems of the South Island of New Zealand from root nodules of the shrub S. prostrata, the small upright tree S. longicarinata (up to 2 m height) and the large tree S. microphylla (up to 25 m height). The increasing rate of species discovery from these studies highlights the possibility of many yet undiscovered species that may have the potential to be used in future agricultural systems.

The host range of Mesorhizobium sp. strains can vary widely, where host range is defined as the ability of the rhizobia to enter legume roots and form nodules (Perret et al., 2000). Some strains can have a very narrow host range where they are only able to form nodules on one or very few legume species, such as M. ciceriT UPM-Ca7 which is only known to nodulate Cicer arietinum (chickpea) (Nour et al.,

1994). In contrast, other strains can exhibit very broad host ranges, being able to nodulate legumes from different genera such as M. tianshanenseT A-1BS which nodulates Caragana polourensis, Glycine max, Glycyrrhiza pallidijlora, Glycyrrhiza sp., Glycyrrhiza uralensis, Halimodendron holodendron,

Sophora alopecuroides and Swainsonia salsula (Chen et al., 1995) or M. plurifariumT LMG1032 which nodulates Acacia nilotica, Acacia senegal, Acacia seyal, Acacia tortilis subsp. raddiana, Leucaena leucocephala and Neptunia oleracea (de Lajudie et al., 1998).

A particularly well-documented case of host range differences in Mesorhizobium sp. has been reported between M. ciceri bv. biserrulae strains WSM1271, WSM1497 and WSM1284, isolated from the root nodules of B. pelecinus from different sites in the Mediterranean basin (Nandasena et al.,

2004). All three strains are highly effective N2-fixing microsymbionts of B. pelecinus (Nandasena et al.,

2004). Strains WSM1271 and WSM1497 have a relatively narrow host range, with WSM1271 nodulating Acacia membranaceus and B. pelecinus, and WSM1497 nodulating Acacia adsurgens, A. membranaceus and B. pelecinus (Nandasena et al., 2004). In contrast, WSM1284 nodulates a much wider group of host legumes consisting of A. adsurgens, A. membranaceus, B. pelecinus, Dorycnium

8 rectum, Dorycnium hirsutum, Glycyrrhiza uralensis, L. leucocephala, Lotus edulis, Lotus glaber, Lotus maroccanus, Lotus ornithopodioides, Lotus pedunculatus, Lotus peregrinus, Lotus subbiflorus and

Ornithopus sativus (Nandasena et al., 2004). Importantly, strains may not always be effective N2-fixing microsymbionts on hosts with which they nodulate, with WSM1284 forming white nodules (indicative of nodules that cannot fix N2) on D. hirsutum, even though this strain is highly effective with many genotypes of B. pelecinus (Nandasena et al., 2004). Therefore, strains that can nodulate one legume species may show very different nodulation phenotypes on other legume species.

Table 1.2: Mesorhizobium species that have been authenticated on host of isolation and are valid species, as defined by LPSN (2019). 45 Mesorhizobium species identified by Type strain, 16S rRNA accession number, effective reference and host of isolation (LPSN, 2019). The additional reference of Jarvis et al. (1997) is included for species re-classified as Mesorhizobium in 1997.

Species Type Strain 16S rRNA Reference Host of isolation Accession number

Mesorhizobium abyssinicae AC98c GQ847896 Degefu et al. (2013) Acacia abyssinica/A. tortilis Mesorhizobium acacia RITF741 NR_137366 Zhu et al. (2015) Acacia melanoxylon R. Br. Mesorhizobium albiziae CCBAU 61158 DQ100066 Wang et al. (2007) Albizia kalkora

Mesorhizobium alhagi CCNWXJ12-2 EU169578 Chen et al. (2010) Alhagi sparsifolia

Mesorhizobium amorphae ACCC 19665 AF041442 Wang et al. (1999) Amorpha fruticosa

Mesorhizobium australicum WSM2073 AY601516 Nandasena et al. (2009) Biserrula pelecinus L.

Mesorhizobium calcicola ICMP 19560 KC237406 De Meyer et al. (2016) Sophora longicarinata

Mesorhizobium camelthorni CCNWXJ 40-4 EU169581 Chen et al. (2011) Alhagi sparsifolia

Mesorhizobium cantuariense ICMP 19515 KC237397 De Meyer et al. (2015) Sophora microphylla

Mesorhizobium caraganae CCBAU 11299 EF149003 Guan et al. (2008) Caragana microphylla

Mesorhizobium chacoense CECT 5336 AJ278249 Velazquez et al. (2001) Prosopis alba

Mesorhizobium ciceri UPM-Ca7 U07934 Nour et al. (1994), Jarvis et al. Cicer arietinum L. (1997) Mesorhizobium delmotii STM4623 KP242314 Mohamad et al. (2017) Anthyllis vulneraria

Mesorhizobium erdmanii USDA 3471 KM192334 Martinez-Hidalgo et al. (2015) Lotus corniculatus

Mesorhizobium gobiense CCBAU 83330 EF035064 Han et al. (2008) Oxytropis glabra

Mesorhizobium hawassense AC99b GQ847899 Degefu et al. (2013) Sesbania sesban

Mesorhizobium CSLC115N KT899885 Marcos-Garcia et al. (2017) Lotus corniculatus helmanticense Mesorhizobium huakuii CCBAU 2609 D13431 Chen et al. (1991), Jarvis et al. Astragalus sinicus (1997)

9

Mesorhizobium japonicum MAFF 303099 NC_002678 Martinez-Hidalgo et al. (2016) Lotus japonicus (genome) Mesorhizobium jarvisii ATCC 33669 KM192335 Martinez-Hidalgo et al. (2015) Lotus corniculatus

Mesorhizobium kowhai ICMP 19512 KC237394 De Meyer et al. (2016) Sophora microphylla

Mesorhizobium loti LMG 6125 X67229 Jarvis et al. (1982), Jarvis et al. Lotus corniculatus (1997) Mesorhizobium LMG 17148 AM181745 Nour et al. (1995), Jarvis et al. Cicer arietinum mediterraneum (1997) Mesorhizobium STM 2683 AM930381 Vidal et al. (2009) Anthyllis vulneraria metallidurans Mesorhizobium muleiense CCBAU 83963 HQ316710 Zhang et al. (2012) Cicer arietinum

Mesorhizobium ICMP 19545 KC237410 De Meyer et al. (2016) Sophora prostrata newzealandense Mesorhizobium olivaresii CPS13 FM203302 Lorite et al. (2016) Lotus corniculatus

Mesorhizobium WSM2075 AY601515 Nandasena et al. (2009) Biserrula pelecinus L. opportunistum Mesorhizobium plurifarium LMG 11892 Y14158 de Lajudie et al. (1998) Acacia senegal

Mesorhizobium STM4891 KP242313 Mohamad et al. (2017) Anthyllis vulneraria prunaredense Mesorhizobium qingshengii CCBAU 33460 JQ339788 Zheng et al. (2013) Astragalus sinicus

Mesorhizobium robiniae CCNWYC 115 EU849582 Zhou et al. (2010) Robinia pseudoacacia

Mesorhizobium sangaii SCAU7 EU514525 Zhou et al. (2013) Astragalus luteolus

Mesorhizobium SDW014 AF508207 Gao et al. (2004) Astragalus adsurgens septentrionale Mesorhizobium shangrilense CCBAU 65327 EU074203 Lu et al. (2009) Caragana bicolor

Mesorhizobium shonense AC39a GQ847890 Degefu et al. (2013) Acacia abyssinica

Mesorhizobium CCBAU 01550 EU399698 Zhao et al. (2012) Astragalus silamurunense membranaceus Mesorhizobium sophorae ICMP 19535 KC237424 De Meyer et al. (2016) Sophora microphylla

Mesorhizobium Ala-3 AM491621 Ramirez-Bahena et al. (2012) Anagyris latifolia tamadayense Mesorhizobium tarimense CCBAU 83306 EF035058 Han et al. (2008) Lotus frondosus

Mesorhizobium temperatum SDW018 AF508208 Gao et al. (2004) Astragalus adsurgens

Mesorhizobium A-1BS AF041447 Chen et al. (1995), Jarvis et al. Glycyrrhiza pallidiflora tianshanense (1997) Mesorhizobium waimense ICMP 19557 KC237387 De Meyer et al. (2015) Sophora longicarinata

Mesorhizobium waitakense ICMP 19523 KC237413 De Meyer et al. (2016) Sophora microphylla

Mesorhizobium wenxiniae WYCCWR 10195 KX132877 Zhang et al. (2018) Cicer arietinum L.

10

Table 1.3: Mesorhizobium species which have not been authenticated by nodulation of host of isolation and are valid species, as defined by LPSN (2019). Five species identified by Type strain, 16S rRNA accession number, effective reference and original host or location of isolation (LPSN, 2019).

Species Type Accession Reference Host/location of isolation Strain number 16S Mesorhizobium oceanicum B7 KT157593 Fu et al. (2017) Isolated from deep-sea water (1km), South China Sea Mesorhizobium sanjuanii BSA136 MF979853 Sannazzaro et al. (2018) Lotus tenuis Mesorhizobium sediminum YIM M12096 KX151664 Yuan et al. (2016) Isolated from deep-sea sediment, Indian Ocean

Mesorhizobium soli NHI-8 KC484966 Nguyen et al. (2015) Rhizosphere of Robinia pseudoacacia

Mesorhizobium thiogangeticum SJT AJ864462 Ghosh and Roy (2006) Rhizosphere soil of Clitoria ternatea

1.4 Mesorhizobium spp. as inoculants in agricultural systems of Australia and

New Zealand

The development of agriculture in Australia and New Zealand has resulted in the introduction of many new, non-native plant species into these environments. In cases where the introduced species of plants were legumes, these plants have usually required the co-introduction of compatible rhizobial inoculants (MacKinnon et al., 1977, Howieson & Ballard, 2004). Inoculation is the practice of adding rhizobia to legume seeds or soil in the vicinity of the seed during sowing (Drew et al., 2014). This addition of rhizobia enables root nodulation to occur, leading to an increase in symbiotic N2 fixation, crop biomass and quality, and the amount of nitrogen returned to the soil when the plant dies (Drew et al., 2014). In several instances, highly effective strains of Mesorhizobium sp. have been introduced into these environments along with the new host legume, to ensure optimal rates of N2 fixation in the field. However, an unintended consequence of the introduction of these strains is that the mobile nature of the symbiosis genes in Mesorhizobium spp. has influenced the diversity of soil organisms nodulating the target legume.

The first reports of the impact of using Mesorhizobium spp. as inoculants came from Ronson and

Sullivan, who were studying the genetic diversity of Lotus corniculatus-nodulating rhizobia in New

11

Zealand (Sullivan et al., 1995). Approximately seven years after inoculation of L. corniculatus with M. loti strain ICMP3153 in an area devoid of compatible L. corniculatus-nodulating organisms, partial 16S rRNA sequencing of diverse isolates confirmed they were not derived from ICMP3153 but from the background soil bacteria (Sullivan et al., 1995). Further analysis revealed that the strains had acquired an ~500 kb region of DNA, harbouring nod, nif and fix symbiosis genes, from the original inoculant strain and that this region had integrated in the recipient genomes at the phe-tRNA gene (Sullivan &

Ronson, 1998). This indicated that that this region had horizontally transferred from the inoculant strain to non-symbiotic rhizobia in the environment (Sullivan & Ronson, 1998). This transferable region of DNA was named a ‘symbiosis island’ on the basis of its similarity to pathogenicity islands, having the ability to transfer to normally non-symbiotic bacteria and enabling them to form a symbiosis

(Sullivan & Ronson, 1998). These symbiosis islands were later classified as monopartite Integrative and

Conjugative Elements (ICEs) (Ramsay et al., 2006). ICEs are defined as elements that excise from their host chromosome in a site‐specific manner, forming a circularized element that is generally transient

(Ramsay et al., 2006). Conjugative transfer is initiated when the ICE is in the circularized form, resulting in a new copy of the ICE integrating into the recipient chromosome (Ramsay et al., 2006).

Further reports of the impact of introducing Mesorhizobium inoculants next came from Australia, where the pasture legume B. pelecinus, originating from the Mediterranean basin, was first introduced into Western Australian agriculture in 1993, along with the inoculant strain M. ciceri bv. biserrulae

WSM1271 (Nandasena et al., 2006). Six years after introduction of the legume at a field site in

Northam (Western Australia), nodules of B. pelecinus were sampled and a range of isolates that were genetically different to the inoculant strain were obtained (Nandasena et al., 2006). Sanger sequencing of a subset of these isolates showed 100% identity of nodA and nifH genes to those in

WSM1271, while the strains harboured highly dissimilar 16S rRNA sequences. Subsequent whole genome sequencing of two isolates (WSM2073 and WSM2075) showed they harboured symbiosis ICEs that were identical to that of the original WSM1271 inoculant strain (Haskett et al., 2016).

Furthermore, all three strains appeared to harbour two other identical regions which were shown to

12 form part of a novel tripartite ICE, which is capable of transfer to recipient cells in a way analogous to that of the monopartite R7A ICE (Haskett et al., 2016). Importantly, Nandasena et al. (2007) and

Haskett et al. (2016) showed that the novel isolates receiving the symbiosis ICE were far less effective at fixing N2 on B. pelecinus than the commercial inoculant strain WSM1271.

Introduction of Mesorhizobium sp. to agriculture in Australia and New Zealand therefore has the potential to lead to the evolution of novel rhizobia that can nodulate the target legume but fix N2 poorly. These poorly effective strains may then compete with the inoculant strain, lowering the amount of N2 fixed and therefore reducing the benefits of legume inoculation in agricultural systems.

It is therefore imperative that mobility of symbiosis ICEs be considered when evaluating

Mesorhizobium inoculant strains for any new legume species into agriculture.

1.5. Scorpiurus muricatus: A potential new annual pasture legume for southern Australia

Over the last three decades, growers have focussed on cropping grains in the southern Australian agricultural areas that receive medium-to-low rainfall (200 – 600 mm). With continuous cropping however, in many cases, it has negatively impacted farming systems causing low soil nitrogen fertility and carbon content, increased plant pathogens, weed herbicide resistance and high fertiliser expense

(Howieson et al., 2000c, Nichols et al., 2012). Consequently, in medium-to-low rainfall areas of southern Australia there is a lack of suitable annual pasture legumes, which is limiting agricultural productivity and profitability (Howieson et al., 2000c).

Since the mid-1990’s extensive research and development in Australia has resulted in the successful domestication of a number of alternative annual pasture legumes from the Mediterranean basin, providing a significant impact on agricultural systems (Loi et al., 2005, Nichols et al., 2007). These

“second generation” of annual legumes for Ley farming possessed traits such as hard seed coats, aerial seeding for easy harvest, deep rooting and early maturation, allowing them to self-regenerate (Loi et al., 2005, Nutt, 2012). This research was further supported by the development of highly effective

13 rhizobial inoculant strains that survived in the targeted soil types that were naturally devoid of rhizobia capable of nodulating these new legumes (Howieson et al., 2000a, Howieson et al., 2000b, Howieson

& Ballard, 2004, Yates et al., 2005).

Scorpiurus muricatus is an annual legume found throughout the Mediterranean basin and

Macaronesia recognised by its radial growth and distinctive ‘Scorpion’s tail’ seed pods (Figure 1.1A and B) (Abbate et al., 2010, Heyn & Raviv, 1966, Beale et al., 1991). It has the potential to be a new

“next generation” alternative pasture legume for the medium-to-low rainfall areas of southern

Australia, with traits including: extreme hard-seededness, deep rooted and has a high adaptability to a wide range of semi-arid environmental conditions, including: soil type, altitude, soil nutrient status, climate and pH (Beale et al., 1991, Ehrman & Cocks, 1990, Atallah et al., 2008). S. muricatus is highly preferred by ruminant livestock (Di Giorgio et al., 2009), has a high crude protein (CP) concentration

(Licitra et al., 1997), a positive effect on the yield of subsequent grain crops (Pülschen, 1992) and good response to repeated defoliations after slow initial plant growth (Ruisi et al., 2017). For all these reasons, this legume is currently being evaluated as a potential new pasture legume species for medium-to-low rainfall areas of southern Australia. However, to ensure the success of the future adoption of this species, elite inoculant rhizobia capable of efficiently fixing N2 with S. muricatus need to be identified (Howieson et al., 2000c).

S. muricatus sampled from various areas in the Mediterranean, have been observed to form elongated-indeterminate or bifurcate nodules (Bouchiba et al., 2017, Atallah et al., 2008, Muresu et al., 2008, Safronova et al., 2004). Few reports exist describing the rhizobial microsymbionts of S. muricatus. Both Muresu et al. (2008) and Bouchiba et al. (2017) report sampling S. muricatus from

Sardinia and Algeria, respectively. While Muresu et al. (2008) failed to successfully isolate any strains from nodules taken from this host, they were able to identify nodule bacteria as belonging to the genus Mesorhizobium by 16S rRNA PCR directly from nodules. In contrast, Bouchiba et al. (2017) did report isolation of 51 strains from S. muricatus however, partial 16S rRNA sequences of 18 strains

14 showed the possible root nodule forming isolates to be most closely related to Rhizobium spp. (15 isolates) and Phyllobacterium spp. (3 isolates), with no isolates matching closely with Mesorhizobium spp.. Crucially though, none of their isolates could be authenticated as rhizobia as they did not nodulate S. muricatus when inoculated back on this host in controlled glasshouse experiments.

Figure 1.1: Photos of Scorpiurus muricatus 95GCN115, Murdoch University, Murdoch, Western Australia. (A) Mature plants with radial branch growth (B) Distinctive ‘Scorpion tail’ seed pods.

Safronova et al. (2004) have reported the recovery of seven isolates from nodules of S. muricatus located on Asinara Island (North Western Sardinia). The growth rate of these isolates was heterogeneous, with both intermediate- (defined as growing within 4-5 days) and slow-growing strains (6-7 days) on YMA at 28°C. 16S rDNA PCR-RFLP analysis grouped the isolates from the sampled legume species with M. loti and M. mediterraneum type strains, consistent with this legume being nodulated by Mesorhizobium spp. In support of this, the isolates were also confirmed to nodulate and fix N2 with L. cytisoides, L. ornithopodioides and O. compressus, and nodulate L. edulis, which are all legumes known to form symbioses with Mesohizobium spp. Therefore, these studies suggest that S. muricatus is possibly nodulated by rhizobia in the genus Mesorhizobium.

15

1.6 Aims

Scorpiurus muricatus is a legume found in the Mediterranean and Macronesia that is currently being evaluated as a new pasture for medium-to-low rainfall areas of southern Australia. Successful introduction of S. muricatus to these agricultural systems will require the co-introduction of a highly effective inoculant strain. There is little information on the microsymbionts that associate with S. muricatus in the literature. However, the Centre of Rhizobium Studies (CRS) has within its collection of rhizobia a range of potential S. muricatus microsymbionts. The key aims of this study are therefore to:

1. Characterise the free-living phenotypes of these organisms.

2. Determine the phylogenetic relationship between these strains and other well-characterised

rhizobia.

3. Evaluate their N2 fixation efficiency in symbiosis with S. muricatus.

4. Determine the ability of S. muricatus-nodulating organisms to form a symbiotic interaction

with current Australian commercial legumes Cicer arietinum, Biserrula pelecinus and Lotus

corniculatus.

16

2.0 METHODS

2.1 Bacterial strains and plant accessions

The 19 strains of bacteria investigated in this thesis were sourced either from the Western Australian

Soil Microbiology (WSM) culture collection (WSM1184, WSM1343 and WSM1386, isolated from nodules of Scorpiurus sulcatus sampled, in Syria, Morocco and Australia respectively) or from S. muricatus nodules collected from soil trapping experiments with soils from Morocco (strains J, N and

T) and Israel (strains A, B, C, D, F, G, H, I, K, L, M, O and S). References strains Mesorhizobium ciceri bv. ciceri CC1192 (commerical inoculant for Cicer arietinum) and M. ciceri bv. biserrulae WSM1497

(commercial inoculant for Biserrula pelecinus) were also sourced from the WSM collection. The

Moroccan and Israeli soils were collected in 2004 and 2006 and have been stored at 4˚C since that time. Seeds of S. muricatus ID: 95GCN115, B. pelecinus cv casbah, Lotus corniculatus cv goldie and C. arietinum cv striker, were acquired from B. Nutt (Murdoch University).

2.2 Growth media and culturing conditions

Bacteria were routinely cultured on Yeast Mannitol Agar (YMA) (Vincent, 1970) with 1.5% w/v agar and incubated at 28°C. For the assessment of growth on different media, individual colonies of each strain were dilution-streaked (Hungria et al., 2016) from YMA plates onto two plates each of ½ LA

(Howieson et al., 1988), YMA and TY (Beringer, 1974) with 1.5% w/v agar and incubated at 28°C for up to 36 days. The number of days required to yield single colonies of 2 mm diameter were recorded and an average of the two plates calculated. If strains failed to reach 2 mm diameter, the colony size was recorded at 36 days. For the assessment of temperature tolerance, individual colonies of each strain were dilution-streaked onto two plates of YMA at 10°C, 16°C, 22°C, 28°C, 32.5°C and 37.5°C for up to

30 days and growth assessed as being either NIL (no growth), poor (1-4 visible single colonies on initial spread and first streak only), limited (≥ 5 visible single colonies on initial spread and 1st streak), growth

17

(≥ 5 visible single colonies on second streak) and good (≥ 5 visible single colonies on third streak), when single colonies visible (≥ 0.3mm diameter) and average calculated.

Table 2.1: List of strains included in this study with country of origin, isolation methods, soil batch number and host of isolation. N/A, not applicable.

STRAIN Country of Origin Isolation method Soil batch Source Host Number WSM1184 ICARDA Nodule N/A S. sulcatus Syria research centre (1989) WSM1343 Oulmes, Morocco (1993) Nodule N/A S. sulcatus

WSM1386 Manjimup research station, Western Nodule N/A S. sulcatus Australia (1994) A Israel Soil trapping 415-419 S. muricatus B C Israel Soil trapping 425-429 S. muricatus D F Israel Soil trapping 435-439 S. muricatus M G Israel Soil trapping 440-444 S. muricatus

H Israel Soil trapping 445-449 S. muricatus S J Morocco Soil trapping 255-258 S. muricatus

I Israel Soil trapping 410-414 S. muricatus K L Israel Soil trapping 420-424 S. muricatus O N Morocco Soil trapping 271-273 S. muricatus T

2.3 Amplified fragment length polymorphism (AFLP) assessment of diversity by RPO1-based PCR

The diversity of the 19 test strains (Table 2.1) plus the two reference strains, CC1192 and WSM1497, was assessed by Amplified Fragment Length Polymorphism (AFLP) with the primer RPO1 (5’-

AATTTTCAAGCGTCGTGCCA-3’) designed by Richardson et al. (1995). Each reaction mixture for PCR contained 1 µL of cell pellet (OD600 = 6), 50 pmol RPO1 primer, 10 µL of GoTaq Green MasterMix

(Promega, Cat. No. 7122) and PCR grade water (Fisher Biotec Australia, Cat. No. UPW-100) to a final volume of 20 µL. The blank consisted of the same constituents but with no added DNA template. PCR cycling conditions were 95°C for 5 min followed by 5 cycles at 95°C for 30 s, 55°C for 10 s, 72°C for 90

18 s and then 30 cycles at 95°C for 30 s, 55°C for 25 s, 72°C for 90 s and a final extension at 72°C for 5 min.

The amplified DNA fragments were analysed by agarose gel electrophoresis, using BIO-RAD PowerPac

1000 (Serial No. 287BR01634) power system and BIO-RAD Sub-cell GT BASIC (Serial No. 61S11350) and

BIO-RAD Sub-cell GT WIDE MINI (Serial No. 711BR08422). The 2% (w/v) agarose gels were buffered with 1 x TAE buffer and a 1 kb molecular weight marker (Promega, Ref no. G5711) was used to estimate product size. Gels were routinely electrophoresed for for 90 min at 70 V.

Agarose gels were post-stained in a solution of 1 x TAE containing 0.5 µg/mL of ethidium bromide

(EtBr), the gel was stained in EtBr for 30 min and de-stained in 1 x TAE buffer for 20 min. Gels were visualised under UV light using the BIO-RAD Molecular Imager XR+ (Serial No. 721BR04626) with Image

Lab software (v. 3.0, build 11). Banding patterns of each strain and characterised strains CC1192 and

WSM1497 were visually compared to one another to determine the diversity within the test group.

2.4 Phylogeny based on 16S rRNA gene

The 16S rRNA gene region of the test strains was amplified using 16S primers forward (pA) 27F (5′-

AGAGTTTGATCCTGGCTCAG-3’) and reverse (pH) 1522R (5′-AAGGAGGTGATCCAGCCGCA-3’) (Edwards et al., 1989). Each PCR reaction mixture contained 1 µL concentrated cell pellet OD600 = 1, 6.25 pmol of pA primer, 6.25 pmol of pH primer, 12.5 µL of GoTaq Green MasterMix (Promega, Cat. No. 7122) made up to a final volume of 25 µl with PCR grade water. The blank consisted of the same constituents but with no added DNA template. PCR cycling conditions were 95°C for 5 min followed by 3 cycles at

95°C for 1 min, 59.4°C for 2 min 15 s, 72°C for 1 min and then 30 cycles at 95°C for 30 s, 59.4°C for 75 s, 72°C for 75 s and a final extension at 72°C for 7 min. The amplified DNA fragments were analysed by agarose gel electrophoresis, stained and visualised as previously described (Section 2.2), with the following modifications: A 5 µL aliquot of the PCR reaction was loaded onto a 1% (w/v) agarose gel which was electrophoresed 60 min at 100 V.

19

The remaining 20 µL of PCR reaction was purified using FavorPrep GEL/PCR Purification Mini Kit (Cat.

No. FAGCK001) following the manufacturer’s instructions. DNA quality and quantity of purified PCR products was assessed by NanodropOne (Thermo Scientific). Purified PCR products were submitted to

Australian Genomics Research Facility (AGRF, Perth, Australia) for Sanger sequencing using forward primers (pA) 27F (5′-AGAGTTTGATCCTGGCTCAG-3’), (Gamma*) 358F (5′-CTCCTACGGGAGGCAGCAGT-

3’) (Vaneechoutte et al., 2000) and (O*) 926F (5′-AACTCAAAGGAATTGACGG-3’) (Lane, 1991), and reverse primers: (pH) 1522R (5′-AAGGAGGTGATCCAGCCGCA-3’), (3) 1093R (5′-

GTTGCGCTCGTTGCGGGACT-3’) and (BKL1) 516R (5′-GTATTACCGCGGCTGCTGGCA-3’). Consensus sequences for partial 16S rRNA sequences of at least 1,388 bp were assembled with Geneious 11.1.5 software. DNA bases were edited when confirmed by overlapping sequences when a minimum of two sequences matched for each base pair. BLASTN comparison of the consensus sequence for each strain was performed on NCBI (NCBI, 2019).

A phylogenetic tree for the partial (1,388 bp) 16S rRNA sequences comprising the 19 test strains compared with selected characterised strains (Table 2.2), was constructed in Geneious 11.1.5, using the Tamura-Nei, neighbour-joining method of 10000 bootstrap replicates with a support threshold of

50%. Azorhizobium caulinodansT ORS 571 was included as the outgroup.

2.5 Phylogeny based on nifH and nodC symbiosis genes

The nifH region of the test strains (excluding WSM1184) was amplified using degenerate primers nif1F

(5’-TAYGGNAARGGNGGNATYGGNAARTC-3’) (Boulygina et al., 2002) and nif439R (5’-

GGCATNGCRAANCCDCCRCA-3’) (De Meyer et al., 2011). Each PCR reaction mixture contained 1 µL concentrated cell pellet (OD600 = 3), 12.5 pmol of nif1F primer, 12.5 pmol of nif439R primer, 12.5 µL of

GoTaq Green MasterMix and made up to a final volume of 25 µl with PCR grade water. The blank consisted of the same constituents but with no added DNA template. PCR cycling conditions were 95°C for 5 min followed by 3 cycles at 95°C for 1 min, 53°C for 2 min, 72°C for 1 min and then 30 cycles at

95°C for 30 s, 53°C for 1 min, 72°C for 1 min and a final extension at 72°C for 7 min.

20

Table 2.2: List of characterised strains used in this study with 16S, nifH and nodC accession numbers. Alternative accession numbers representing (G) genome and (C) contig, provided when used. N/A, not applicable.

STRAIN 16S accession nifH accession nodC accession

Azorhizobium caulinodansT ORS 571 D11342 AP009384 (G) AP009384 (G)

M. australicumT WSM2073 AY601516 AY601522 NC_019973 (G)

M. ciceri CC1192 NZ_CP015062 CP015062 (G) NZ_CP015062

M. ciceriT UPM-Ca7 U07934 DQ450928 DQ407292

M. ciceri WSM1271 AY601513 AY601521 NC_015675 (G)

M. ciceri WSM1497 CP021070 (G) CP021070 (G) CP021070 (G)

M. delmotiiT BQ8482 FUIG01000092 (C) FUIG01000049 (C) FUIG01000058 (C)

M. erdmanii Opo-242 MZXX01000021 (C) MZXX01000018 (C) MZXX01000003 (C)

M. erdmaniiT USDA 3471 KM192334 N/A KM192344

M. helmanticenseT CSLC115N KT899885 PZJX01000070 (C) PZJX01000067 (C)

M. japonicumT MAFF303099 BA000012 (G) BA000012 (G) BA000012 (G)

M. lotiT DSM 2626 QGGH01000001 (C) QGGH01000015 (C) QGGH01000021 (C)

M. muleienseT CGMCC 1.11022 FNEE01000049 (C) FNEE01000025 (C) FNEE01000043 (C)

M. opportunistumT WSM2075 AY601515 AY601524 CP002279 (G)

M. prunaredenseT STM4891 KP242313 FTPD01000017 (C) KU984252

M. tamadayenseT Ala-3 AM491621 LN824202 AM491624

M. tianshanenseT A-1BS AF041447 N/A N/A

M. tianshanenseT CCBAU 3306 N/A GQ167282 KP251785

Mesorhizobium sp. AA22 LYTO01000050 (C) LYTO01000326 (C) LYTO01000324 (C)

Mesorhizobium sp. AA23 PRJNA323413 (G) LYTP01000007 (C) LYTP01000015 (C)

21

The amplified DNA fragments were separated and visualised by agarose gel electrophoresis and the quality and quantity of purified PCR products analysed as previously described (Section 2.3). Purified products were sequenced as detailed in earlier (Section 2.4), using the amplification primers nif1F and nif439R. Consensus partial nifH sequences of at least 257 bp were assembled and a phylogenetic tree constructed with Geneious (Section 2.4). Strains from the 16S rRNA tree were included in the nifH tree, except in cases where there was no nifH sequence available. In these instances, these strains were substituted for related strains for which nifH sequences were available such that M. erdmaniiT

USDA 3471, was replaced with M. erdmanii Opo-242, M. tianshanenseT A-1BS replaced with M. tianshanenseT CCBAU 3306, M. ciceriT UPM-Ca7 sequence length was only 85 bp but was left in the tree as it matched closely with M. ciceri CC1192.

The nodC region of the test strains (excluding WSM1184) was amplified using degenerate primers nodC540F (5’-TGATYGAYATGGARTAYTGGCT-3’) and reverse: nodC1164R (5’-GAYARCCARTCGCTRTTG-

3’) (Laguerre et al., 2001). PCR reaction and cycling conditions and gel electrophoresis and visualisation were as per nifH, with the exceptions that 6.25 pmol was added for each primer, with an annealing temperature of 47°C for 2 min 15 s for 3 cycles and an annealing time of 75 s for the 30 cycle of amplifications. Sequencing was completed as previously described for nifH using the amplification primers nodC540F and nodC1164R.

A phylogenetic tree for partial (496 bp) nodC sequence was built as described for nifH, with both M. erdmaniiT USDA 3471 and M. erdmanii Opo-242 included. Only strain I had a slightly shorter sequence than other tests strains of 429 bp.

2.6 Assessment of symbiotic phenotype of strains on Scorpiurus muricatus

Equal proportions of yellow sand and coarse river sand (1:1) were combined, with 5 g/L Fe2(SO4)3 at a rate of 2 L per cement-mixer of sand to reduce the pH to 6.59. Combined and pH-adjusted sand was added to 1 L pots lined with paper towel and steamed for 3 h, allowed to cool and then flushed twice with hot, boiled deionised (DI) water (Yates et al., 2016).

22

S. muricatus seeds were mechanically scarified with grade 50 sandpaper and surface sterilised in 70%

(v/v) ethanol for 1 min, 4% (v/v) NaOCl for 3 min, followed by six washes in sterile DI water. Following surface sterilisation, the seed coat of each individual seed was aseptically cut with a scalpel.

Approximately 40 seeds were then spread on sterile 0.9 % (w/v) water agar plates with 4 mL sterile DI water, covered with aluminium foil and allowed to imbibe at room temperature for 24 h, after which excess DI water was removed with a sterile pipette, aluminium foil replaced, and plates were inverted for a further 16 h. The germination rate of seeds was greater than 99%.

Due to space limitations in the glasshouse, the 19 strains were split into two groups, and two sequential effectiveness experiments were conducted, with experiment 1 commencing on 27 March

2019 and harvested 69 days later, and experiment 2 commencing on 25 April 2019 and harvested 76 days later.

Each inoculated treatment consisted of eight pots, except for WSM1386 in experiment 2 which only consisted of four pots. Uninoculated N-starved and N-Fed controls were included in each experiment, with 16 pots per treatment, except for the N-starved control in experiment 2 which consisted of 8 pots. In all treatments and both experiments, each pot contained three seedlings at sowing. To prepare the inoculant strains, single colonies of each inoculant strain were streaked in triplicate on ½

LA media and the plates incubated at 28°C for up to 14 days. Plates were individually resuspended into a total volume of 40 ml of sterile 1% (w/v) sucrose solution (Yates et al., 2016) and a 1 mL aliquot of this suspension was inoculated onto each seedling at sowing, with 1 mL of sterile 1% (w/v) sucrose solution added to each of the uninoculated N-fed and N-starved control plants. To assess inoculant strain viable cell numbers, a 1 ml aliquot of each inoculant suspension was serially diluted eight-fold in sterile 1% (w/v) sucrose and a 100 µL aliquot of 10-5, 10-6, 10-7 and 10-8 dilutions were spread over separate ½ LA plates and incubated at 28°C. Colonies were enumerated on plates, with those yielding counts between 30 – 300 colony forming units (CFU) used to calculate the number of inoculant cells/mL (O'Hara et al., 2016a). Sterile Polyvinyl chloride tubes (2.5 cm diameter and 25 cm length)

23 were inserted into the sand for nutrient and water supply, sterile lids were used to close tubes. After planting and inoculation, sterile alkathene beads were spread on the surface of the soil to prevent air- borne infection and cross-contamination (Yates et al., 2016).

Plants were watered at least twice-weekly with sterile DI water as required and received 20 ml of CRS

Plant nutrient solution (Yates et al., 2016) once a week. Two mL of sterile KNO3 (10 g/L) was applied to N-fed control plants in each of the first two weeks, and four mL each following week. Nine days after inoculation, plants were thinned to one plant per pot and pots randomised.

At harvest, plants were carefully removed from the soil, roots washed, and nodules counted and described. The shoots and roots were separated, and oven dried at 60°C for 72 hours prior to weighing.

One plant from each treatment had between two and six nodules removed for bacterial isolation to assess nodule occupancy. Isolation of nodule bacteria was carried out as described by Hungria et al.

(2016), where nodules were first sterilised by soaking in 70% (v/v) ethanol for 40 s, then 4% NaOCl

(v/v) for 4 min, followed by washing six times in sterile DI water prior to being crushed and dilution streaked onto ½ LA plates and incubated at 28°C. Single colonies were restreaked onto ½ LA plates and then single colonies inoculated into ½ LA broth and incubated on a gyratory shaker at 250 rpm until cultures reached OD600 ~ 1.0, after which cells were harvested by repeated centrifugation (14,000 x g, 2 min) to concentrate cell pellets to a final OD600 = 6. Strain identification was carried-out with

AFLP PCR with the RPO1 primer as previously described (Section 2.3), with the profile of isolate strains compared to that of the original strain profile.

Means and standard errors were calculated for all data. One-way ANOVA was calculated in RStudio

(Version 1.2.1335) to test if treatment groups were statistically equivalent to one another (P ≤ 0.05).

WELCH correction for heterogeneity of variance was used for mean shoot dry weight/plant and mean nodules/plant.

24

2.7 Assessment of nodulation and nitrogen fixation on other legume species

Six representative strains (WSM1343, WSM1386, A, D, J and T) were chosen to test the ability of S. muricatus-nodulating strains to nodulate and fix N2 on Biserrula pelecinus, Cicer arietinum and Lotus corniculatus, with S. muricatus included as a control. Glasshouse experiments were conducted as per section 2.6, with several exceptions. Each treatment for each plant species consisted of four pots. C. arietinum seeds were not scarified, B. pelecinus seeds were additionally scarified prior to sterilisation by hand using grade 50 sandpaper in place of a scalpel, and L. corniculatus seeds were not manually scarified. Seeds were germinated on 0.9% (w/v) water agar plates, 40 h for B. pelecinus, C. arietinum and S. muricatus and 64 h for L. corniculatus. Inoculants were prepared on YMA and enumerated as per section 2.5, except that viable cell counts were performed by the Miles and Misra method, where

25 µl aliquots of the 10-1, 10-2, 10-3, 10-4 10-5, 10-6, 10-7 and 10-8 dilutions were pipetted onto YMA agar plates and incubated at 28°C, for up to 10 days with those yielding counts between 5 – 50 CFU used to calculate the number of inoculants cells/mL (O'Hara et al., 2016a).

Initially, three seedlings were planted into pots under aseptic conditions, with C. arietinum and S. muricatus thinned to one plant per pot after 11 days, B. pelecinus after 27 days and L. corniculatus after 33 days. C. arietinum and S. muricatus were harvested 47 days post inoculation while B. pelecinus and L. corniculatus were harvested 48 days post inoculation. Plants were harvested and nodules collected as described in section 2.6. Nodules were sterilised by soaking in 70% (v/v) ethanol for 30 s, then 4% NaOCl (v/v) for 2 min, followed by washing six times in sterile DI water prior to being crushed and dilution-streaked onto YMA plates. Nodule occupancy was assessed as per Section 2.6, except that isolates were cultured in Yeast-mannitol broth (YMB). Statistical analyses were as previously described (section 2.6), however the WELCH correction was not carried out on the S. muricatus mean shoot dry weight/plant data, as these data were not heterogenous.

25

3.0 RESULTS

3.1 Free-living phenotype of strains

3.1.1 Growth on ½ LA, YMA and TY

To investigate the growth of the bacterial isolates from Scorpiurus spp. on different media, pure cultures of each strain were streaked separately onto YMA, ½ LA and TY agar and incubated for up to

36 days at 28°C. All 19 strains grew to single colonies on YMA, with cultures requiring between two –

13 days to yield single colonies of 2 mm diameter (Table 3.1). Strain WSM1184 showed the fastest rate of growth in forming a colony of 2 mm diameter after two days, followed by strains WSM1386, C and D forming the same sized colonies after six days. The remaining 15 strains took between eight-13 days to produce to 2 mm diameter colonies on this medium. Apart from strains WSM1184 and

WSM1386, growth of all strains was markedly slower on ½ LA than on YMA, with only nine (WSM1184,

WSM1386, A, B, F, H, K, O and S) of the 19 strains achieving a colony size of 2 mm diameter. On TY, growth of five strains (WSM1343, C, D, N and T) was equivalent to that on ½ LA. Growth of most strains was restricted on TY, with only ten strains (WSM1343, WSM1386, C, D, I, K, L, M, N and T) yielding colonies <2 mm diameter, and eight strains (A, B, F, G, H, J, O and S) not growing during 36 days of incubation at 28˚C. WSM1184 was the only strain to grow rapidly and yield 2 mm diameter colonies on TY medium, taking three days.

3.1.2 Temperature tolerance

The temperature tolerance of the 19 strains was investigated by streaking pure cultures of each strain on YMA agar plates and incubating at 10, 16, 22, 28, 32.5 and 37.5°C for up to 30 days. Rate of growth was recorded as days taken to form visible single colonies (≥ 0.3 mm diameter). Strains grew most rapidly at 28°C, producing visible colonies between two and five days after inoculation (Table 3.2).

Apart from F, K and G, all strains showed good growth at 32.5°C, however at 37.5°C, growth was severely inhibited with eight strains (WSM1343, WSM1386, C, D, I, N, S and T) yielding limited growth, one strain showing poor growth (J) and nine strains (A, B, F, G, H, K, L, M and O) yielding no growth at

26 all. All strains grew at 10°C, with growth rate and cell yield decreasing at lower temperatures for most strains. Strains WSM1184, WSM1343, WSM1386, S and T were the only strains that showed good growth at 10°C. Only strain WSM1184 had good growth at both higher and lower temperatures.

Table 3.1: Growth of pure cultures of 19 strains on YMA, ½ LA and TY at 28°C. Strain growth after dilution streaking two plates on three different media; ½ LA, YMA and TY at 28°C. Days of growth were recorded when single colonies reached 2 mm diameter. Strain colony size was recorded after 36 days of growth when single colonies did not reach 2 mm. YMA: Yeast Mannitol Agar, ½ LA: ½ lupin agar, TY: Tryptone yeast agar. YMA ½ LA TY STRAIN Days to 2 mm Days to 2 mm Size at 36 days Days to 2 mm Size at 36 diameter diameter diameter days

WSM1184 2 2 3

WSM1343 9 0.5 0.6

WSM1386 6 6 0.5

A 8 19 No growth

B 12 21 No growth

C 6 0.5 0.7

D 6 0.8 1.5

F 11 20 No growth

G 11 0.3 No growth

H 8 21 No growth

I 13 1.1 0.5

J 8 0.6 No growth

K 10 15 0.4

L 9 1.1 0.6

M 13 1.1 <0.1

N 11 <0.1 0.5

O 9 20 No growth

S 8 19 No growth

T 9 0.4 0.9

27

Table 3.2: Growth of strains after dilution streaking on YMA at 10, 16, 22, 28, 32.5 and 37.5°C for up to 30 days. NIL = no growth; poor growth, (+) 1-4 visible single colonies on initial spread and first streak only; Limited growth, (++) ≥ 5 visible single colonies on initial spread and 1st streak; Growth, (+++) ≥ 5 visible single colonies on second streak; Good growth, (++++) ≥ 5 visible single colonies on third streak. Days recorded when single colonies visible ≥ 0.3mm diameter. Temperature 10°C 16°C 22°C 28°C 32.5°C 37.5°C

Strain Growth Days Growth Days Growth Days Growth Days Growth Days Growth Days WSM1184 ++++ 2 ++++ 2 ++++ 2 ++++ 1 ++++ 1 ++++ 2 WSM1343 ++++ 6 ++++ 5 ++++ 4 ++++ 4 ++++ 4 ++ 5 WSM1386 ++++ 6 ++++ 5 ++++ 4 ++++ 4 ++++ 4 ++ 5 A +++ 11 +++ 7 ++++ 5 ++++ 4 ++++ 4 NIL 30 B +++ 8 +++ 7 ++++ 5 ++++ 4 ++++ 4 NIL 30 C +++ 6 ++++ 5 ++++ 4 ++++ 4 ++++ 4 ++ 5 D +++ 7 ++++ 5 ++++ 4 ++++ 4 ++++ 4 ++ 5 F +++ 11 +++ 7 ++++ 4 ++++ 4 +++ 4 NIL 30 G +++ 11 +++ 7 ++++ 5 ++++ 5 +++ 7 NIL 30 H +++ 8 ++++ 6 ++++ 4 ++++ 4 ++++ 4 NIL 30 I +++ 7 ++++ 6 ++++ 4 ++++ 4 ++++ 4 ++ 5 J +++ 8 +++ 7 ++++ 4 ++++ 4 ++++ 4 + 5 K +++ 7 ++++ 7 ++++ 4 ++++ 4 ++ 7 NIL 30 L +++ 8 ++++ 7 ++++ 4 ++++ 4 ++++ 4 NIL 30 M +++ 8 +++ 7 ++++ 5 ++++ 4 ++++ 6 NIL 30 N ++ 8 +++ 8 +++ 5 ++++ 4 ++++ 4 ++ 5 O +++ 6 +++ 6 ++++ 4 ++++ 4 ++++ 4 NIL 30 S ++++ 7 ++++ 6 ++++ 4 ++++ 4 ++++ 4 ++ 5 T ++++ 7 ++++ 6 ++++ 4 ++++ 4 ++++ 4 ++ 6

28

3.2 Diversity and phylogenetic relationships

The diversity of the 19 strains was initially investigated by Amplified Fragment Length Polymorphism

(AFLP) with the RPO1 primer, with subsequent phylogenetic characterisation determined by 16S rRNA

(for ancestral relatedness) and nifH and nodC (for symbiotic diversity) sequencing.

3.2.1 Diversity based on AFLP

The overall genetic diversity of the 19 strains was first assessed by RPO1-based (AFLP) PCR, with M. ciceri CC1192 (Australian commercial inoculant for Cicer arietinum) and M. ciceri WSM1497

(Australian commercial inoculant for Biserrula pelecinus) included as reference strains. A total of 15 of the strains yielded a profile not shared by any other strain, including WSM1497 and CC1192, indicating a high degree of genetic diversity among the strains (Figure 3.1). Strains H and S showed identical profiles, suggesting that the strains are highly similar to each other. The profile of strains N and T was very similar, with an additional band at approximately 850 bp present in the lane with strain T, which is absent in the profile for strain N. The additional overall differences in intensity for several other bands in both lanes makes it difficult to confidently determine relatedness of these two strains by this method. Overall, RPO1-based PCR revealed a large degree of genetic diversity within the 19 strains.

3.2.2 Phylogeny based on 16S rRNA gene

Analysis of the partial (1,388 bp) 16S rRNA gene sequences separated 18 of the 19 strains into four clades (Figure 3.2). All four clades aligned closely with Mesorhizobium type strains, indicating that the

18 strains could all be classified in the genus Mesorhizobium. BLASTN comparison of these 18 strains matched at a minimum of 99% to known Mesorhizobium spp. (Appendix Table 1). In contrast, strain

WSM1184 did not align well with the 16S rRNA gene sequence of any of the reference strains, with subsequent BLASTN comparison showing the WSM1184 sequence to be 99.79% identical to

Agrobacterium tumefaciens strain A6 over 1413 bp. Therefore, WSM1184 is a likely to be a strain of

Agrobacterium tumefaciens.

29

Figure 3.1: Images of agarose gel electrophoresis of RPO1-based Amplified Fragment Length Polymorphism (AFLP) PCR banding pattern of 21 strains. (A): Strains A, B, C, D, F, G, H, S, I, J and K (B): Strains L, M, N, T, O, WSM1343, WSM1386, WSM1184, characterised M. ciceri strains WSM1497, CC1192 and blank. Promega 1kb ladder used with 13 bands. The blank sample contains the PCR master mix with no DNA template added.

30

Clade I was strongly supported (100% bootstrap value), comprising three strains (WSM1343, C and D) with M. lotiT DSM 2626 and M. australicumT WSM2073 isolated from B. pelecinus growing in Northam,

Western Australia, and four M. ciceri strains, including CC1192 (Figure 3.2). The partial 16S rRNA sequences of strains C and D were identical, while strain WSM1343 was most closely aligned with M. lotiT DSM2626. Clade II contained two strains (N and T) together with M. helmanticenseT CSLC115N,

M. tamadayenseT Ala-3 and M. tianshanenseT A-1BS. Strains N and T matched 100% over the consensus sequence length, consistent with their very similar RPO1 banding pattern (Figure 3.1). Clade

III comprised strain WSM1386 which aligned most closely to M. erdmanii Opo-242, M. erdmaniiT USDA

3471, M. japonicumT MAFF303099, M. opportunistumT WSM2075 and Mesorhizobium sp. AA23. Clade

IV contained 12 strains (A, B, F, G, H, I, J, K, L, M, O and S) together with M. delmotiiT BQ8482, M. muleienseT CGMCC 1.11022, M. prunaredenseT STM4891 and Mesorhizobium sp. AA22 making it the largest grouping of strains. Within clade IV, many strains shared 100% identity over the consensus sequence length, (B and I; A, G, H, J, K, L, M, O and S). In fact, the latter group of nine strains also shared identical 16S rRNA sequences with M. muleienseT CGMCC 1.11022.

Therefore, these data are consistent with all strains, apart from the A. tumefaciens strain WSM1184, being classified in the genus Mesorhizobium. It also shows a considerable degree of genetic diversity among the strains, supporting the data from the earlier AFLP (RPO1) analysis (Section 3.2.1).

3.2.3 nifH and nodC

To assess the symbiotic phylogeny of the strains, nifH (encoding the Fe-protein of nitrogenase) and nodC (encoding N-acetylglucosaminyltransferase involved in Nod factor synthesis) sequence analysis was performed. Due to WSM1184 being classified as A. tumefaciens, nifH and nodC sequencing was not performed on this strain.

31

Figure 3.2: Phylogenetic tree of 19 strains (in blue) based on a partial 16S rRNA sequence (1,388 bp). Type strains M. australicumT WSM2073, M. ciceriT UPM-Ca7, M. delmotiiT BQ8482, M. erdmaniiT USDA 3471, M. helmanticenseT CSLC115N, M. japonicumT MAFF303099, M. lotiT DSM2626, M. muleienseT CGMCC 1.11022, M. opportunistumT WSM2075, M. prunaredenseT STM4891, M. tamadayenseT Ala-3 and M. tianshanenseT A-1BS. Mesorhizobium sp. AA22 and Mesorhizobium sp. AA23 isolated from Modjo, Ethiopia. Three alternative M. ciceri strains: M. ciceri CC1192, M. ciceri WSM1271 and M. ciceri WSM1497. Outgroup Azorhizobium caulinodansT ORS 571. Tamura-Nei genetic distance model was calculated using geneious tree builder software, Geneious 11.1.5. Numbers at nodes indicate levels of bootstrap support based on neighbour-joining analysis of 10000 bootstrap replicates. Bootstrap values below 50% are not shown.

32

Analysis of partial (257 bp) nifH sequences grouped the remaining 18 strains into two clades (Figure

3.3). Clade I was strongly supported (100% bootstrap value) and comprised of M. australicumT

WSM2073, M. ciceri WSM1271, M. ciceri WSM1497, M. delmotiiT BQ8482, M. opportunistumT

WSM2075, M. prunaredenseT STM4891, Mesorhizobium sp. AA22 and all strains except strain J. Some strains in clade I shared identical nifH sequences (B, H and S; N and T) while others differed very little, such as strains C and D and strains L and O whose nifH sequences varied by only one bp with each other over the consensus sequence. Clade II was strongly supported (100% bootstrap value) and contained strain J as the only experimental strain, grouping with M. japonicumT MAFF303099, M. helmanticenseT CSLC115N, M. lotiT DSM 2626 and Mesorhizobium sp. AA23.

Analysis of partial (496 bp) nodC sequences similarly grouped the 18 strains into two different, well- supported clades (clade I bootstrap value of 100% and clade II 76%), although with some variation in groupings compared to the nifH sequence analysis (Figure 3.4). Clade I contained 16 strains (A, B, C,

D, F, G, H, I, J, K, L, M, N, O, S and T) together with M. erdmaniiT USDA 3471, M. helmanticenseT

CSLC115N, M. japonicumT MAFF303099, M. lotiT DSM 2626, Mesorhizobium sp. AA22 and

Mesorhizobium sp. AA23. Unlike the nifH results where strain J clustered away from the other test strains in clade II (Figure 3.3), the strain grouped with clade I of nodC (Figure 3.4) however it still was only distantly related to the other strains and again most closely aligned with Mesorhizobium sp. AA23.

Among other strains within clade I, there was a very high degree of similarity, with five groups of strains sharing identical nodC sequences over the consensus region (A, B, M and Mesorhizobium sp.

AA22; C and D; H and S; N and T; O and L). Clade II was formed by the two remaining strains WSM1343 and WSM1386 in addition to M. erdmanii Opo-242, M. tamadayenseT Ala-3 and M. tianshanenseT

CCBAU 3306 (Figure 3.4), which contrasts with their nifH grouping (Figure 3.3). The nodC sequences of these two strains showed a nine bp difference over the consensus length and a minimum difference of 114 base pairs (strain I) to the other experimental strains. These strains were not closely related to any of the characterised strains, with a minimum difference of 87 base pairs, or 82% match to M. tianshanenseT CCBAU 3306 over the consensus length. Therefore, these data indicate that there is a

33 reasonable amount of genetic similarity in nifH and nodC sequences across the test strains, with a small number showing more distant relatedness for these particular symbiosis genes.

Figure 3.3: Phylogenetic analysis of 18 strains (in blue) based on a partial nifH sequence (257 bp). Type strain M. ciceriT UPM-Ca7 sequence length 85 bp. Type strains M. australicumT WSM2073, M. ciceriT UPM-Ca7, M. delmotiiT BQ8482, M. helmanticenseT CSLC115N, M. japonicumT MAFF303099, M. lotiT DSM 2626, M. muleienseT CGMCC 1.11022, M. opportunistumT WSM2075, M. prunaredenseT STM4891, M. tamadayenseT Ala-3 and M. tianshanenseT CCBAU 3306. Mesorhizobium sp. AA22 and Mesorhizobium sp. AA23 isolated from Modjo, Ethiopia. Three alternative M. ciceri strains: M. ciceri CC1192, M. ciceri WSM1271 and M. ciceri WSM1497, and M. erdmanii Opo-242 were used. Outgroup Azorhizobium caulinodansT ORS 571. Tamura-Nei genetic distance model was calculated using geneious tree builder software, Geneious 11.1.5. Numbers at nodes indicate levels of bootstrap support based on neighbour-joining analysis of 10000 bootstrap replicates. Bootstrap values below 50% are not shown.

34

Figure 3.4: Phylogenetic analysis of 18 strains (in blue) based on a partial nodC sequence (496 bp). Strain I sequence length 429 bp and M. tianshanenseT CCBAU 3306 sequence length 484 bp. Type strains M. australicumT WSM2073, M. ciceriT UPM-Ca7, M. delmotiiT BQ8482, M. erdmaniiT USDA 3471, M. helmanticenseT CSLC115N, M. japonicumT MAFF303099, M. lotiT DSM 2626, M. muleienseT CGMCC 1.11022, M. opportunistumT WSM2075, M. prunaredenseT STM4891, M. tamadayenseT Ala-3 and M. tianshanenseT CCBAU 3306. Mesorhizobium sp. AA22 and Mesorhizobium sp. AA23 isolated from Modjo, Ethiopia. Three alternative M. ciceri strains: M. ciceri CC1192, M. ciceri WSM1271 and M. ciceri WSM1497. M. erdmanii Opo-242 added, matching 16S tree. Outgroup Azorhizobium caulinodansT ORS 571. Tamura-Nei genetic distance model was calculated using geneious tree builder software, Geneious 11.1.5. Numbers at nodes indicate levels of bootstrap support based on neighbour- joining analysis of 10000 bootstrap replicates. Bootstrap values below 50% are not shown.

35

3.3 Symbiotic phenotype of strains on S. muricatus

3.3.1 Yield and nodulation of S. muricatus

To investigate the symbiotic effectiveness of the strains, the nodulation and N2 fixation phenotypes of the strains was assessed by inoculating each strain separately onto S. muricatus and growing plants in

N-limited glasshouse conditions. The nodulation and mean shoot dry weight of inoculated S. muricatus

was compared to uninoculated plants grown either without added N (N-starved) or with added KNO3

(N-Fed). N-starved plants were relatively small in comparison to N-fed plants and had yellowing of leaves with tips turning pink/red in colour (Figure 3.5). For ease of handling, the 19 strains were split into two experimental groups, with the previously characterised S. muricatus-nodulating strain

WSM1386 included in both groups as a control.

Figure 3.5: Photo of Scorpiurus muricatus (experiment 1) 49 days after sowing, N-fed (left) and N-starved (right).

In the first experiment harvested 69 days post-inoculation, nine of the 11 inoculated treatments

(WSM1343, WSM1386, A, D, H, I, G, K and S) out-yielded the mean shoot dry weight/plant of the uninoculated N-starved control (P ≤ 0.05; Figure 3.7A). S. muricatus inoculated with either strains O or WSM1184 produced mean shoot dry weight/plant equivalent to the N-starved control.

36

Uninoculated N-fed control plants significantly out-yielded all the inoculated treatments, with a mean shoot dry weight/plant 1.8 times the highest yielding strain, WSM1343. Nine of the 11 inoculated treatments (WSM1386, D, H, I, G, S, K, O and A) yielded mean shoot dry weights/plant that were not statistically different from each other, ranging from 53.8% (strain WSM1386) to 33.3% (strain A) of the N-fed mean shoot dry weight/plant (Figure 3.7A).

In the second experiment, six of the nine inoculated treatments (C, F, L, M, N and T) out-yielded the mean shoot dry weight/plant of the uninoculated N-starved control, harvested 76 days post- inoculation (P ≤ 0.05; Figure 3.8A). S. muricatus inoculated with strains WSM1386, B and J produced mean shoot dry weight/plant equivalent to the N-starved control. However, unlike experiment 1, the

N-fed control plants did not produce significantly more mean shoot dry weight than the highest yielding inoculated treatments WSM1386, C, F, L, M, N and T.

The mean shoot dry weight/plant in experiment 2 were substantially reduced compared with experiment 1, with N-fed controls yielding a mean shoot dry weight/plant of 0.415 g and 0.617 g respectively, and N-starved controls 0.036 g and 0.048 g respectively. S. muricatus inoculated with strain WSM1386 in experiment 1 produced 53.8% of the N-fed control mean shoot dry weight/plant and was statistically different, in contrast to experiment 2, where S. muricatus inoculated with strain

WSM1386 produced 63.8% of the N-fed control mean shoot dry weight/plant and were not statistically different. However, while all other inoculated treatments in experiments 1 and 2 consisted of eight plants per treatment, the WSM1386 treatment in experiment 2 consisted of only four plants, with this treatment showing a large variation in shoot dry weight/plant. Similar to experiment 1, seven of the nine inoculated treatments in experiment 2 yielded mean shoot dry weight/plant that were not statistically different from each other (Figure 3.8A groupings ‘a’, ‘b’ and ‘c’).

All strains in both experiments nodulated S. muricatus, producing nodules that were indeterminate, rarely coralloid, cylindrical and pink in colour (Figure 3.6A and B) apart from strain WSM1184 which did not form nodules. S. muricatus inoculated with strain WSM1386 formed the highest mean nodule

37 count per plant at 149 (Figure 3.7B), followed by strain C forming 127 nodules/plant and WSM1386 forming 126 nodules/plant (Figure 3.8B). Nodule number varied considerably across treatments, with

S. muricatus inoculated with strain J forming the lowest mean of 7.5 nodules/plant, followed by strain

B forming 28 nodules/plant and strain O forming 34 nodules/plant, while the mean for strain

WSM1386 across both experiments was 141 nodules/plant. However, there was no clear correlation between mean shoot dry weight/plant and mean nodules/plant. Six of the 11 inoculated treatments in experiment 1 yielded mean nodules/plant that were not statistically different from each other

(Figure 3.7B groupings ‘b’ and ‘c’) and five of the nine treatments in experiment 2 yielded mean nodules/plant that were not statistically different from each other (Figure 3.8B groupings ‘a’, ‘b’ and

‘c’). Plants inoculated with strain WSM1184 produced white growths that were highly dissimilar in structure to the nodules formed on plants in other inoculated treatments. No bacteria could be reisolated from tissue crushes of these growths. This is consistent with the 16S rRNA sequencing data indicating strain WSM1184 to be A. tumefaciens.

Figure 3.6: Photos of nodules formed on Scorpiurus muricatus at harvest, 69 days post-inoculation, inoculated with strain D. (A) Plant D 4 with large pink coralloid nodule (12 mm x 14 mm) and pink indeterminate cylindrical nodules (B) Plant D 8 showing pink indeterminate cylindrical nodules to 8 mm in length. Scale bars represent 1 cm.

38

0.7 a (A)

0.6

0.5

b 0.4 bc bc bc bc bc bc 0.3 c bcd c

0.2 Mean shoot (g/plant) shoot weight dry Mean

0.1 d d

0.0

Treatment

180 (B) a 160

140 ab 120 abc abc abc bc 100 bcd cde 80

60 de Mean nodules/plant Mean e 40

20 f 0 1343 1386 D H I G S K O A 1184

Treatment

Figure 3.7 Symbiotic effectiveness of 11 strains (WSM1184, WSM1343, WSM1386, A, D, G, H, I, K, O and S) on S. muricatus, showing (A) Mean shoot dry weight/plant and (B) Mean nodule number/plant. Plants were grown in N-limited conditions and either uninoculated or inoculated separately with 11 strains. The uninoculated N- starved plants received no added N while N-fed plants were supplied nitrogen as KNO3. Plants were harvested at 69 days post-inoculation. Means with the same letter are not statistically different, while different letters indicate significant differences according to one-way ANOVA using RStudio (Version 1.2.1335) at a significance level of P ≤ 0.05 with WELCH correction for heterogeneity of variance. Error bars show standard error.

39

0.7 (A)

0.6

0.5 a

0.4

a abcd abc 0.3 abc abc abc ab

0.2 Mean shoot shoot Mean (g/plant) weight dry bcd cd 0.1 d

0.0

Treatment

180 (B)

160 ab a 140

120 abc 100 abc abc 80 bcd cd

60 Mean nodules/plant Mean

40 de

20 e

0 T 1386 F N L C M B J

Treatment

Figure 3.8: Symbiotic effectiveness of nine strains (WSM1386, B, C, F, J, L, M, N and T) on S. muricatus, showing (A) Mean shoot dry weight/plant and (B) Mean nodule number/plant. Plants were grown in N-limited conditions and either uninoculated or inoculated separately with nine strains. The uninoculated N-starved plants received no added N while N-fed plants were supplied nitrogen as KNO3. Plants were harvested at 76 days post- inoculation. Means with the same letter are not statistically different, while different letters indicate significant differences according to one-way ANOVA using RStudio (Version 1.2.1335) at a significance level of P ≤ 0.05 with WELCH correction for heterogeneity of variance. Error bars show standard error.

40

Nodule occupant strain identity was determined by comparing the RPO1 banding pattern of nodule isolates to the original inoculant strain (Figures 3.9A and B, and 3.10A and B). All inoculant strains, except strain G, were confirmed as occupying nodules on S. muricatus. The non-template control in one of the RPO1 analyses did yield a very faint band at 750 bp (Figure 3.9A), however this did not match any of the strains and was therefore unlikely to have affected the profiling. The single nodule isolate obtained from nodule crushes of S. muricatus inoculated with strain G did not match any of the inoculant banding patterns, suggesting a non-experimental strain may have nodulated this treatment. In total, five of the N-starved control plants were nodulated in experiments 1 and 2, the

RPO1 profile of the N-starved isolate obtained in experiment 1 matching strain WSM1386, and the

RPO1 profile of N-starved isolates in experiment 2 matching strains N and T. This indicates likely cross- contamination of the N-starved controls by these strains. These N-starved plants were removed from the mean shoot dry weight/plant analysis (Figures 3.7A and 3.8A). Therefore, effectiveness experiments 1 and 2 showed that all Mesorhizobium strains (apart from strain G) formed nodules and fixed N2 on S. muricatus with a variation in N2 fixation efficiency observed.

3.4 Host range of Mesorhizobium strains.

A subset of six of the 18 Mesorhizobium strains (WSM1343, WSM1386, A, D, J and T) was selected to investigate their ability to nodulate and fix N2 on legumes known to form symbiotic interactions with

Mesorhizobium spp. A selection of strains isolated from Israel (A and D), Morocco (WSM1343, J and

T) and Australia (WSM1386) were chosen. Strains were also selected based on their relative N2 fixation effectiveness level on S. muricatus, so that representative effective (WSM1343, WSM1386, D, and T) and less effective (A and J) strains were included (Figures 3.7A and 3.8A). Strains were inoculated separately onto S. muricatus, Biserrula pelecinus, Lotus corniculatus and Cicer arietinum in N-limited glasshouse conditions. The nodulation and mean shoot dry weight of inoculated plants was compared

to uninoculated plants either devoid of (N-starved) or fed KNO3 (N-Fed). S. muricatus was included to confirm the nodulation phenotype of the strains.

41

Figure 3.9: Images of agarose gel electrophoresis of RPO1-based Amplified Fragment Length Polymorphism (AFLP) PCR banding pattern of bacteria isolated from S. muricatus nodules. Original strains represented by identifying letter or number located in lane prior to nodule crush isolates 1 or 2. (A): Strains A, D, G, H, I, K, O and S with corresponding nodule isolates and blank (B): Strains WSM1343 and WSM1386 with corresponding nodule isolates and N-starved nodule isolate (N- P7 1). Promega 1kb ladder used with 13 bands. The blank sample contains the PCR master mix with no DNA template added.

42

Figure 3.10: Images of agarose gel electrophoresis of RPO1-based Amplified Fragment Length Polymorphism (AFLP) PCR banding pattern of bacteria isolated from S. muricatus nodules. Original strains represented by identifying letter or number located in lane prior to nodule crush isolates 1, 2 or 3. (A): Strains B, C, F, J, L, M, N and T with corresponding nodule isolates (B): Strain WSM1386 and corresponding nodule isolates, N-starved nodule isolates (N- P7 1, N- P7 2, N- P8 1 and N- P8 2), empty (no PCR reaction loaded in lane) and Blank. Promega 1kb ladder used with 13 bands. The blank sample contains the PCR master mix with no DNA template added.

43

C. arietinum and S. muricatus were harvested 47 days after inoculation while L. corniculatus and B. pelecinus plants were harvested 48 days after inoculation. All S. muricatus plants formed pink nodules when inoculated with strains WSM1343, WSM1386, D and T, although only half of the plants inoculated with A and J were nodulated (Table 3.4). All inoculated B. pelecinus formed nodules, except for strain WSM1343, where three plants inoculated with WSM1343 formed a mixture of pink, light pink and green nodules while the fourth plant was not nodulated. Nodule colour also varied for B. pelecinus inoculated with strain J, forming light pink, green and white nodules. L. corniculatus formed pink nodules for all treatments, although the percentage of plants nodulated in each treatment varied widely from 100% nodulated for strains WSM1386 and D, 75% for strain A, 50% for strains WSM1343 and T, and 25% for strain J. In contrast, no inoculated or uninoculated C. arietinum treatments were nodulated, indicating that these Mesorhizobium strains do not form a symbiotic interaction with C. arietinum.

Table 3.4: Nodulation of Mesorhizobium strains WSM1343, WSM1386, A, D, J and T on three host plants: Biserrula pelecinus, Lotus corniculatus and Scorpiurus muricatus. Plants were grown in N-limited conditions and either uninoculated or inoculated separately with six strains. The uninoculated N-starved plants received no added N while N-fed plants were supplied nitrogen as KNO3. Plants were harvested 47 (S. muricatus) and 48 days (B. pelecinus and L. corniculatus) post inoculation. (0 – 100) percentage of plants that formed nodules. Nodule colour described. C. arietinum did not form nodules and was excluded.

Treatment Plant species S. muricatus L. corniculatus B. pelecinus % plants Nodule colour % plants Nodule % plants Nodule nodulated nodulated colour nodulated colour WSM1343 100 Pink 50 Pink 75 Pink, light pink and green WSM1386 100 Pink 100 Pink 100 Pink A 50 Pink and white 75 Pink 100 Pink D 100 Pink 100 Pink 100 Pink J 50 Pink and white 25 Pink 100 Light pink, green and white T 100 Pink 50 Pink 100 Pink N-fed - - 100 Light pink and white N-starved 100 Pink - 100 Light pink

For both S. muricatus and B. pelecinus, nodules were observed on uninoculated N-starved, and N-fed and N-starved controls, respectively (Table 3.4). Bacteria isolated separately from nodules excised

44 from plants from each of these treatments showed RPO1 profiles that matched strain WSM1386, indicating a likely unintentional cross-inoculation of N-starved plants during inoculation (Figures 3.11B and 3.12B). RPO1 profiles of bacteria isolated from all inoculated treatments were consistent with their respective inoculant strain in 28 out of 35 cases (Figures 3.11, 3.12 and 3.13). In five cases, the profile of one of the isolates matched the inoculant strain, while the profile of the second isolate taken from a different nodule of a plant in the same treatment did not match any of the reference strain profiles [S. muricatus inoculated with strain WSM1343 (Figure 3.11B), B. pelecinus with strain J (Figure

3.12A) and L. corniculatus inoculated separately with strains WSM1343, WSM1386 and T (Figure 3.13A and B)]. Only for B. pelecinus inoculated with strain D, did both nodule isolates not match the reference strain profile. This indicates that in these small number of instances, a different, non- experimental strain likely nodulated these individual plants.

The mean shoot dry weight data for S. muricatus inoculated separately with the six strains showed the strains were fixing N2 on this host, with a variation in efficiency of N2 fixation consistent with data obtained in earlier effectiveness experiments 1 and 2 (Figures 3.7A, 3.8A and 3.14A). Plants inoculated with strain WSM1386 produced 59.4% of N-fed mean shoot dry weight/plant, in comparison to experiment 1 (53.8%) and experiment 2 (63.8%). S. muricatus inoculated with strain D produced statistically-equivalent mean shoot dry weight/plant to N-fed control, while plants inoculated separately with the remaining strains (WSM1343, WSM1386, A, J and T) produced significantly less mean shoot dry weight/plant than the N-fed control (Figure 3.14A). B. pelecinus plants inoculated separately with strains WSM1343, WSM1386, A, D, J and T produced significantly less mean shoot dry weight/plant than the N-fed control, instead producing plant shoot biomass that was not statistically different to the N-starved (nodulated) control (Figure 3.14B). This indicates that although B. pelecinus was nodulated by the strains, they do not fix N2 effectively with this host. In contrast, although all inoculated L. corniculatus treatments produced mean shoot dry weight/plant that was more than 4.9- fold lower than the N-fed control plants, these data were not statistically different to either N-Fed or

N-starved (non-nodulated) control plants (Figure 3.14C). Therefore, while the means are suggestive

45 that the strains fix N2 poorly on this host, this cannot be definitively stated with the data from this experiment.

Overall, these data suggest that while this subset of Mesorhizobium strains do not nodulate C. arietinum, they are capable of nodulating B. pelecinus and L. corniculatus, with the effectiveness data indicating they are likely to be poorly effective N2 fixers on the latter two hosts.

Figure 3.11: Images of agarose gel electrophoresis of RPO1-based Amplified Fragment Length Polymorphism (AFLP) PCR banding pattern of bacteria isolated from S. muricatus nodules. Original strains represented by identifying letter or number located in lane prior to nodule crush isolates ‘1’ or ‘2’. (A): Strains WSM1343, A, D, J and T with corresponding nodule isolates (B): Isolate WSM1343 ‘2’, Strain WSM1386 and corresponding nodule isolates, N-starved nodule isolates (N- P1 1, N- P1 2, N- P2 1, N- P2 2, N- P3 1, N- P3 2, N- P4 1 and N- P4 2), and Blank. Promega 1kb ladder used with 13 bands. The blank sample contains the PCR master mix with no DNA template added.

46

Figure 3.12: Images of agarose gel electrophoresis of RPO1-based Amplified Fragment Length Polymorphism (AFLP) PCR banding pattern of bacteria isolated from B. pelecinus nodules. Original strains represented by identifying letter or number located in lane prior to nodule crush isolates ‘1’ or ‘2’. (A): Strains WSM1343, A, D, J and T with corresponding nodule isolates, WSM1343 isolates identified as ‘small’ isolated from small green nodules. (B): Strain WSM1386 and corresponding nodule isolates, N-starved (N- P1 1, N- P1 2, N- P4 1 and N- P4 2) and N-fed (N+ P1 1, N+ P1 2, N+ P2 1 and N+ P2 2) isolates and blank. Promega 1kb ladder used with 13 bands. The blank sample contains the PCR master mix with no DNA template added.

47

Figure 3.13: Images of agarose gel electrophoresis of RPO1-based Amplified Fragment Length Polymorphism (AFLP) PCR banding pattern of bacteria isolated from L. corniculatus nodules. Original strains represented by identifying letter or number located in lane prior to nodule crush isolates ‘1’ or ‘2’. (A): Strains A, D and T with corresponding nodule isolates (B): Strains WSM1343, WSM1386 and J with corresponding nodule isolates and blank. Promega 1kb ladder used with 13 band. The blank sample contains the PCR master mix with no DNA template added.

48

0.18 a (A) 0.15

0.12 ab bc bc bc 0.09 bc 0.06 c 0.03

Mean shoot shoot Mean (g/plant) weight dry 0.00 N-fed D 1386 1343 T A J Treatment

0.25 a (B)

0.20

0.15

0.10

b 0.05 b b b b

b b Mean shoot shoot Mean (g/plant) weight dry 0.00 N-fed T D A 1386 1343 N-starved J Treatment

(C) 0.07 ab 0.06

0.05

0.04

0.03

0.02 b ab ab 0.01 a a a ab

Mean shoot shoot Mean (g/plant) weight dry 0.00 N-fed D 1386 1343 A T N-starved J Treatment

Figure 3.14: Symbiotic effectiveness, shown as mean shoot dry weight/plant, of six strains (WSM1343, WSM1386, A, D, J and T) on three host plants (A) S. muricatus (B) B. pelecinus (C) L. corniculatus. Plants were grown in N-limited conditions and either uninoculated or inoculated separately with six strains. The uninoculated N-starved plants received no added N while N-fed plants were supplied nitrogen as KNO3. Plants were harvested 47 (S. muricatus) and 48 days (B. pelecinus and L.corniculatus) post inoculation. Means with the same letter are not statistically different, while different letters indicate significant differences according to one- way ANOVA using RStudio (Version 1.2.1335) at a significance level of P ≤ 0.05 with WELCH correction for heterogeneity of variance for B. pelecinus and L. corniculatus. Effectively nodulated N-starved S. muricatus were omitted. Error bars show standard error.

49

4.0 DISCUSSION

The aims of this thesis were to characterise strains isolated from Scorpiurus sp. by investigating the free-living phenotype and phylogenetic relationship of the organisms and to assess their ability to form a symbiotic relationship with S. muricatus, Cicer arietinum, Biserrula pelecinus and Lotus corniculatus.

4.1 Growth characteristics and diversity of Mesorhizobium strains

A total of 18 of the 19 test strains were classified as Mesorhizobium spp. based on 16S rRNA sequence analysis, and these strains showed considerable variation in their growth rates and temperature tolerance. On YMA, only three strains (WSM1386, C and D) took less than eight days to grow colonies to 2 mm diameter at 28°C, the remainder taking between eight – 13 days. This latter growth rate is substantially slower than what is generally considered to be characteristic of the genus, with rates of between five to seven days (O'Hara et al., 2016b) to form colonies of at least 2 mm diameter at 28°C on YMA. In fact, the growth rate of the majority of these Mesorhizobium strains aligns more closely with that described for slow-growing Bradyrhizobium species, at five to eight days to form colonies of less than 1 mm diameter (O'Hara et al., 2016b). In their study of organisms isolated from Asinara Island

(Italy), Safronova et al. (2004) describe the growth rate of their S. muricatus isolates as being between four to seven days at 28°C to form visible colonies on YMA, although they do not specify the colony size achieved in this time frame. Therefore, it appears that the growth rate of many of the S. muricatus isolates investigated in this thesis may be atypical for this bacterial genus. Isolating more strains from

S. muricatus nodules and assessing their growth rate would be informative to see if this slower-growth rate is a general feature of rhizobia that associate with this legume.

On ½ LA medium, strain growth rates were substantially slower than on YMA, with more than half of the strains unable to form colonies of 2 mm diameter after 36 days. ½ LA medium is widely used for rhizobial culturing and nodule isolations of rhizobia from many genera, including Mesorhizobium sp.

50

(Hungria et al., 2016), with its low phosphate concentration specifically chosen to reduce strain gumminess, allowing for ease of subculturing (Howieson et al., 1988, Hungria et al., 2016). The poor growth of some of the isolates on ½ LA is consistent with the generally slower growth rate observed for these rhizobia on YMA. However, it could also indicate that the specific nutritional requirements of these Mesorhizobium strains may differ to other previously characterised organisms. Further work characterising the nutritional requirements of these strains is therefore required. On TY, growth of the strains was further restricted, with only ten of 18 strains achieving any growth at all. TY is a rich medium originally developed from LB, a medium well-suited to the culturing of enteric organisms

(Bertani, 1951). Although TY is routinely used as the medium of choice for the molecular characterisation of model rhizobia, many strains of Mesorhizobium grow poorly on this medium (J.

Terpolilli, unpublished results).

Optimal growth on YMA for all Mesorhizobium strains ranged between 22 – 32.5°C, with 28°C identified as the apparent optimum growth temperature. All strains grew at the lowest tested temperature of 10°C, but only nine grew at 37.5°C. O'Hara et al. (2016b) describe the optimal growth temperature for Mesorhizobium spp. as 25 – 30°C, with temperature tolerance ranging from 4 – 10°C to 37 – 42°C. The tolerance of these Mesorhizobium strains to temperature is therefore consistent with that reported in the literature for this genus.

Interestingly, in their study of S. muricatus organisms from Algeria, Bouchiba and coworkers (2017) isolated bacteria from S. muricatus nodules by culturing isolates on YMA at 26°C for 3-7 days.

However, all characterised isolates subsequently failed to nodulate the original host in authentication experiments, with partial 16S rRNA sequencing indicating that many of the strains were most closely related to Rhizobium species. Rhizobium spp. are generally considered to be fast-growing rhizobia, achieving colonies of 2 mm in diameter in 3-5 days on YMA (O'Hara et al., 2016b). The work in this thesis has shown some Mesorhizobium strains from Scorpiurus can have growth rates of greater than

8 days at 28°C. Therefore, it is possible that Bouchiba et al. (2017) may have failed to isolate slower-

51 growing S. muricatus-nodulating organisms in their study due to insufficient incubation time and possibility slightly reduced incubation temperatures.

Assessing strain ancestral relatedness by 16S rRNA sequencing classified 18 of the 19 strains as

Mesorhizobium, with alignment of 16S sequences showing the strains to be related to a diversity of microsymbionts within this genus. Half of the Mesorhizobium strains matched 100% with M. muleienseT CGMCC 1.11022 isolated from a root nodule of C. arietinum, Xinjiang, China (Zhang et al.,

2012), while the remaining Mesorhizobium isolates grouped with characterised Mesorhizobium spp. known to nodulate legumes across a wide diversity of different legume genera, including C. arietinum,

B. pelecinus, L. corniculatus and Lotus japonicus. Strain WSM1184 sourced from ICARDA (International

Centre for Agricultural Research in the Dry Areas, Syria), was identified as Agrobacterium tumefaciens and was unable to nodulate S. muricatus. It is likely that the fast growth rate of WSM1184 meant that it may have outcompeted other bacteria when isolated from S. sulcatus nodules and mistakenly identified as a possible nodule-forming bacterium.

Most of the Mesorhizobium strains shared highly similar nifH and nodC sequences, with one large clade comprising the majority of strains in the phylogenetic trees for both genes. The close grouping of WSM1343 and WSM1386 in both nifH and nodC trees, while their 16S sequences are only very distantly related, is an interesting finding, as both strains were isolated from geographically distant regions, with WSM1343 isolated from S. sulcatus located in Oulmes, Morocco (1993), while WSM1386 was isolated from S. sulcatus growing at Manjimup research station in 1994, near a site that had previously been used for a B. pelecinus trial (Western Australia). Scorpiurus spp. are not native to

Australia, with the plants at the Manjiump station being part of a small evaluation trial and consequently are likely to have been inoculated, although records of this are not available to confirm whether inoculation took place and with which strain. The 16S sequence of WSM1386 grouped this strain most closey to M. opportunistumT WSM2075, a strain isolated from B. pelecinus growing in

Northam (Western Australia) that was shown to have acquired the symbiosis genes from the mobile

52 symbiosis ICE of M. ciceri WSM1271, the original inoculant for this pasture legume (Nandasena et al.,

2007, 2009, Haskett et al., 2016). Therefore, WSM1386 may be a native Australian strain similar to M. opportunistumT WSM2075 that has acquired symbiosis genes from an inoculant applied to the S. sulcatus field site that harboured nifH and nodC genes similar to those of WSM1343. This acquisition would have enabled WSM1386 to form a symbiosis with S. muricatus. This suggests that, like M. ciceri

WSM1271, and M. japonicum R7A, the S. muricatus Mesorhizobium strains may also harbour mobile symbiosis ICEs. Further work is required to determine whether the Mesorhizobium strains analysed in this thesis harbour these mobile genetic elements.

4.2 Symbiotic effectiveness and host range

The Mesorhizobium strains investigated in this thesis were able to fix N2 with S. muricatus, with comparison of mean shoot dry weights showing a wide range of symbiotic effectiveness of strains, ranging from equivalent to the N-fed control to equivalent to the N-starved control. Each strain was authenticated and confirmed to fix N2 on S. muricatus, except for strain G where the RPO1 banding pattern of the single cultured isolate obtained from nodules taken from plants in the effectiveness experiment, did not match the original inoculant. For strain G, additional symbiotic testing is required to confirm the N2 fixation phenotype of this strain on S. muricatus.

The mean shoot dry weights of control plants in the second effectiveness experiment were substantially lower than those achieved in the first experiment, suggesting a difference in experimental conditions between these two experiments. These symbiotic effectiveness assays were conducted at different times of the year, with experiment 1 commencing on 27 March, while experiment 2 commenced 29 days later on 25 April 2019. Plant growth in the glasshouse in which the experiments were conducted is driven by natural sunlight, with a limited degree of temperature control. It is possible therefore that the differences in growth yield between these two experiments is a consequence of reduced daylength and/or decrease in temperature in the second experiment, leading to slower plant growth rates. To minimise these effects, a larger, single experiment containing

53 all test strains could have been conducted, however a lack of sufficient space in the glasshouse at the time of conducting the first effectiveness experiment precluded this option. Nevertheless, the data from these experiments suggest that some of the more effective Mesorhizobium strains investigated in this thesis may be candidate commercial inoculant strains for S. muricatus, such as strains

WSM1343, WSM1386, D, F, N and T.

The host range experiment evaluated whether S. muricatus microsymbionts can nodulate and fix N2 on B. pelecinus, C. arietinum and L. corniculatus, which are legumes of agricultural importance in

Australia. Except for strain D, all strains were confirmed as nodulating B. pelecinus and L. corniculatus, while no strains were able to nodulate C. arietinum. The lack of nodulation of C. arietinum by these strains indicates that S. muricatus-nodulating Mesorhizobium sp. may be unable to symbiotically interact with this host. This could be confirmed by testing some of the other 12 Mesorhizobium strains characterised in this thesis. If there is shown to be broad incompatibility of these strains with C. arietinum, then it suggests that these strains would not compete with C. arietinum inoculants if released into agriculture, consequently reducing inoculant efficacy.

Nandasena et al. (2004) have shown that M. ciceri WSM1284 which can nodulate B. pelecinus, was also able to nodulate L. corniculatus, L. ornithopodiodes and L. pedunculatus, while two other B. pelecinus-nodulating strains (WSM1283 and WSM1497) could not. Similarly, the Lotus-nodulating strain M. japonicum R7A is able to nodulate B. pelecinus (Haskett et al., 2016) however none of these four strains is capable of nodulating C. arietinum [Nandasena et al. (2004) & Y. Hill, unpublished results for R7A]. Conversely, while the nodulation phenotype of CC1192 (the commercial inoculant for C. arietinum) on Lotus spp. is not known, it does not nodulate B. pelecinus (Nandasena et al., 2004).

Therefore, there appears to be a distinct host range difference between strains that nodulate annual pastures such as B. pelecinus, Lotus sp. and perhaps S. muricatus and those that nodulate the grain legume C. arietinum. It will be informative to further characterise the host range across other well- characterised B. pelecinus, Lotus sp. and C. arietinum-nodulating strains.

54

The nodulation of N-starved and N-fed control plants in some of the treatments for the effectiveness and host range experiments was undesirable, however, in all cases the RPO1 profile obtained from isolates matched that of one of the inoculant strains. This indicates that the cause of nodulation was likely to have been cross-contamination during inoculation. The N-starved and N-Fed controls received

1 ml of sterile 1% sucrose per seedling at the time of inoculation, and this solution was delivered after the inoculated treatments received their respective strain using the same pipette. It is possible therefore that residual quantities of inoculant may have been inadvertently transferred from these inoculated treatments to the control plants, leading to the nodulation observed. Future experiments will benefit from inoculating with separate sterile syringes for each treatment to eliminate this source of cross-contamination.

Although WSM1343, WSM1386, A, J and T were able to nodulate B. pelecinus, and L. corniculatus, they did not fix N2 effectively, producing mean shoot dry weights equivalent to the N-starved control plants. However, all inoculated treatments did produce pink nodules on the plant species tested, indicating the possibility that the strains could fix N2, albeit not efficiently. Given the uncertainty in results interpretation due to the nodulation of some N-fed and N-starved uninoculated controls for these host plants, the N2-fixing capability of these strains on these hosts needs to be quantified in additional experiments before firm conclusions on their efficiency can be made.

4.3 Concluding statement and future directions

Scorpiurus muricatus has the potential to meet the requirements of an annual legume for medium-to- low rainfall areas of southern Australia. The future adoption of this pasture legume into these agricultural systems requires the co-introduction of an effective microsymbiont. This project characterised the phylogeny as well as free-living and symbiotic phenotype of a range of S. muricatus nodulating organisms as a first step to identifying an elite inoculant strain for this host plant.

Many of the S. muricatus-nodulating Mesorhizobium spp. investigated in this thesis showed growth rates on YMA that were slower than what is generally accepted for this genus. This suggests that these

55 organisms may represent a new “slow growing” group of Mesorhizobium spp.. It would therefore be informative to measure the mean generation time of these strains in liquid media and further investigate their nutritional requirements, compared with other well-characterised Mesorhizobium strains. Future studies should take account of the potential for a slower growth rate and therefore allow sufficient time for growth when isolating microsymbionts from S. muricatus and potentially other Scorpiurus spp..

The phylogenetic analysis of chromosomal 16S rRNA compared to symbiosis (nifH and nodC) genes, particularly in the case of WSM1386 and WSM1343, is highly suggestive that these S. muricatus microsymbionts harbour mobile symbiosis ICEs, as has been described for other strains of

Mesorhizobium. In the case of the Mesorhizobium inoculant strain for B. pelecinus, horizontal transfer of the symbiosis ICE in the soil resulted in the evolution of novel microsymbionts that were suboptimally effective N2 fixers with this host, potentially reducing the efficacy of the inoculant. It is therefore imperative that the strains characterised in this thesis be interrogated for the presence of symbiosis ICEs. The mobility of any ICEs detected can subsequently be investigated by conducting conjugation experiments between the S. muricatus strains and the R7ANS ICE-devoid recipient strain, as has been previously described (Haskett et al., 2016).

The symbiotic effectiveness of the best-performing Mesorhizobium strain tested in this thesis was

67.5% of the N-fed control. The ideal inoculant for this pasture legume would have as high a N2 fixation rate as possible and preferably one that produced shoot dry weights >90% of the N-Fed control.

Therefore, more strains of S. muricatus-nodulating organisms should be screened for their symbiotic effectiveness. It is worth noting that the Mesorhizobium strains characterised in this thesis that were collected from nodulated plants in the field (WSM1343 and WSM1386), were isolated from S. sulcatus plants rather than S. muricatus, and that the soil trapping isolates were collected from locations described as having Scorpiurus sp. present, although not specifically S. muricatus. It is not known if S. sulcatus and S. muricatus microsymbionts are equally effective N2-fixers on both species, however,

56 further studies on S. muricatus symbionts may benefit by isolating microsymbionts from S. muricatus nodules and soils in the field.

A subset of the Mesorhizobium strains tested in this thesis were able to nodulate B. pelecinus and L. corniculatus (WSM1343, WSM1386, A, D, J and T), with the effectiveness data suggesting that they fixed N2 poorly on these hosts. As B. pelecinus and Lotus sp. are pre-existing pasture legumes in

Australian agricultural systems, release of an S. muricatus isolate that nodulates but fixes poorly with these legumes could have substantive negative impacts on their N2 fixation. The N2 fixation phenotype of these strains on B. pelecinus and Lotus sp. therefore needs to be confirmed in subsequent effectiveness experiments. These experiments could also be expanded to include all the isolates tested in this thesis, as some may show a different phenotype on these hosts. A strain that fixed N2 well on

S. muricatus but did not nodulate either B. pelecinus or L. corniculatus, as was the case with the subset of strains tested on C. arietinum in this thesis, would be ideal. Alternatively, a strain that was effective across all three species would be a candidate commercial strain to recommend for all these pasture legumes, provided that the strain didn’t nodulate C. arietinum. Furthermore, future work also needs to investigate whether there is any symbiotic interaction between current Australian commercial inoculants for B. pelecinus (WSM1497), C. arietinum (CC1192) and Lotus sp. (SU343, CC829) with S. muricatus. This will determine the chance of any possible negative impacts of these strains on the N2 fixation of S. muricatus if this legume is introduced as an annual pasture for medium-to-low rainfall areas of southern Australia.

57

5.0 BIBLIOGRAPHY

Abbate, V., Maugeri, G., Cristaudo, A. & Gresta, F. (2010) Scorpiurus muricatus L. subsp. subvillosus

(L.) Thell., a potential forage legume species for a Mediterranean environment: a review.

Grass and Forage Science 65: 2-10.

Atallah, T., Rizk, H., Cherfane, A., Daher, F.B., El-Alia, R., De Lajuide, P. & Hajj, S. (2008) Distribution

and Nodulation of Spontaneous Legume Species in Grasslands and Shrublands in

Mediterranean Lebanon. Arid Land Research and Management 22: 109-122.

Barry, R.G. & Chorley, R.J., (2003) Atmosphere, weather, and climate. In. London ; New York:

Routledge, pp. 471.

Beale, P.E., Lahlou, A. & Bounejmate, M. (1991) Distribution of wild annual legume species in

Morocco and relationship with soil and climatic factors. Australian Journal of Agricultural

Research 42: 1217-1230.

Beringer, J.E. (1974) R factor transfer in Rhizobium leguminosarum. Journal of General Microbioly 84:

188-198.

Bertani, G. (1951) Studies on lysogenesis. I. The mode of phage liberation by lysogenic Escherichia

coli. Journal of Bacteriology 62: 293-300.

Black, M., Moolhuijzen, P., Chapman, B., Barrero, R., Howieson, J., Hungria, M. & Bellgard, M.

(2012) The genetics of symbiotic nitrogen fixation: comparative genomics of 14 rhizobia

strains by resolution of protein clusters. Genes 3: 138-166.

Bouchiba, Z., Boukhatem, Z.F., Ighilhariz, Z., Derkaoui, N., Kerdouh, B., Abdelmoumen, H., Abbas,

Y., Missbah El Idrissi, M. & Bekki, A. (2017) Diversity of nodular bacteria of Scorpiurus

muricatus in western Algeria and their impact on plant growth. Canadian Journal of

Microbiology 63: 450-463.

58

Boulygina, E.S., Kuznetsov, B.B., Marusina, A.I., Tourova, T.P., Kravchenko, I.K., Bykova, S.A.,

Kolganova, T.V. & Galchenko, V.F. (2002) A study of nucleotide sequences of nifH genes of

some methanotrophic bacteria. Microbiology 71: 425-432.

Callaham, D.A. & Torrey, J.G. (1981) The structural basis for infection of root hairs of Trifolium

repens by Rhizobium. Canadian Journal of Botany 59: 1647-1664.

Capela, D., Filipe, C., Bobik, C., Batut, J. & Bruand, C. (2006) Sinorhizobium meliloti differentiation

during symbiosis with alfalfa: a transcriptomic dissection. Molecular Plant-Microbe

Interactions 19: 363-372.

Chen, W., Wang, E., Wang, S., Li, Y., Chen, X. & Li, Y. (1995) Characteristics of Rhizobium

tianshanense sp. nov., a moderately and slowly growing root nodule bacterium isolated from

an arid saline environment in Xinjiang, People's Republic of China. International Journal of

Systematic Bacteriology 45: 153-159.

Chen, W.M., Laevens, S., Lee, T.M., Coenye, T., De Vos, P., Mergeay, M. & Vandamme, P. (2001)

Ralstonia taiwanensis sp. nov., isolated from root nodules of Mimosa species and sputum of

a cystic fibrosis patient. International Journal of Systematic and Evolutionary Microbiology

51: 1729-1735.

Chen, W.M., Zhu, W.F., Bontemps, C., Young, J.P. & Wei, G.H. (2010) Mesorhizobium alhagi sp.

nov., isolated from wild Alhagi sparsifolia in north-western China. International Journal of

Systematic and Evolutionary Microbiology 60: 958-962.

Chen, W.M., Zhu, W.F., Bontemps, C., Young, J.P. & Wei, G.H. (2011) Mesorhizobium camelthorni

sp. nov., isolated from Alhagi sparsifolia. International Journal of Systematic and

Evolutionary Microbiology 61: 574-579.

Chen, W.X., Li, G.S., Qi, Y.L., Wang, E.T., Yuan, H.L. & Li, J.L. (1991) Rhizobium huakuii sp. nov.

Isolated from the Root Nodules of Astragalus sinicus. International Journal of Systematic

Bacteriology 41: 275-280.

59

Chen, W.X., Yan, G.H. & Li, J.L. (1988) Numerical Taxonomic Study of Fast-Growing Soybean Rhizobia

and a Proposal that Rhizobium fredii Be Assigned to Sinorhizobium gen. nov. International

Journal of Systematic Bacteriology 38: 392-397.

D'Haeze, W. & Holsters, M. (2002) Nod factor structures, responses, and perception during initiation

of nodule development. Glycobiology 12: 79-105. de Lajudie, P., Willems, A., Nick, G., Moreira, F., Molouba, F., Hoste, B., Torck, U., Neyra, M.,

Collins, M.D., Lindstrom, K., Dreyfus, B. & Gillis, M. (1998) Characterization of tropical tree

rhizobia and description of Mesorhizobium plurifarium sp. nov. International Journal of

Systematic Bacteriology 48 Pt 2: 369-382.

De Meyer, S.E., Tan, H.W., Andrews, M., Heenan, P.B. & Willems, A. (2016) Mesorhizobium

calcicola sp. nov., Mesorhizobium waitakense sp. nov., Mesorhizobium sophorae sp. nov.,

Mesorhizobium newzealandense sp. nov. and Mesorhizobium kowhaii sp. nov. isolated from

Sophora root nodules. International Journal of Systematic and Evolutionary Microbiology 66:

786-795.

De Meyer, S.E., Tan, H.W., Heenan, P.B., Andrews, M. & Willems, A. (2015) Mesorhizobium

waimense sp. nov. isolated from Sophora longicarinata root nodules and Mesorhizobium

cantuariense sp. nov. isolated from Sophora microphylla root nodules. International Journal

of Systematic and Evolutionary Microbiology 65: 3419-3426.

De Meyer, S.E., Van Hoorde, K., Vekeman, B., Braeckman, T. & Willems, A. (2011) Genetic diversity

of rhizobia associated with indigenous legumes in different regions of Flanders (Belgium).

Soil Biology and Biochemistry 43: 2384-2396.

Degefu, T., Wolde-Meskel, E., Liu, B., Cleenwerck, I., Willems, A. & Frostegard, A. (2013)

Mesorhizobium shonense sp. nov., Mesorhizobium hawassense sp. nov. and Mesorhizobium

abyssinicae sp. nov., isolated from root nodules of different agroforestry legume trees.

International Journal of Systematic and Evolutionary Microbiology 63: 1746-1753.

60

Di Giorgio, G., Graziano, D., Ruisi, P., Amato, G. & Giambalvo, D. (2009) Pheno-morphological and

agronomic diversity among Scorpiurus muricatus (Fabaceae) natural populations collected in

Sicily. The Journal of Agricultural Science 147: 411-422.

Dixon, R. & Kahn, D. (2004) Genetic regulation of biological nitrogen fixation. Nature Reviews

Microbiology 2: 621-631.

Downie, J.A. (2010) The roles of extracellular proteins, polysaccharides and signals in the

interactions of rhizobia with legume roots. FEMS Microbiol Reviews 34: 150-170.

Drew, E., Herridge, D.F., Ballard, R., O’Hara, G., Deaker, R., Denton, M., Yates, R.J., Gemell, G.,

Hartley, E., Phillips, L., Seymour, N., Howieson, J. & Ballard, N., (2014) Inoculating Legumes:

A Practical Guide. In. D.o. Agriculture (ed). Kingston, ACT: Grains Research and Development

Corporation, pp. 72.

Edwards, U., Rogall, T., Blocker, H., Emde, M. & Bottger, E.C. (1989) Isolation and direct complete

nucleotide determination of entire genes. Characterization of a gene coding for 16S

ribosomal RNA. Nucleic Acids Research 17: 7843-7853.

Egener, T., Martin, D.E., Sarkar, A. & Reinhold-Hurek, B. (2001) Role of a ferredoxin gene

cotranscribed with the nifHDK operon in N(2) fixation and nitrogenase "switch-off" of

Azoarcus sp. strain BH72. Journal of Bacteriology 183: 3752-3760.

Ehrman, T. & Cocks, P.S. (1990) Ecogeography of Annual Legumes in Syria - Distribution Patterns.

Journal of Applied Ecology 27: 578-591.

Frank, B. (1889) Über die Pilzsymbiose der Leguminosen. Berichte der Deutschen Botanischen

Gesellschaft 7: 332-346.

Fu, G.Y., Yu, X.Y., Zhang, C.Y., Zhao, Z., Wu, D., Su, Y., Wang, R.J., Han, S.B., Wu, M. & Sun, C.

(2017) Mesorhizobium oceanicum sp. nov., isolated from deep seawater. International

Journal of Systematic and Evolutionary Microbiology 67: 2739-2745.

61

Gage, D.J. (2004) Infection and Invasion of Roots by Symbiotic, Nitrogen-Fixing Rhizobia during

Nodulation of Temperate Legumes. Microbiology and Molecular Biology Reviews 68: 280-

300.

Gao, J.L., Turner, S.L., Kan, F.L., Wang, E.T., Tan, Z.Y., Qiu, Y.H., Gu, J., Terefework, Z., Young, J.P.,

Lindstrom, K. & Chen, W.X. (2004) Mesorhizobium septentrionale sp. nov. and

Mesorhizobium temperatum sp. nov., isolated from Astragalus adsurgens growing in the

northern regions of China. International Journal of Systematic and Evolutionary Microbiology

54: 2003-2012.

Ghosh, W. & Roy, P. (2006) Mesorhizobium thiogangeticum sp. nov., a novel sulfur-oxidizing

chemolithoautotroph from rhizosphere soil of an Indian tropical leguminous plant.

International Journal of Systematic and Evolutionary Microbiology 56: 91-97.

Graham, P.H. & Vance, C.P. (2003) Legumes: importance and constraints to greater use. Plant

Physiology 131: 872-877.

Guan, S.H., Chen, W.F., Wang, E.T., Lu, Y.L., Yan, X.R., Zhang, X.X. & Chen, W.X. (2008)

Mesorhizobium caraganae sp. nov., a novel rhizobial species nodulated with Caragana spp.

in China. International Journal of Systematic and Evolutionary Microbiology 58: 2646-2653.

Han, T.X., Han, L.L., Wu, L.J., Chen, W.F., Sui, X.H., Gu, J.G., Wang, E.T. & Chen, W.X. (2008)

Mesorhizobium gobiense sp. nov. and Mesorhizobium tarimense sp. nov., isolated from wild

legumes growing in desert soils of Xinjiang, China. International Journal of Systematic and

Evolutionary Microbiology 58: 2610-2618.

Haskett, T.L., Terpolilli, J.J., Bekuma, A., O'Hara, G.W., Sullivan, J.T., Wang, P., Ronson, C.W. &

Ramsay, J.P. (2016) Assembly and transfer of tripartite integrative and conjugative genetic

elements. Proceedings of the National Academy of Sciences of the United States of America

113: 12268-12273.

Heiser, C.B., (1981) Seed to civilization: the story of food, p. 254. W. H. Freeman, San Francisco.

62

Herridge, D.F., Peoples, M.B. & Boddey, R.M. (2008) Global inputs of biological nitrogen fixation in

agricultural systems. Plant and Soil 311: 1-18.

Heyn, C.C. & Raviv, V. (1966) Experimental Taxonomic Studies in the Genus Scorpiurus

(Papilionaceae). Bulletin of the Torrey Botanical Club 93: 259-267.

Howieson, J. & Ballard, R. (2004) Optimising the legume symbiosis in stressful and competitive

environments within southern Australia - some contemporary thoughts. Soil Biology &

Biochemistry 36: 1261-1273.

Howieson, J.G., Ewing, M.A. & D'Antuono, M.F. (1988) Selection for acid tolerance in Rhizobium

meliloti. Plant and Soil 105: 179-188.

Howieson, J.G., Malden, J., Yates, R.J. & O'Hara, G.W. (2000a) Techniques for the selection and

development of elite inoculant strains of Rhizobium leguminosarum in southern Australia.

Symbiosis 28: 33-48.

Howieson, J.G., Nutt, B. & Evans, P. (2000b) Estimation of host-strain compatibility for symbiotic N-

fixation between Rhizobium meliloti, several annual species of Medicago and Medicago

sativa. Plant and Soil 219: 49-55.

Howieson, J.G., O’Hara, G.W. & Carr, S.J. (2000c) Changing roles for legumes in Mediterranean

agriculture: developments from an Australian perspective. Field Crops Research 65: 107-122.

Hungria, M., O’Hara, G.W., Zilli, J.E., Araujo, R.S., Deaker, R. & Howieson, J.G., (2016) Isolation and

growth of rhizobia. In: Working with rhizobia. J.G. Howieson & M.J. Dilworth (eds). Canberra:

Australian Centre for International Agricultural Research, pp. 40-62.

Hwang, J.H., Ellingson, S.R. & Roberts, D.M. (2010) Ammonia permeability of the soybean nodulin

26 channel. FEBS Letters 584: 4339-4343.

Ibanez, F., Wall, L. & Fabra, A. (2017) Starting points in plant-bacteria nitrogen-fixing symbioses:

intercellular invasion of the roots. Journal of Experimental Botany 68: 1905-1918.

Jarvis, B.D.W., Pankhurst, C.E. & Patel, J.J. (1982) Rhizobium-Loti, a New Species of Legume Root

Nodule Bacteria. International Journal of Systematic Bacteriology 32: 378-380.

63

Jarvis, B.D.W., VanBerkum, P., Chen, W.X., Nour, S.M., Fernandez, M.P., CleyetMarel, J.C. & Gillis,

M. (1997) Transfer of Rhizobium loti, Rhizobium huakuii, Rhizobium ciceri, Rhizobium

mediterraneum, and Rhizobium tianshanense to Mesorhizobium gen. nov. International

Journal of Systematic Bacteriology 47: 895-898.

Jordan, D.C. (1982) Transfer of Rhizobium japonicum Buchanan 1980 to Bradyrhizobium gen. nov., a

Genus of Slow-Growing, Root Nodule Bacteria from Leguminous Plants. International Journal

of Systematic Bacteriology 32: 136-139.

Karunakaran, R., Ramachandran, V.K., Seaman, J.C., East, A.K., Mouhsine, B., Mauchline, T.H.,

Prell, J., Skeffington, A. & Poole, P.S. (2009) Transcriptomic analysis of Rhizobium

leguminosarum biovar viciae in symbiosis with host plants Pisum sativum and Vicia cracca.

Journal of Bacteriology 191: 4002-4014.

Laguerre, G., Nour, S.M., Macheret, V., Sanjuan, J., Drouin, P. & Amarger, N. (2001) Classification of

rhizobia based on nodC and nifH gene analysis reveals a close phylogenetic relationship

among Phaseolus vulgaris symbionts. Microbiology 147: 981-993.

Lane, D.J., (1991) 16S/23S rRNA sequencing. In: Nucleic Acid Techniques in Bacterial Systematic. E.

Stackebrandt & M. Goodfellow (eds). New York: John Wiley and Sons, pp. 115-175.

Laranjo, M., Alexandre, A. & Oliveira, S. (2014) Legume growth-promoting rhizobia: an overview on

the Mesorhizobium genus. Microbiological Research 169: 2-17.

LeBauer, D.S. & Treseder, K.K. (2008) Nitrogen Limitation of Net Primary Productivity in Terrestrial

Ecosystems Is Globally Distributed. Ecology 89: 371-379.

Lewis, G., Schrire, B., Mackinder, B. & Lock, M., (2005) Legumes of the world, p. 577. The Royal

Botanic Gardens, Kew, Richmond, UK.

Lewis, G.P., Schrire, B.D., Mackinder, B.A., Rico, L. & Clark, R. (2013) A 2013 linear sequence of

legume genera set in a phylogenetic context — A tool for collections management and taxon

sampling. South African Journal of Botany 89: 76-84.

64

Licitra, G., Carpino, S., Schadt, I., Avondo, M. & Barresi, S. (1997) Forage quality of native pastures

in a Mediterranean area. Animal Feed Science and Technology 69: 315-328.

Loi, A., Howieson, J.G., Nutt, B.J. & Carr, S.J. (2005) A second generation of annual pasture legumes

and their potential for inclusion in Mediterranean-type farming systems. Australian Journal

of Experimental Agriculture 45: 289-299.

Lorite, M.J., Flores-Felix, J.D., Peix, A., Sanjuan, J. & Velazquez, E. (2016) Mesorhizobium olivaresii

sp. nov. isolated from Lotus corniculatus nodules. Systematic and Applied Microbiology 39:

557-561.

LPSN (2019) List of prokaryotic names with standing in nomenclature, Genus Mesorhizobium.

http://www.bacterio.net/mesorhizobium.html. Accessed: 20 September 2019.

LPWG (2013) Legume phylogeny and classification in the 21st century: Progress, prospects and

lessons for other species-rich clades. Taxon 62: 217-248.

Lu, Y.L., Chen, W.F., Wang, E.T., Han, L.L., Zhang, X.X., Chen, W.X. & Han, S.Z. (2009)

Mesorhizobium shangrilense sp. nov., isolated from root nodules of Caragana species.

International Journal of Systematic and Evolutionary Microbiology 59: 3012-3018.

MacKinnon, P.A., Robertson, J.G., Scott, D.J. & Hale, C.N. (1977) Legume inoculant usage in New

Zealand. New Zealand Journal of Experimental Agriculture 5: 35-39.

Marcos-Garcia, M., Menendez, E., Ramirez-Bahena, M.H., Mateos, P.F., Peix, A., Velazquez, E. &

Rivas, R. (2017) Mesorhizobium helmanticense sp. nov., isolated from Lotus corniculatus

nodules. International Journal of Systematic and Evolutionary Microbiology 67: 2301-2305.

Martinez-Hidalgo, P., Ramirez-Bahena, M.H., Flores-Felix, J.D., Igual, J.M., Sanjuan, J., Leon-

Barrios, M., Peix, A. & Velazquez, E. (2016) Reclassification of strains MAFF 303099T and

R7A into Mesorhizobium japonicum sp. nov. International Journal of Systematic and

Evolutionary Microbiology 66: 4936-4941.

Martinez-Hidalgo, P., Ramirez-Bahena, M.H., Flores-Felix, J.D., Rivas, R., Igual, J.M., Mateos, P.F.,

Martinez-Molina, E., Leon-Barrios, M., Peix, A. & Velazquez, E. (2015) Revision of the

65

taxonomic status of type strains of Mesorhizobium loti and reclassification of strain USDA

3471T as the type strain of Mesorhizobium erdmanii sp. nov. and ATCC 33669T as the type

strain of Mesorhizobium jarvisii sp. nov. International Journal of Systematic and Evolutionary

Microbiology 65: 1703-1708.

Mohamad, R., Willems, A., Le Quere, A., Maynaud, G., Pervent, M., Bonabaud, M., Dubois, E.,

Cleyet-Marel, J.C. & Brunel, B. (2017) Mesorhizobium delmotii and Mesorhizobium

prunaredense are two new species containing rhizobial strains within the symbiovar

anthyllidis. Systematic and Applied Microbiology 40: 135-143.

Moulin, L., Munive, A., Dreyfus, B. & Boivin-Masson, C. (2001) Nodulation of legumes by members

of the beta-subclass of Proteobacteria. Nature 411: 948-950.

Muresu, R., Polone, E., Sulas, L., Baldan, B., Tondello, A., Delogu, G., Cappuccinelli, P., Alberghini,

S., Benhizia, Y., Benhizia, H., Benguedouar, A., Mori, B., Calamassi, R., Dazzo, F.B. &

Squartini, A. (2008) Coexistence of predominantly nonculturable rhizobia with diverse,

endophytic bacterial taxa within nodules of wild legumes. FEMS Microbiology Ecology 63:

383-400.

Nandasena, K., O’Hara, G.W., Tiwari, R.P., Yates, R.J., Kishinevsky, B.D. & Howieson, J.G. (2004)

Symbiotic relationships and root nodule ultrastructure of the pasture legume Biserrula

pelecinus L.- a new legume in agriculture. Soil Biology and Biochemistry 36: 1309-1317.

Nandasena, K.G., O'Hara G, W., Tiwari, R.P. & Howieson, J.G. (2006) Rapid in situ evolution of

nodulating strains for Biserrula pelecinus L. through lateral transfer of a symbiosis island

from the original mesorhizobial inoculant. Applied and Environmental Microbiology 72:

7365-7367.

Nandasena, K.G., O'Hara, G.W., Tiwari, R.P., Sezmis, E. & Howieson, J.G. (2007) In situ lateral

transfer of symbiosis islands results in rapid evolution of diverse competitive strains of

mesorhizobia suboptimal in symbiotic nitrogen fixation on the pasture legume Biserrula

pelecinus L. Environmental Microbiology 9: 2496-2511.

66

Nandasena, K.G., O'Hara, G.W., Tiwari, R.P., Willems, A. & Howieson, J.G. (2009) Mesorhizobium

australicum sp. nov. and Mesorhizobium opportunistum sp. nov., isolated from Biserrula

pelecinus L. in Australia. International Journal of Systematic and Evolutionary Microbiology

59: 2140-2147.

NCBI (2019) BLASTN, National Center for Biotechnology Information.

https://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastn&PAGE_TYPE=BlastSearch&LINK_L

OC=blasthome. Accessed: 1 September 2019.

Nguyen, T.M., Pham, V.H. & Kim, J. (2015) Mesorhizobium soli sp. nov., a novel species isolated

from the rhizosphere of Robinia pseudoacacia L. in South Korea by using a modified culture

method. Antonie Van Leeuwenhoek 108: 301-310.

Nichols, P.G.H., Loi, A., Nutt, B.J., Evans, P.M., Craig, A.D., Pengelly, B.C., Dear, B.S., Lloyd, D.L.,

Revell, C.K., Nair, R.M., Ewing, M.A., Howieson, J.G., Auricht, G.A., Howie, J.H., Sandral,

G.A., Carr, S.J., de Koning, C.T., Hackney, B.F., Crocker, G.J., Snowball, R., Hughes, S.J., Hall,

E.J., Foster, K.J., Skinner, P.W., Barbetti, M.J. & You, M.P. (2007) New annual and short-

lived perennial pasture legumes for Australian agriculture—15 years of revolution. Field

Crops Research 104: 10-23.

Nichols, P.G.H., Revell, C.K., Humphries, A.W., Howie, J.H., Hall, E.J., Sandral, G.A., Ghamkhar, K. &

Harris, C.A. (2012) Temperate pasture legumes in Australia—their history, current use, and

future prospects. Crop and Pasture Science 63.

Nour, S.M., Cleyetmarel, J.C., Normand, P. & Fernandez, M.P. (1995) Genomic Heterogeneity of

Strains Nodulating Chickpeas (Cicer arietinum L) and Description of Rhizobium

mediterraneum sp. nov. International Journal of Systematic Bacteriology 45: 640-648.

Nour, S.M., Fernandez, M.P., Normand, P. & Cleyet-Marel, J.C. (1994) Rhizobium ciceri sp. nov.,

consisting of strains that nodulate chickpeas (Cicer arietinum L.). International Journal of

Systematic Bacteriology 44: 511-522.

67

Nutt, B.J., (2012) Incidence and inheritance of hard-seededness and early maturity in Ornithopus

sativus. In: School of Biological Sciences and Biotechnology. Murdoch University, pp. 274.

O'Hara, G.W., Hungria, M., Woomer, P. & Howieson, J.G., (2016a) Counting rhizobia. In: Working

with rhizobia. J.G. Howieson & M.J. Dilworth (eds). Canberra: Australian Centre for

International Agricultural Research, pp. 109-124.

O'Hara, G.W., Zilli, J.E., Poole, P.S. & Hungria, M., (2016b) and physiology of rhizobia. In:

Working with rhizobia. J.G. Howieson & M.J. Dilworth (eds). Canberra: Australian Centre for

International Agricultural Research, pp. 125-145.

Oldroyd, G.E. & Downie, J.A. (2004) Calcium, kinases and nodulation signalling in legumes. Nature

Reviews Molecular Cell Biology 5: 566-576.

Perret, X., Staehelin, C. & Broughton, W.J. (2000) Molecular basis of symbiotic promiscuity.

Microbiology and Molecular Biology Reviews 64: 180-201.

Peters, N.K. & Verma, D.P.S. (1990) Phenolic-Compounds as Regulators of Gene-Expression in Plant-

Microbe Interactions. Molecular Plant-Microbe Interactions 3: 4-8.

Poole, P., Ramachandran, V. & Terpolilli, J. (2018) Rhizobia: from saprophytes to endosymbionts.

Nature Reviews Microbiology 16: 291-303.

Pülschen, L. (1992) Effects of Two Underseed Species, Medicago polymorpha L. and Scorpiurus

muricatus L., on the Yield of Main Crop (Durum Wheat) and Subsequent Crop (Teff) Under

Humid Moisture Regimes in Ethiopia. Journal of Agronomy and Crop Science 168: 249-254.

Ramirez-Bahena, M.H., Hernandez, M., Peix, A., Velazquez, E. & Leon-Barrios, M. (2012)

Mesorhizobial strains nodulating Anagyris latifolia and Lotus berthelotii in Tamadaya ravine

(Tenerife, Canary Islands) are two symbiovars of the same species, Mesorhizobium

tamadayense sp. nov. Systematic and Applied Microbiology 35: 334-341.

Ramsay, J.P., Sullivan, J.T., Stuart, G.S., Lamont, I.L. & Ronson, C.W. (2006) Excision and transfer of

the Mesorhizobium loti R7A symbiosis island requires an integrase IntS, a novel

68

recombination directionality factor RdfS, and a putative relaxase RlxS. Molecular

Microbiology 62: 723-734.

Richardson, A.E., Viccars, L.A., Watson, J.M. & Gibson, A.H. (1995) Differentiation of Rhizobium

strains using the polymerase chain reaction with random and directed primers. Soil Biology

and Biochemistry 27: 515-524.

Robertson, G.P. & Vitousek, P.M. (2009) Nitrogen in Agriculture: Balancing the Cost of an Essential

Resource. Annual Review of Environment and Resources 34: 97-125.

Ruisi, P., Amato, G., Frenda, A.S., Giambalvo, D. & Di Miceli, G., (2017) Scorpiurus muricatus L.: an

interesting legume species for Mediterranean forage systems. In: Grassland resources for

extensive farming systems in marginal lands: major drivers and future scenarios. C.

Porqueddu, A. Franca, G. Lombardi, G. Molle, G. Peratoner & A. Hopkins (eds). Alghero, Italy:

Organising Committee of the 19th Symposium of the European Grassland Federation, pp.

394-396.

Safronova, V.I., Piluzza, G., Belimov, A.A. & Bullitta, S. (2004) Phenotypic and genotypic analysis of

rhizobia isolated from pasture legumes native of Sardinia and Asinara Island. Antonie Van

Leeuwenhoek 85: 115-127.

Sannazzaro, A.I., Torres Tejerizo, G., Fontana, M.F., Cumpa Velasquez, L.M., Hansen, L.H., Pistorio,

M. & Estrella, M.J. (2018) Mesorhizobium sanjuanii sp. nov., isolated from nodules of Lotus

tenuis in the saline-alkaline lowlands of Flooding Pampa, Argentina. International Journal of

Systematic and Evolutionary Microbiology 68: 2936-2942.

Shimada, N., Aoki, T., Sato, S., Nakamura, Y., Tabata, S. & Ayabe, S. (2003) A cluster of genes

encodes the two types of chalcone isomerase involved in the biosynthesis of general

flavonoids and legume-specific 5-deoxy(iso)flavonoids in Lotus japonicus. Plant Physiology

131: 941-951.

Sprent, J.I., Ardley, J. & James, E.K. (2017) Biogeography of nodulated legumes and their nitrogen-

fixing symbionts. New Phytologist 215: 40-56.

69

Sullivan, J.T., Patrick, H.N., Lowther, W.L., Scott, D.B. & Ronson, C.W. (1995) Nodulating strains of

Rhizobium loti arise through chromosomal symbiotic gene transfer in the environment.

Proceedings of the National Academy of Sciences of the United States of America 92: 8985-

8989.

Sullivan, J.T. & Ronson, C.W. (1998) Evolution of rhizobia by acquisition of a 500-kb symbiosis island

that integrates into a phe-tRNA gene. Proceedings of the National Academy of Sciences of

the United States of America 95: 5145-5149.

Timmers, A.C., Auriac, M.C. & Truchet, G. (1999) Refined analysis of early symbiotic steps of the

Rhizobium-Medicago interaction in relationship with microtubular cytoskeleton

rearrangements. Development 126: 3617-3628.

Udvardi, M. & Poole, P.S. (2013) Transport and metabolism in legume-rhizobia symbioses. Annual

Review of Plant Biology 64: 781-805.

Vandamme, P. & Coenye, T. (2004) Taxonomy of the genus Cupriavidus: a tale of lost and found.

International Journal of Systematic and Evolutionary Microbiology 54: 2285-2289.

Vaneechoutte, M., Claeys, G., Steyaert, S., De Baere, T., Peleman, R. & Verschraegen, G. (2000)

Isolation of Moraxella canis from an ulcerated metastatic lymph node. Journal of Clinical

Microbiology 38: 3870-3871.

Velazquez, E., Igual, J.M., Willems, A., Fernandez, M.P., Munoz, E., Mateos, P.F., Abril, A., Toro, N.,

Normand, P., Cervantes, E., Gillis, M. & Martinez-Molina, E. (2001) Mesorhizobium

chacoense sp. nov., a novel species that nodulates Prosopis alba in the Chaco Arido region

(Argentina). International Journal of Systematic and Evolutionary Microbiology 51: 1011-

1021.

Vidal, C., Chantreuil, C., Berge, O., Maure, L., Escarre, J., Bena, G., Brunel, B. & Cleyet-Marel, J.C.

(2009) Mesorhizobium metallidurans sp. nov., a metal-resistant symbiont of Anthyllis

vulneraria growing on metallicolous soil in Languedoc, France. International Journal of

Systematic and Evolutionary Microbiology 59: 850-855.

70

Vincent, J.M., (1970) A manual for the practical study of root-nodule bacteria, p. 164. Published for

the International Biological Programme by Blackwell Scientific, Oxford.

Wang, D., Yang, S., Tang, F. & Zhu, H. (2012) Symbiosis specificity in the legume: rhizobial

mutualism. Cellular Microbiology 14: 334-342.

Wang, E.T., van Berkum, P., Sui, X.H., Beyene, D., Chen, W.X. & Martinez-Romero, E. (1999)

Diversity of rhizobia associated with Amorpha fruticosa isolated from Chinese soils and

description of Mesorhizobium amorphae sp. nov. International Journal of Systematic

Bacteriology 49 Pt 1: 51-65.

Wang, F.Q., Wang, E.T., Liu, J., Chen, Q., Sui, X.H., Chen, W.F. & Chen, W.X. (2007) Mesorhizobium

albiziae sp. nov., a novel bacterium that nodulates Albizia kalkora in a subtropical region of

China. International Journal of Systematic and Evolutionary Microbiology 57: 1192-1199.

Yahara, T., Javadi, F., Onoda, Y., de Queiroz, L.P., Faith, D.P., Prado, D.E., Akasaka, M., Kadoya, T.,

Ishihama, F., Davies, S., Slik, J.W.F., Yi, T.S., Ma, K.P., Bin, C., Darnaedi, D., Pennington,

R.T., Tuda, M., Shimada, M., Ito, M., Egan, A.N., Buerki, S., Raes, N., Kajita, T., Vatanparast,

M., Mimura, M., Tachida, H., Iwasa, Y., Smith, G.F., Victor, J.E. & Nkonki, T. (2013) Global

legume diversity assessment: Concepts, key indicators, and strategies. Taxon 62: 249-266.

Yates, R.J., Howieson, J.G., Hungria, M., Bala, A., O’Hara, G.W. & Terpolilli, J., (2016)

Authentication of rhizobia and assessment of the legume symbiosis in controlled plant

growth systems. In: Working with rhizobia. J.G. Howieson & M.J. Dilworth (eds). Canberra:

Australian Centre for International Agricultural Research, pp. 73-110.

Yates, R.J., Howieson, J.G., Real, D., Reeve, W.G., Vivas-Marfisi, A. & O'Hara, G.W. (2005) Evidence

of selection for effective nodulation in the Trifolium spp. symbiosis with Rhizobium

leguminosarum biovar trifolii. Australian Journal of Experimental Agriculture 45: 189-198.

Yuan, C.G., Jiang, Z., Xiao, M., Zhou, E.M., Kim, C.J., Hozzein, W.N., Park, D.J., Zhi, X.Y. & Li, W.J.

(2016) Mesorhizobium sediminum sp. nov., isolated from deep-sea sediment. International

Journal of Systematic and Evolutionary Microbiology 66: 4797-4802.

71

Zhang, J., Guo, C., Chen, W., de Lajudie, P., Zhang, Z., Shang, Y. & Wang, E.T. (2018) Mesorhizobium

wenxiniae sp. nov., isolated from chickpea (Cicer arietinum L.) in China. International Journal

of Systematic and Evolutionary Microbiology 68: 1930-1936.

Zhang, J.J., Liu, T.Y., Chen, W.F., Wang, E.T., Sui, X.H., Zhang, X.X., Li, Y., Li, Y. & Chen, W.X. (2012)

Mesorhizobium muleiense sp. nov., nodulating with Cicer arietinum L. International Journal

of Systematic and Evolutionary Microbiology 62: 2737-2742.

Zhao, C.T., Wang, E.T., Zhang, Y.M., Chen, W.F., Sui, X.H., Chen, W.X., Liu, H.C. & Zhang, X.X. (2012)

Mesorhizobium silamurunense sp. nov., isolated from root nodules of Astragalus species.

International Journal of Systematic and Evolutionary Microbiology 62: 2180-2186.

Zheng, W.T., Li, Y., Jr., Wang, R., Sui, X.H., Zhang, X.X., Zhang, J.J., Wang, E.T. & Chen, W.X. (2013)

Mesorhizobium qingshengii sp. nov., isolated from effective nodules of Astragalus sinicus.

International Journal of Systematic and Evolutionary Microbiology 63: 2002-2007.

Zhou, P.F., Chen, W.M. & Wei, G.H. (2010) Mesorhizobium robiniae sp. nov., isolated from root

nodules of Robinia pseudoacacia. International Journal of Systematic and Evolutionary

Microbiology 60: 2552-2556.

Zhou, S., Li, Q., Jiang, H., Lindstrom, K. & Zhang, X. (2013) Mesorhizobium sangaii sp. nov., isolated

from the root nodules of Astragalus luteolus and Astragalus ernestii. International Journal of

Systematic and Evolutionary Microbiology 63: 2794-2799.

Zhu, Y.J., Kun, J., Chen, Y.L., Wang, S.K., Sui, X.H. & Kang, L.H. (2015) Mesorhizobium acaciae sp.

nov., isolated from root nodules of Acacia melanoxylon R. Br. International Journal of

Systematic and Evolutionary Microbiology 65: 3558-3563.

72

6.0 APPENDIX

40 35 30 25 20 15 10 5

Mean nodules/plant Mean 0 1386 D 1343 T N-starved A J

Treatment Appendix Figure 1: Mean nodule number/plant on inoculated and N-starved S. muricatus plants. Plants were grown in N-limited conditions and either uninoculated or inoculated separately with six strains. The uninoculated N-starved plants received no added N. Plants harvested 47 days post inoculation. Error bars show standard error.

25

20

15

10

5

Mean nodules/plants Mean 0 1386 D T 1343 A J Strain

Appendix Figure 2: Mean nodule number/plant on inoculated L. corniculatus plants. Plants were grown in N- limited conditions and inoculated separately with six strains. Plants harvested 48 days post inoculation. Error bars show standard error.

60 50 40 30 20 10 0

Mean nodules/plant Mean T N-fed D 1386 A N-starved J 1343

Treatment

Appendix Figure 3: Mean nodule number/plant on B. pelecinus. Plants were grown in N-limited conditions and either uninoculated or inoculated separately with six strains. The uninoculated N-starved plants received no added N while N-fed plants were supplied nitrogen as KNO3. Plants harvested 48 days post inoculation. Error bars show standard error.

73

Appendix Table 1: List of experimental strains with closest species matches using BLASTN of full 16S rRNA consensus sequences (NCBI, 2019). Consensus length, closest species matches, query cover %, percent identity, and country of isolation and soil batch ID.

STRAIN CONSENSUS CLOSEST MATCH Query Percent Country of LENGTH (bp) cover % identity isolation and soil batch ID WSM1184 1413 Agrobacterium tumefaciens 100 99.86 Morocco WSM1343 1394 Mesorhizobium ciceri 100 99.86 Morocco Mesorhizobium loti 100 99.86 Mesorhizobium sp. WSM1497 100 99.86 Mesorhizobium ciceri biovar 100 99.86 biserrulae Mesorhizobium amorphae 99 99.71 WSM1386 1407 Mesorhizobium japonicum 100 99.93 Manjimup Mesorhizobium erdmanii 100 99.93 Mesorhizobium loti 100 99.93 Mesorhizobium opportunistum 100 99.86 Mesorhizobium huakuii 100 99.72 A 1416 Mesorhizobium delmotii 100 99.93 Israel Mesorhizobium muleiense 100 99.86 415-419 Mesorhizobium prunaredense 99 99.93 Mesorhizobium tianshanense 100 99.72 Mesorhizobium mediterraneum 99 99.79 B 1404 Mesorhizobium delmotii 100 100 Israel Mesorhizobium prunaredense 100 100 415-419 Mesorhizobium mediterraneum 100 99.79 Mesorhizobium muleiense 100 99.79 Mesorhizobium tianshanense 100 99.72 C 1417 Mesorhizobium cicero 100 99.93 Israel Mesorhizobium ciceri biovar 100 99.93 425-429 biserrulae WSM1271 Mesorhizobium loti 100 99.79 Mesorhizobium australicum 100 99.51 WSM2073 Mesorhizobium amorphae 99 99.65 D 1393 Mesorhizobium cicero 100 99.93 Israel Mesorhizobium ciceri biovar 100 99.93 425-429 biserrulae WSM1271 Mesorhizobium loti 100 99.78 Mesorhizobium amorphae 99 99.64 F 1416 Mesorhizobium delmotii 100 99.86 Israel Mesorhizobium tianshanense 100 99.86 435-439 Mesorhizobium mediterraneum 99 99.93 Mesorhizobium prunaredense 99 99.86 Mesorhizobium muleiense 100 99.72 G 1391 Mesorhizobium delmotii 100 99.93 Israel Mesorhizobium prunaredense 100 99.93 440-444 Mesorhizobium muleiense 100 99.86 Mesorhizobium mediterraneum 99 99.78 Mesorhizobium tianshanense 100 99.57

74

H 1409 Mesorhizobium delmotii 100 99.93 Israel Mesorhizobium prunaredense 99 99.93 445-449 Mesorhizobium muleiense 100 99.86 Mesorhizobium mediterraneum 99 99.79 Mesorhizobium tianshanense 100 99.72 I 1396 Mesorhizobium delmotii 100 100 Israel Mesorhizobium prunaredense 99 100 410-414 Mesorhizobium muleiense 100 99.79 Mesorhizobium tianshanense 100 99.71 Mesorhizobium mediterraneum 99 99.78 J 1411 Mesorhizobium delmotii 100 99.93 Morocco Mesorhizobium muleiense 100 99.86 255-258 Mesorhizobium prunaredense 99 99.93 Mesorhizobium tianshanense 100 99.72 Mesorhizobium mediterraneum 99 99.79 K 1404 Mesorhizobium delmotii 100 99.93 Israel Mesorhizobium prunaredense 100 99.93 410-414 Mesorhizobium muleiense 100 99.86 Mesorhizobium mediterraneum 100 99.79 Mesorhizobium tianshanense 100 99.72 L 1414 Mesorhizobium delmotii 100 99.93 Israel Mesorhizobium prunaredense 99 99.93 420-424 Mesorhizobium muleiense 100 99.86 Mesorhizobium mediterraneum 99 99.79 Mesorhizobium tianshanense 100 99.72 M 1395 Mesorhizobium delmotii 100 99.93 Israel Mesorhizobium muleiense 100 99.86 435-439 Mesorhizobium prunaredense 99 99.93 Mesorhizobium tianshanense 100 99.71 Mesorhizobium mediterraneum 99 99.78 N 1416 Mesorhizobium tianshanense 100 99.93 Morocco Mesorhizobium helmanticense 100 99.79 271-273 Mesorhizobium metallidurans 99 99.86 Mesorhizobium tamadayense 100 99.72 Mesorhizobium tarimense 99 99.86 O 1411 Mesorhizobium delmotii 100 99.93 Israel Mesorhizobium muleiense 100 99.86 420-424 Mesorhizobium prunaredense 99 99.93 Mesorhizobium tianshanense 100 99.72 Mesorhizobium mediterraneum 99 99.79 S 1399 Mesorhizobium delmotii 100 99.93 Israel Mesorhizobium muleiense 100 99.86 445-449 Mesorhizobium prunaredense 99 99.93 Mesorhizobium tianshanense 100 99.71 Mesorhizobium mediterraneum 99 99.78 T 1416 Mesorhizobium tianshanense 100 99.93 Morocco Mesorhizobium helmanticense 100 99.79 271-273 Mesorhizobium metallidurans 99 99.86 Mesorhizobium tamadayense 100 99.72 Mesorhizobium tarimense 99 99.86

75

Appendix Table 2: Percentage of S. muricatus plants nodulated when inoculated with 19 experimental strains in two effectiveness experiments. Plants were grown in N-limited conditions and inoculated separately with 19 strains. Experiment 1 plants inoculated on 27 March and harvested 69 days post-inoculation. Experiment 2 plants inoculated 25 April and harvested 76 days post-inoculation.

* One plant inoculated with strain WSM1184 formed nodules at the bottom of the pot, consistent with contamination.

** Plants inoculated with strain WSM were 100% nodulated in both experiments. STRAIN % plants Experiment nodulated WSM1184 12.5* 1 WSM1343 100 1 WSM1386 100** 1,2 A 100 1 B 87.5 2 C 100 2 D 100 1 F 100 2 G 100 1 H 100 1 I 100 1 J 75 2 K 100 1 L 100 2 M 100 2 N 100 2 O 100 1 S 100 1 T 100 2

76