Diversity of Silverleaf Nightshade in Australia and Implications For

Diversity of silverleaf nightshade in Australia and

implications for management

Xiaocheng Zhu

B.Sc., Fujian Agriculture and Forestry University, China

M.Sc. (by research), Fujian Agriculture and Forestry University, China

A thesis presented to Charles Sturt University in fulfilment of the

requirements for the degree of Doctor of Philosophy

Faculty of Science

Charles Sturt University

Wagga Wagga, NSW, Australia

February 2013

TABLE OF CONTENTS

Table of Contents ...... i

Certificate of Authorship ...... iii

Acknowledgements ...... iv

List of Publications ...... vi

Abstract ...... viii

Chapter 1 General introduction ...... 1

Chapter 2 Literature review ...... 9

2.1 Importance and distribution ...... 9

2.2 Management ...... 13

2.2.1 Chemical control ...... 13

2.2.2 Biocontrol ...... 15

2.2.3 Other management strategies ...... 16

2.3 Biology and ecology ...... 17

2.3.1 Description ...... 17

2.3.2 Lifecycle and genetic diversity ...... 22

2.3.3 Germination and emergence...... 26

2.3.4 Spread ...... 28

i Chapter 3 Morphological variation of silverleaf nightshade (Solanum elaeagnifolium Cav.) in south-eastern Australia ...... 31

Chapter 4 Evaluation of simple sequence repeat (SSR) markers from

Solanum crop species for Solanum elaeagnifolium ...... 43

Chapter 5 Development of SSR markers for genetic analysis of silverleaf nightshade (Solanum elaeagnifolium) and related species...... 51

Chapter 6 SSR marker analysis to determine the genetic variation in

Solanum elaeagnifolium in Australia ...... 59

Chapter 7 Genetic variation and structure of Solanum elaeagnifolium in Australia analysed by AFLP markers ...... 67

Chapter 8 Identification of silverleaf nightshade using microsatellite markers and microstructure ...... 75

Chapter 9 Time of emergence impacts the growth and reproduction of silverleaf nightshade (Solanum elaeagnifolium Cav.) ...... 83

Chapter 10 General discussion and conclusions ...... 91

Appendix ...... 104

References ...... 109

ii CERTIFICATE OF AUTHORSHIP

I, Xiaocheng Zhu, hereby declare that this submission is my own work and to the best of my knowledge and belief, understand that it contains no material previously published or written by another person, nor material which to a substantial extent has been accepted for the award of any other degree or diploma at Charles Sturt University or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by colleagues with whom I have worked at Charles Sturt University or elsewhere during my candidature is fully acknowledged.

I agree that this thesis be accessible for the purpose of study and research in accordance with normal conditions established by the Executive

Director, Library Services, Charles Sturt University or nominee, for the care, loan and reproduction of thesis, subject to confidentiality provisions as approved by the University.

Signature

8th February 2013

iii ACKNOWLEDGEMENTS

I would like to thank to my supervisory team: principle supervisor:

Professor Deirdre Lemerle and co-supervisors: Drs Hanwen Wu, Harsh

Raman, Geoffrey E. Burrows and Rex Stanton for their support, guidance and understanding through my PhD. Dr Rex Stanton also provided some great pictures of silverleaf nightshade for this thesis.

Great appreciation also goes to Drs Neil Coombes, De Li Liu,

Guangdi Li – NSW Department of Primary Industries and Mr Craig Poynter of Spatial Data Analysis Network (SPAN), Charles Sturt University and the

National Climate Centre, Bureau of Meteorology, for their help on my data analysis.

I thank Mr John Garvie of the Natural Resources Management

Board, Mr Robert Thompson of the NSW Department of Primary Industries and a number of state, and local organisations across Australia for their assistance in field sampling.

Thanks to Dr Rosy Raman, Ms Belinda Taylor, Dr Ata Rehman, Dr

Ben Stodart, Dr John Harper, Ms Natalie Allison, Dr Bree Wilson and Mr

Joe Moore for their assistance in the laboratory.

Thanks to Ms Kamala Anggamuthu and Ms Kim Maree Kendell for their support and help.

Thanks to my wife Mrs Meifang Liu for data input and proofreading, and also for her love and tolerance. Thanks to my parents Mr Yulin Zhu and

iv Mrs Lijun Chen, and my daughters Banruo Zhu and Chenxi Zhu for their support and understanding throughout my entire study.

Thanks also to Charles Sturt University for providing a scholarship and funding for my research. The EH Graham Centre and the Council of

Australasian Weed Societies Inc are acknowledged for their travel support to attend academic conferences.

v LIST OF PUBLICATIONS

Journal papers

Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., Stanton, R., & Burrows, G.

E. (2012). Evaluation of simple sequence repeat (SSR) markers from

Solanum crop species for Solanum elaeagnifolium. Weed Research, 52(3),

217-223. doi: 10.1111/j.1365-3180.2012.00908.x.

Zhu, X., Raman, H., Wu, H., Lemerle, D., Burrows, G., & Stanton, R.

(2013). Development of SSR markers for genetic analysis of silverleaf nightshade (Solanum elaeagnifolium) and related species. Plant Molecular

Biology Reporter, 31(1), 248-254. doi: 10.1007/s11105-012-0473-z.

Zhu, X. C., Wu, H. W., Stanton, R., Burrows, G. E., Lemerle, D., & Raman,

H. (2013). Morphological variation of Solanum elaeagnifolium in south- eastern Australia. Weed Research, doi: 10.1111/wre.12032.

Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., Stanton, R., & Burrows,

G. E. (2013). Genetic variation and structure of Solanum elaeagnifolium in

Australia analysed by amplified fragment length polymorphism markers.

Weed Research, doi: 10.1111/wre.12029.

Zhu, X. C., Wu, H. W., Stanton, R., Raman, H., Lemerle, D., & Burrows, G.

E. (2013). Time of emergence impacts the growth and reproduction of silverleaf nightshade (Solanum elaeagnifolium Cav.). Weed Biology and

Management. doi:10.1111/wbm.12015.

vi Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., Stanton, R., & Burrows, G.

E. (2013). SSR marker analysis to determine the genetic variation in

Solanum elaeagnifolium in Australia. Plant Protection Quarterly (accepted).

Conference proceedings

Zhu, X. C., Burrows, G., Wu, H., Raman, H., Stanton, R., & Lemerle, D.

(2011). Identification of silverleaf nightshade using microsatellite markers and microstructure. Proceedings of the 23rd Asian-Pacific Weed Science

Society Conference, Cairns, pp. 604-609.

Zhu, X. C., Wu, H., Stanton, R., Raman, H., Lemerle, D., & Burrows, G.

(2012). The impact of emergence time on silverleaf nightshade (Solanum elaeagnifolium) development and growth. Proceedings of the 18th

Australasian Weeds Conference, Melbourne, pp. 329-332.

Co-authored journal paper

Burrows G.E., White R.G., Harper J.D.I., Heady R.D., Stanton R.A., Zhu

X., Wu H., Lemerle D. (2013) Entry of fluorescent tracers into leaf trichomes of silverleaf nightshade (Solanum elaeagnifolium Cav.).

American Journal of Botany. (requested revision).

vii ABSTRACT

Silverleaf nightshade (Solanum elaeagnifolium) or SLN is a Weed of

National Significance in Australia, mainly infesting the southern cereal cropping zone and with the potential to infest 400 million hectares. SLN management relies on herbicides, but efficacy is influenced by factors such as plant morphology, genetic background, growth stage and environmental conditions. This study investigated the extent and cause of morphological and genetic variation, the distribution of different phenotypes and genotypes, and how the modes of reproduction contribute to the adaptability of SLN in

Australia.

SLN is a herbaceous, perennial weed that reproduces both sexually and asexually. In Australia, plants germinate from the soil seedbank and rootbank from September (spring) to April (autumn), fruits normally form in

January, and the aerial growth senesces in May.

High morphological variation was found between 642 SLN individuals from south-eastern Australia. Leaf length, width and area ranged from 1.44 to 10.6 cm, 0.39 to 4.09 cm, and 0.41 to 25.8 cm2, respectively.

High trichome densities were found on both leaf surfaces (67 and 132 trichomes/mm2 on the adaxial and abaxial surface, respectively). Larger leaves usually have lower adaxial trichome densities than smaller leaves. On average, there were 603 and 814 stomata/mm2 on the adaxial (ranged from

284 to 942 stomata/mm2) and abaxial (ranged from 455 to 1519 stomata/mm2) surfaces, respectively. These morphological variations may influence foliar herbicide coverage, retention and uptake.

viii Nineteen SSR primer-pairs and four AFLP primer combinations detected a high level of genetic diversity of 94 SLN populations, with average Jaccard’s coefficient for SSR and AFLP analysis at 0.43 and 0.26, respectively. High intra- and inter-population genetic diversity was found, suggesting SLN has capacity to adapt and persist under a range of management systems and environments.

This study also improved the differentiation of SLN from the morphologically similar native perennial Solanum species S. esuriale

(quena). Misidentification can result in delay of management as S. esuriale is not a problem unless high densities occur. Compared to S. esuriale, SLN has a significantly longer trichome intrusive base structure on the adaxial leaf surface. In addition, five and seven species-specific SSR bands (alleles) were identified for SLN and S. esuriale, respectively.

Distribution of different SLN phenotypes and genotypes were correlated with abiotic factors. Individuals that were taller and had a larger leaf area were distributed in areas where higher rainfall was received during the growing season (2009-2010), indicating the likely impact of rainfall on

SLN morphology. In addition, the Bayesian model-based genetic structure analysis assigned Australian SLN into two gene pools. The spatial distribution of these two gene pools correlated well with the early records

(1901 and 1918) of SLN in Australia, which may suggest the possible sites of first establishment.

The reproductive strategies of SLN can help to explain why it covers such a broad area in Australia. Plants that emerged in spring produced

ix significantly more seeds and root biomass than plants that emerged in the late summer or early autumn, especially for root-generated individuals, thereby resulting in the replenishment of the soil seedbank and rootbank, and contributing to establishment. In addition, accidental transportation of seed in contaminated agricultural products will promote expansion of the species, lead to gene flow within and between populations and contribute to the genetic diversity and adaptability. By contrast, those plants that emerged later (plants from seeds later than January and plants from root later than

March) only produced very limited root biomass (dry weight < 0.1 g per plant) before winter. It is not clear whether such a small amount of root biomass is sufficient for the plant to survive over winter and reproduce in subsequent years.

The results of this PhD study provide an insight into the morphological and genetic diversity of SLN in Australia. The research also improves SLN identification and enhances an understanding on the adaptability of SLN. These findings will assist in designing effective management strategies.

CHAPTER 1 GENERAL INTRODUCTION

Silverleaf nightshade (Solanum elaeagnifolium Cav. Solanaceae) or

SLN is a worldwide significant weed introduced to Australia from south- western United States or northern Mexico in 1901 (Cuthbertson et al., 1976).

It competes with pastures and crops and leads to considerable yield loss (up to 75% in cotton (Abernathy & Keeling, 1979) and 69% in wheat (Lemerle

& Leys, 1991)). SLN grows in the temperate regions in Australia with annual rainfall around 250 - 600 mm. Currently it infests at least 350,000 hectares in Australia (Feuerherdt, 2009), with potential to infest up to 398 million hectares (Kwong, 2006).

SLN reproduces sexually by seeds and asexually by root fragments, which mainly contribute to long and short distance distribution, respectively. Seed production can be as high as 247 million seeds per hectare in USA (Cooley & Smith, 1972) and seeds are viable for at least six years in Australia (Stanton et al., 2009b). Regeneration can occur from the extensive root system; a 5-cm long root fragment can survive for 15 months under moist conditions (Fernandez & Brevedan, 1972).

The dry summer in Australia may restricts the success of the biocontrol program of SLN (Kwong, 2006), while the high regenerative ability of the extensive root system limited the efficacy of mechanical management (Stanton et al., 2011), thus herbicide control is the only useful management strategy in Australia. Even so, there are problems due to costs, herbicide residues in the soil, and unreliable performance (Stanton et al.,

1 2009a). Efficacy of herbicides is affected by plant morphology (Brewer et al., 1991) which is influenced by genetic diversity (Marshall & Moss, 2008), growth stage (Greenfield, 2003) and environmental factors. However, such information on SLN is not available for Australian populations. An understanding of the variation in SLN morphology, and genetic diversity is essential to improve efficacy of management strategies.

Morphological variation has been previously reported within

Australian SLN based on limited herbarium samples (Bean, 2004; Symon,

1981). Morphological variation such as leaf size and trichome density influences the retention and contact area of herbicide droplets, and impacts chemical control (Kraemer et al., 2009). SLN is widely distributed over a large area in Australia covering different climate zones (Kwong, 2006), thus a large scale morphological study is required to understand SLN variation and adaptability across this range, and development of effective management strategies.

Such morphological variations may be caused by genetic diversity in

SLN. Genetic diversity is a predominant factor contributing to weed success

(Dekker, 1997). It may result in new phenotypes and contribute to weed resilience. High level of genetic diversity within and between populations of

SLN was reported in South Australia (SA) (Hawker et al., 2006). However genetic diversity of SLN growing across Australia is still unknown.

SLN can be misidentified as an Australian native Solanum species S. esuriale Lindl. (quena). Solanum esuriale is a non-invasive species and is generally not an agricultural problem unless it occurs at high densities

(Johnson et al., 2006). Misidentification caused the delay of management of

SLN in SA (Hosking et al., 2000). The understanding of morphological and genetic diversity will improve the identification of SLN.

Information on the morphology and genetics of SLN will also improve the understanding of distribution of different SLN phenotypes and genotypes, to help predict future spread of the species. Abiotic factors such as rainfall influence morphology in many species thus impacting management efficacy (Steptoe et al., 2006). In addition, SLN was first recorded at Bingara, NSW in 1901 followed by rapidly subsequent records in Melbourne (1909) and Hopetoun (1918), VIC, which suggests that multiple introductions of SLN may have happened in Australia (Cuthbertson et al., 1976). However, it is still unknown whether different introductions are genetically diverse and whether the distribution of different genotypes correlates with distinct introduction events of SLN in Australia.

Besides morphology and genetics, study on the species reproductive strategies is required to understand how and why SLN could adapt to such a broad range of environments in Australia.

Three hypotheses will be tested.

1. Levels of morphological and genetic diversity of SLN have contributed to its expansion in Australia;

• SLN is highly morphologically diverse in Australia;

• SLN is highly genetically diverse in Australia;

3 • Morphological and genetic tools can improve the

identification of SLN in Australia.

2. The spatial distribution of SLN phenotypes and genotypes is correlated to certain abiotic factors;

• Phenotypes are correlated with climate conditions such

as rainfall;

• Genotype distribution is related to introductions to

Australia.

3. The reproductive strategies of SLN have enabled it to establish and spread in Australia.

The thesis structure is listed below. It consists of a general introduction, literature review, a series of papers, and finally a general discussion and conclusions.

Chapter 1 General introduction:

This chapter provides the background information and hypotheses of this study;

Chapter 2 Literature review:

This chapter provided an overview of:

• The importance and distribution of SLN;

4 • The current management strategies;

• The biology and ecology of SLN.

Chapter 3 Research paper (published):

This chapter is a large scale study of SLN morphological variation and highlights the potential correlation between rainfall and phenotypes.

The highly morphological variations detected in this chapter suggested the requirement of genetic diversity study. The following chapters (4 and 5) therefore aim to develop genetic markers for investigating the genetic diversity in SLN.

Zhu, X. C., Wu, H. W., Stanton, R., Burrows, G. E., Lemerle, D., &

Raman, H. (2013). Morphological variation of silverleaf nightshade

(Solanum elaeagnifolium Cav.) in south-eastern Australia. Weed Research. doi: 10.1111/wre.12032.

Chapter 4 Research paper (published):

Thirteen cross-species SSR primer pairs for SLN were developed in this chapter while the Chapter 5 attempted to find some SLN-specific SSR markers.

Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., Stanton, R., &

Burrows, G. E. (2012). Evaluation of simple sequence repeat (SSR) markers from Solanum crop species for Solanum elaeagnifolium. Weed Research,

52(3), 217-223. doi: 10.1111/j.1365-3180.2012.00908.x.

Chapter 5 Research paper (published):

5 A total of 23 SLN-specific SSR markers were developed from publicly available nucleotide and expressed sequence tags (ESTs) databases of SLN in this study. In the following chapter, the highly polymorphic SSR primer pairs were selected to study the genetic diversity of SLN in Australia.

Zhu, X., Raman, H., Wu, H., Lemerle, D., Burrows, G., & Stanton,

R. (2013). Development of SSR markers for genetic analysis of silverleaf nightshade (Solanum elaeagnifolium) and related species. Plant Molecular

Biology Reporter, 31(1), 248-254. doi: 10.1007/s11105-012-0473-z.

Chapter 6 Research paper (accepted):

This chapter investigates genetic diversity between SLN populations in Australia using SSR markers developed in this PhD study. A subset of individuals was further evaluated using AFLP markers in Chapter 7.

Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., Stanton, R., &

Burrows, G. E. (2013). SSR marker analysis to determine the genetic variation in Solanum elaeagnifolium in Australia. Plant Protection

Quarterly. (accepted).

Chapter 7 Research paper (published):

This genetic structure study highlights the correlation between introduction events of SLN with genetic demes using AFLP markers.

Some of the markers developed in Chapters 4-7, together with some micro-features, were used to improve the identification of SLN (Chapter 8).

6 Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., Stanton, R., &

Burrows, G. E. (2013). Genetic variation and structure of Solanum elaeagnifolium in Australia analysed by AFLP markers. Weed Research. doi: 10.1111/wre.12029.

Chapter 8 Conference proceeding:

The morphological and genetic tools described in this conference paper were used for correct identification of SLN, which successfully separated SLN from morphologically similar species, quena.

The previous chapters highlighted the morphological and genetic variation of SLN in Australia. In the Chapter 9 the reproductive strategies of

SLN was studied. Reproduction could be one of the important factors affecting the morphological and genetic variation (Chapters 3-7).

Zhu, X. C., Burrows, G., Wu, H., Raman, H., Stanton, R., &

Lemerle, D. (2011). Identification of silverleaf nightshade using microsatellite markers and microstructure. Proceedings of the 23rd Asian-

Pacific Weed Science Society Conference, Cairns, Australia, pp. 604-609.

Chapter 9 Research paper (published):

Study of the reproductive strategies of SLN.

Zhu, X. C., Wu, H. W., Stanton, R., Raman, H., Lemerle, D., &

Burrows, G. E. (2013). Time of emergence impacts the growth and reproduction of silverleaf nightshade (Solanum elaeagnifolium Cav.). Weed

Biology and Management. doi:10.1111/wbm.12015.

Chapter 10 General discussion and conclusion:

This chapter discusses the outcome of this PhD study and aspects of further research.

CHAPTER 2 LITERATURE REVIEW

2.1 Importance and distribution

Silverleaf nightshade (Solanum elaeagnifolium Cav.) or SLN is a deep-rooted, summer-growing perennial weed that originated in south- western United States and northern Mexico and is now widely spread in southern Australia (Stanton, et al., 2009a). It is a worldwide significant weed that competes with pastures and crops and leads to considerable yield loss. SLN was first reported in Australia in 1901 near Bingara, New South

Wales (NSW) (Cuthbertson, et al., 1976). It is now listed as a Weed of

National Significance in Australia (Australian Weeds Committee, 2012). It infests at least 350,000 hectares in Australia (Feuerherdt, 2009) and can potentially infest 398 million hectares (Kwong, 2006). It costs $10 million annually for control in SA (Kwong et al., 2008).

SLN competes with pastures and crops, leads to the loss of soil moisture (Green et al., 1988), and causes up to 69%, 65%, 64% and 21% yield reduction in wheat-Australia (Lemerle & Leys, 1991), peanuts-Spain

(Hackett & Murray, 1982), corn-Morocco (Baye & Bouhache, 2007) and cotton-USA (Smith et al., 1990), respectively. Allelopathic potential has been reported for SLN as well. For example, aqueous extract of SLN leaves

(100 g leaves: 1 L water) reduced seed germination by 38% and 97% in cotton and lettuce, respectively (Bothma, 2006). In addition, Mkula (2006) has found that SLN possesses allelopathic activity on the early growth of cotton (Gossypium hirsutum L.).

9 SLN is considered an alternative host of some insects infesting cotton (Esquivel & Esquivel, 2009; Idol & Slosser, 2005), pepper (Tejada &

Reyes, 1986), potato (Tscheulin et al., 2009) and sorghum (Hall & Teetes,

1981). Similarly, viruses, bacteria and nematodes are also found on SLN such as pepper mottle virus (Rodriguez-Alvarado et al., 2002), potato virus

Y (Boukhris-Bouhachem et al., 2007), zebra complex disease of potato

(Wen et al., 2009) and a nematode (Orrina phyllobia) of fig tree (Thorne,

1934).

SLN fruits contain alkaloids that may form neurotoxins such as glycoalkaloids (Boyd et al., 1984). Cattle can suffer moderate poisoning symptoms if they ingest ripe fruits equal to 0.1-0.3% of their body weight

(Dollahite & Allen, 1960). By contrast, sheep and goats are more resistant than cattle.

SLN is now recorded in many areas beyond its native range, including Africa, Asia, Europe, Oceania, South and North America (Mekki,

2007) (Fig. 1). SLN has recently been reported to infest 25,000 hectares in

Syria (Al-Mouemar & Azmeh, 2009) and 20,000 hectares in Kariouan,

Tunisia (Chalghaf et al., 2007).

Fig. 1 Worldwide distribution of SLN (University of Georgia, 2013).

In Australia, SLN infests the inland pasture and cereal cropping areas (Fig. 2) across NSW, SA and Victoria (VIC) (Kwong, 2006; Stanton, et al., 2009a). At least 350,000 hectares are infested (Feuerherdt, 2009), with potential to increase to 398 million hectares (Kwong, 2006). In SA alone, the affected area was estimated at 50,000-60,000 hectares in 1997

(Heap et al., 1997), a fivefold increase since 1974 (McKenzie & Douglas,

1974). Isolated populations also occur in Western Australia (WA) and

Queensland (QLD) and there is also one herbarium sample recorded from

Alice Springs, Northern Territory (NT) in 1978 (Fig. 2). The isolated outbreaks of SLN in NT, QLD and SA (marked with star on Fig. 2) may result from adventitious translocations of the species. Alternatively, they could be the misidentifications of morphologically similar native Solanum species such as S. esuriale (quena) (Bean, 2004), S. coactiliferum (western nightshade) or S. karsenses (menindee nightshade). Morphological and genetic analyses may be required to confirm these records. The reasons for such broad infestation may be attributed to morphological flexibility across different climate zones, modes of reproduction and genetic variations.

11 However, such information is scarce. The potential distribution of SLN in

Australia (Fig. 2) was estimated using software CLMATE and highlighted the regions which have similar climatic conditions with its native area

(Kwong, 2006). This potential region of occurrence may change, as

Australia is predicted to have hotter and wetter summers in future (Porch et al., 2009). In addition, Lemerle & Leys (1991) reported that SLN caused much higher yield loss on wheat in the drought years. Therefore, the lower winter rainfall of Australia in future (Timbal & Jones, 2008) may lead to a greater impact of SLN on crops and pastures.

Fig. 2 Current (Parsons & Cuthbertson, 2001) and potential (Australian

Weeds Committee, 2013) distribution of SLN in Australia.

12 2.2 Management

Although SLN has been present in Australia for more than 100 years, no effective management options have been found for large and dense infestations (Stanton, et al., 2009a). Management strategies for SLN include non-chemical and chemical control. However, chemical control is the most widely used and useful option for SLN control in Australia.

2.2.1 Chemical control

The registered commercial products for SLN management in

Australia are Tordon 75-D® (2,4-D and picloram), Starane Advanced®

(fluroxypyr) and glyphosate (360g/L, various trade names) (Ensbey et al.,

2011). Glyphosate, 2,4-D and fluroxypyr are useful to stop seed set, but have limited impact on the root system, resulting in vegetative regeneration.

Glyphosate is translocated to the roots but its efficacy is strongly influenced by environmental factors, and is not effective in hot, dry and dusty conditions. The picloram-based herbicides can effectively target the root system, stopping re-growth in subsequent years (Rodriguez, 1972), but picloram is expensive and has residual effects that damage the following crops or pastures.

Choudhary and Bordovsky (2006) reported a 89% reduction of SLN density by using glyphosate applications in the early, middle and later season of Roundup Ready cotton. However, such control is temporary, as new plants will regenerate from the root system. Effect of 2,4-D on SLN management is limited. Lemerle and Leys (1991) reported an increased SLN density from 8 to 11.4 shoot/mm2, after four years control with multiple 2,4-

13 D application throughout summer. Herbicide efficacy is often influenced by many factors such as plant morphology (Carvalho et al., 2009), genetic background (Marshall & Moss, 2008) and growth stage (Greenfield, 2003).

Foliar herbicide efficacy can be influenced by leaf characteristics such as trichomes (Brewer, et al., 1991) and stomatal density (Ricotta &

Masiunas, 1992). Larger leaves tend to intercept more herbicide droplets, allowing more volume of retained herbicide available for uptake (Kraemer et al., 2009; Leaper & Holloway, 2000). However, trichomes on the leaf surface can form a water repellent surface (Brewer, et al., 1991) and reduce the contact area between leaf surface and herbicide droplets (Chachalis et al., 2001; Kraemer, et al., 2009). A negative correlation between trichome density and herbicide efficacy has been reported in mustard greens

(Brassica juncea) (Huangfu et al., 2009). Modified seed oil adjuvants can help water droplets (500 µm diameter) penetrate the trichome barrier of peppermint-scented geranium (Pelargonium tomentosum) leaves, adhere to the leaf surface and increase the wetted area from 0 to 6.5 cm2 (Xu et al.,

2011).

Contrary to trichomes, stomatal guard cells can be a positive factor for herbicide uptake (Wang & Liu, 2007). Compared to other leaf cells, guard cells provide an easier pathway for herbicide uptake, due to a thinner cuticle and better connection with subjacent cells (Ricotta & Masiunas,

1992). Ricotta & Masiunas (1992) reported that tomato genotypes with higher stomatal density were more sensitive to acifluorfen herbicides.

Uptake through stomata can also be improved by adjuvants such as silicone surfactants (Wang & Liu, 2007). However, the function of adjuvants

depends on the types and concentration of adjuvants (Xu, et al., 2011) and varies among species (Wang & Liu, 2007).

In addition, the growth stage of SLN also affects herbicidal control.

In one study, around 10-40% of applied glyphosate was absorbed by SLN during anthesis (January), which is useful to stop seed set, while only 0-1% of the absorbed glyphosate penetrated into the root system (Greenfield,

2003). However, up to 80% of the applied glyphosate can be absorbed at the green to yellow berry stage (March) and 70% of the absorbed herbicide can be translocated into the root system (Greenfield, 2003).

2.2.2 Biocontrol

The biocontrol of SLN is unreliable in Australia (Wapshere, 1988).

At least 30 potential biocontrol agents including 25 insects, three mites, one fungus and one nematode have been identified for SLN around the world

(Kwong, 2006). A moth (Frumenta nephelomicta) is host specific to SLN and was released in 1978-1983 in South Africa, but failed to establish due to the impact of native parasitoids (Olckers & Zimmermann, 1991). There are two leaf beetles (Leptinotarsa texana and L. defecta) that have been released in South Africa in 1992 (Olckers & Hulley, 1994). Although both species can infest eggplant and other South Africa native Solanum species, such infestations are minor (Olckers et al., 1995). Leptinotarsa texana can reduce aboveground biomass of SLN by 75% and has successfully established in

South Africa, while L. defecta failed to establish (Hoffmann et al., 1998;

Olckers et al., 1999). A nematode (Orrinia phyllobia) was released in Texas,

15 USA (Keeling & Abernathy, 1985), and achieved more than 80% control of

SLN within four years (Orr, 1981).

Wapshere (1988) evaluated the feasibility of ten potential biocontrol agents in Australia, including the four agents (F. nephelomicta, L. texana, L. defecta and O. phyllobia) mentioned above. Dry summers in eastern

Australia prevent these biocontrol agents establishing. Subsequently, it was concluded that biocontrol of SLN is more likely to succeed in northern

NSW where there is adequate summer rainfall than in the southern temperate zones. Kwong (2006) assessed 30 potential biocontrol agents and highlighted six species (Gargaphia arizonica, F. nephelomicta, F. solanophaga, Frumenta (Sp.A), Symmetrischema ardeola and

Gnorimoschema sp.) with medium potential for biocontrol in Australia. This research group also reported that beside SLN, O. phyllobia also infested 13

Australian native Solanum species and some eggplant cultivars, thus it was not suitable for biocontrol (Field et al., 2009).

2.2.3 Other management strategies

Previous studies reported the potential effect of competition on suppressing SLN infestations using lucerne (Medicago sativa), pastures

(such as Phalaris species) (Tideman, 1960) and smuts finger grass (Digitaria eriantha) (Viljoen & Wassermann, 2004). Competition can be combined with mechanical control (such as cultivation, slashing and chipping): cultivation in early summer followed by growth of competitive sorghum and cotton crops, eradicated SLN in the third year in South Africa

(Davis et al., 1945). However, control required sufficient soil moisture for growth of the competitive crops (Davis, et al., 1945) and is heavily reliant

16 on seasonal and environmental conditions (Viljoen & Wassermann, 2004).

Other non-chemical controls such as burning and slashing can stop SLN seed set but are ineffective for controlling the root system and therefore result in regeneration. Mechanical control may reduce the vigour of the root system and improve the efficacy of chemical control. For example, glyphosate application following by cultivation can reduce SLN infestations by 97% (Cooley & Smith, 1973). Cultivation alone is ineffective as a long- term strategy for control of this perennial weed (Stanton et al., 2011) as it fragments the root system and intensifies the infestation. Nowadays, conservation farming with no- or minimum tillage is widely adopted in

Australia, with reduced soil erosion and increased soil water (Kelly &

Reeder, 2002). The reduced tillage practices will limit the spread of SLN, as the fragmentation of the root system is minimised due to less cultivation.

2.3 Biology and ecology

2.3.1 Description

SLN is an erect, herbaceous, perennial weed, which can grow up to one metre tall (Boyd, et al., 1984). It belongs to the “Leptostemonum” subgenus of Solanum (Levin et al., 2006). The phylogenetic (Martine et al.,

2006) and detailed taxonomy (Nee, 1999; Whalen, 1984) relationships of this subgenus can also be found in other studies. SLN is closely related to horticultural crops such as eggplant (Solanum melongena), potato (Solanum tuberosum) and tomato (Solanum lycopersicum). Tetraploid (2n = 4x = 48) and hexaploid (2n = 6x = 72) SLN were found in Argentina (Scaldaferro et

17 al., 2012), but only diploid (2n = 2x = 24) individuals have been reported from Australia, based on single plant collected in Burra, SA (Randell &

Symon, 1976).

Fruit anatomy and microscopic features were also reported for many

Solanaceae species and are crucial for species identification (Chiarini &

Barboza, 2009). The unripe fruits of SLN are round, smooth, green striped and turn yellow when ripe (Kidston et al., 2007), usually 1 cm in diameter, without stomata and covered by cuticular wedges (Chiarini & Barboza,

2007a, b).

Dense stellate trichomes are found on SLN leaves, with a diameter ranging from 300 to 500 µm (Christodoulakis, et al., 2009). Trichomes of

SLN consist of 10-18 horizontal rays and a central, vertical spine

(Christodoulakis, et al., 2009). Similar stellate trichomes are also found in other Solanum species such as S. esuriale, huaritar (S. mandonis), borrachero (S. umbellatum) (Roe, 1971) and eggplant (S. melongena)

(Seithe, 1979). Trichomes on SLN have a very unusual lignified intrusive base structure which is only found in a few other species (Bothma, 2006;

Christodoulakis, et al., 2009), such as Microlepis oleaefolia

(Melastomataceae). The comparison of the trichome structure between SLN,

Solanum juvenale and S. hieronymi also highlighted the intrusive base structure in SLN (Cosa et al., 1998, 2000). The function of this intrusive structure is still unknown. The presence of dense trichomes could protect plants from high summer temperatures (Perez-Estrada et al., 2000), solar radiation (Jordan et al., 2005) and herbivores (Canosantana & Oyama, 1992;

Levin, 1973).

Stomata are present on both leaf surfaces of SLN (Christodoulakis, et al., 2009). The amphistomatic structure, a characteristic of many xeromorphic plants, increases the CO2 conductance, reduces limitation of total CO2 diffusion into mesophyll cells, increases the photosynthetic capacity and helps SLN adapt to hot and arid climates (Fahn & Cutler,

1992).

Prickles are usually present on the stem, sometimes on petioles and leaves (Boyd, et al., 1984; Symon, 1981). Leaves are entire or shallowly lobed and covered by dense trichomes on both sides, which gives SLN a silvery-white appearance (Boyd, et al., 1984; Symon, 1981). Detailed descriptions of SLN have been provided by Bean (2004), Boyd et al. (1984) and Symon (1981).

Morphological variations have been reported in many Solanaceae species (Anderson et al., 1999; Clausen & Crisci, 1989; Edmonds, 1978;

Kardolus, 1999; Spooner & van den berg, 2001; Van den Berg &

Groendijk-Wilders, 1999). For example, length of the fourth leaf ranged from 1-17 cm and 5-18 cm in Solanum megistacrolobum and S. toralapanum, respectively (Giannattasio & Spooner, 1994).

Morphological plasticity in SLN has also been reported based on limited herbarium samples (Bean, 2004; Encomidou & Yannitsaros, 1975;

Symon, 1981). Plants vary in height and prickle density (Encomidou &

Yannitsaros, 1975). Encomidou & Yannitsaros (1975) reported that prickle of SLN tend to increase under dry condition. By contrast, after observed 83

Australian native Solanum species, Symon (1986) suggested that prickles in

19 Solanum are not a response to dry environment but to avoiding damage by marsupials. Leaf size of SLN ranges from 3 to 6.5 cm in length and 0.8 to

1.4 cm in width (Bean, 2004). Leaf colour ranges from silvery to pale yellowish-green, which indicates variations in trichome density. Several researchers have reported the presence of stellate trichomes (Bean, 2004;

Bothma, 2006; Christodoulakis et al., 2009) and stomata (Christodoulakis, et al., 2009) on both leaf surfaces of SLN, while limited information is available on the densities of trichomes and stomata. A recently published paper (Travlos, 2013) highlighted that under dry conditions total leaf area of

SLN decreased from around 200 to 80 cm2. A similar reduction on plant height and total biomass were noted as well (Travlos, 2013). Plant morphology can be impacted by many factors including genetics (Clements

& Ditommaso, 2011; Diggle, 1993), DNA content (Achigan-Dako et al.,

2008; Smarda & Bures, 2006), selective pressure (Sharma & Esler, 2008;

Symon, 1986), nutrition (Wakhloo, 1975, 1970) and light (Xu et al., 2012).

The morphological variation within species can make identification of the 120 Solanum species in Australia difficult. Solanum esuriale, S. coactiliferum and S. karsenses are morphologically similar to SLN. Solanum karsenses is restricted to south-western plains of NSW beyond the SLN infesting area. Solanum coactiliferum and SLN can be differentiated by the corolla lobe or stamen number (Table 1), but identification before anthesis can be difficult. Bean (2004) noted that SLN is very similar in morphology to the Australian native S. esuriale, and that microscopic examination of branchlet and calyx trichome is usually required for correct identification.

Generally, S. esuriale is easier to manage than SLN and is not a problem

unless high densities occur (Johnson, et al., 2006) thus landholders may be less diligent with implementing it control. Consequently, when SLN is misidentified as S. esuriale (as occurred in SA, 1918), opportunities for early eradication of an outbreak can be missed and delayed the control efforts (Hosking, et al., 2000). Currently identification of SLN is based on morphological characteristics such as stamen length, spine density or fruit shape (Table 1) (Kidston et al., 2007). Other useful diagnostic features in

Leptostemonum subgenus include sympodial units (Whalen, 1984), trichome

(Seithe, 1979) and root and stem anatomy (Cosa et al., 1998). However, those characters listed in table 1 are plastic, vary considerably within

Australian SLN populations and overlap with S. esuriale. Therefore identification based solely on morphological traits can be unreliable and lead to misidentification.

Table 1. Morphological characteristics of SLN, S. esuriale and Solanum coactiliferum (Kidston, et al., 2007).

21 2.3.2 Lifecycle and genetic diversity

SLN is a herbaceous perennial weed growing in spring, summer and autumn (Stanton et al. 2009). SLN reproduces both sexually as a self- incompatible outcrosser and asexually from adventitious buds in the root system (Petanidou et al., 2012) (Fig. 3). Under Australian climatic conditions, new plants generate from perennial roots or seeds between

October (spring) and April (autumn), and flowering commences from

December (summer) and continues through to March (autumn) (Stanton, et al., 2009a). Fruits normally form in January (but can be as early as

December) and mature within 4-8 weeks after fruit set (McKenzie, 1980).

The aerial growth senesces after frosts in late autumn (May), while the perennial roots remain dormant until next spring.

Fig. 3 Reproductive lifecycle of SLN.

Each SLN individual can produce 40 to 60 fruits annually, and each mature fruit in the USA reportedly contains 24-149 seeds, depending on the

time of germination (Boyd & Murray, 1982b). In Texas (USA), SLN seed production ranges from 12 million to 247 million seeds per hectare, with a population density ranging from 17,000 to 99,000 plants per hectare

(Cooley & Smith, 1972). Similarly, fruit and seed production of SLN varies between locations in NSW, Australia from 45 to 74 fruits and 1,814 to 2,945 seeds per plant, which is probably due to variations in total rainfall (Stanton et al., 2012b).

Asexual reproduction produces genetically identical individuals, while sexual reproduction contributes to genetic diversity, which is a predominant factor contributing to weed success (Green et al., 2001). Single gene variations may result in phenotypes with greater fitness to natural and artificial selection pressures (Dekker, 1997). A recently published paper highlighted that SLN seed production was significantly different between different Greece populations under dry condition (Travlos, 2013), suggesting the impact of genetic background on SLN adaptability and reproduction.

There are several factors which influence species genetic diversity, including mating system, hybridization, mutation and bottlenecks (Barrett,

1992; Prentis et al., 2008). Generally, the sexually reproducing species display a higher level of genetic variation than species that reproduce asexually (O'Hanlon et al., 2000). Sexual reproduction increases the level of allele transmission and heterozygosity (Barrett, 1992).

Intra- and inter-species hybridizations cause rapid gene flow, and promote recombination of the genome, therefore leading to novel genotypes

23 and/or adaptive gene introgression (Prentis et al., 2008). Multiple introductions (Cuthbertson et al., 1976) and inter-species hybridization

(Hardin et al., 1972) were reported in SLN, and these could increase genetic diversity in SLN.

Single base pair mutation is common in many weed species and it can lead to herbicide resistant genotypes. In eastern black nightshade

(Solanum ptycanthum), imidazolinone-resistant genotypes is caused by a single base pair mutation in acetolactate synthase (ALS) genes, which leads to an alanine to threonine mutation (Milliman et al., 2003).

By contrast, bottlenecks usually reduce within-population genetic variation, restrict the evolution, and decrease adaptability (Van Buskirk &

Willi, 2006). However, population bottlenecks may also promote genetic drift and lead to genetic variation (van Heerwaarden et al., 2008), in some conditions such as extreme inbreeding (Prentis et al., 2008). An invasive species usually starts with a small population size. Due to the size and isolation of the population, bottlenecks are common during the introduction and establishment phases (Prentis et al., 2008). A long lag phase after introduction and/or multiple introductions is usually required for a successful invasive species (Ellstrand & Schierenbeck, 2000). Both cases are true in SLN. SLN was introduced many times (Cuthbertson et al., 1976) and has been considered a noxious weed for 50 years (long lag phase) after its first introduction in Australia (McKenzie, 1980).

The adaptation of an invasive species to selection pressure is largely determined by the level of genetic diversity in populations (Barrett, 1992;

Dekker, 1997; O'Hanlon et al., 2000). Therefore, study of genetic diversity

is critical for weed management. Although herbicide resistant SLN has not been reported, herbicide resistant genotypes have been found in three other

Solanum species: American black nightshade (S. americanum, Chase et al.,

1998), eastern black nightshade (Milliman et al., 2003) and black nightshade (S. nigrum, Stankiewicz et al., 2001). Different plant genotypes can also significantly influence the success of biocontrol agents (Nissen et al., 1995), as found in hydrilla (Hydrilla verticillata) (Schmid et al., 2010).

The rapid development in molecular technologies has advanced the ability to detect and examine genetic diversity within plants (O'Hanlon et al.,

2000). Numerous molecular methods have been developed to assay genetic diversity, such as simple sequence repeat (SSR), amplified fragment length polymorphism (AFLP), cleaved amplified polymorphic sequences (CAPS), random amplified polymorphic DNA (RAPD), internal transcribed spacer

(ITS), single-nucleotide polymorphism (SNP), DNA barcoding and diversity arrays technology (DArT). These markers have been used extensively in weed research. For example, Marulanda et al. (2007) used

AFLP and SSR markers to differentiate six Rubus species. SSR markers detected high Simpson’s genetic diversity (0.8) between native Cortaderia jubata (Bolivia, Peru and Ecuador) individuals and low diversity (0.19) between invasive (USA) individuals (Okada et al., 2009). In addition, CAPS markers were used to reveal six diverse acetolactate synthase (ALS)- resistant genes in Lolium rigidum (Yu et al., 2008). Such studies provide insight to the identification, genetic diversity and herbicide resistance of weeds, thus contribute to management.

25 Genetic markers are widely used in Solanaceaeous species, especially for the three Solanum crops, potato (S. tuberosum) (Akkale et al.,

2010), tomato (S. lycopersicum) (Shirasawa et al., 2010) and eggplant (S. melongena) (Fukuoka et al., 2010). For example, Akkale et al. (2010) employed six AFLP primer combinations to study the genetic diversity of

26 potato genotypes in Turkey and detected an average Jaccard’s genetic distance at 0.36. Waycott et al. (2011) developed six highly polymorphic

SSR markers for bush tomato (S. centrale) and detected high genetic diversity between samples, with a Nei’s genetic distance (Nei, 1972) ranging from 0.11 to 1.66.

Using RAPD markers, Hawker et al. (2006) reported a high genetic variation within and between SLN populations in SA. Sexual reproduction of SLN was considered as the predominant factor contributing to this genetic diversity. However, the genetic diversity of SLN growing across

Australia is largely unknown. In addition, AFLP and SSR markers are more reproducible and informative than RAPD in Solanum (Jones et al., 1997;

McGregor et al., 2000). These markers could be employed to better evaluate the genetic diversity of SLN.

2.3.3 Germination and emergence

The soil seedbank of SLN can last for at least 10 years in the USA

(Boyd & Murray, 1982a). Under Australian climatic conditions, seedbank viability decreased to 20% after three years, with seeds buried deeper persisting longer (Stanton, et al., 2012b). Seeds only lasted for three months when ensiled in chopped cereal forage (Stanton et al., 2012a).

The seedbank can reach 4,000 seeds/m2 in the top 10 cm of soil in

Australia (McKenzie, 1980). Germination events are infrequent (Wapshere,

1988) and may be inhibited by the sticky seed coat (Rutherford, 1978) as seedlings are usually found after heavy summer thunderstorms (Molnar &

McKenzie, 1976). Seedlings emerged in spring are exposed to hot and dry summer conditions, while those germinated in autumn may be killed by the cold winter, thus only a few of these seedlings may last to the next growing season (McKenzie, 1980).

Diurnal temperature fluctuations promote germination, while germination at constant temperatures is less than 5% (Boyd & Murray,

1982b; Stanton, et al., 2012b; Trione & Cony 1990). Germination rates of around 42% were achieved with fluctuating temperatures of 10/ 25, 15/ 25 and 15/ 30 ºC (Stanton, et al., 2012b), while higher germination (57%) was reported at fluctuating temperatures of 20/ 30 ºC (Boyd & Murray, 1982b).

There are contrasting reports on the effect of pH on germination.

Boyd and Murray (1982b) found that germination peaked (59%) at pH between 6 and 7, and declined dramatically with no germination detected at pH 3 and 9. By contrast, Stanton et al. (2012b) found that germination increased linearly with pH (4 to 10, r = 0.8). Factors such as seed pre- treatment, incubation conditions and duration, and the maintenance of petri dish solution volumes, and therefore pH, could have contributed to the reported differences.

The SLN root system can grow up to 4 m in depth (Australian

Weeds Committee, 2012), and has strong ability to regenerate (Fig. 4)

27 (Cuthbertson, et al., 1976). Shoots can emerge from laterals as deep as 0.5 m in cultivated areas (Monaghan & Brownlee, 1979). Regeneration can occur from a root fragment as short as 10 mm length (Stanton, et al., 2011).

Under moist conditions, taproot fragments (5 cm in length) can survive for

15 months (Fernandez & Brevedan, 1972).

Fig. 4 SLN regeneration from a 10 cm long root fragment.

2.3.4 Spread

Sexual reproduction resulting in seed production contributes to the initial establishment and long distance dispersal of this species (Wapshere,

1988). Livestock, especially sheep, are the most important vector for seed dispersal in Australia (Heap & Honan, 1993; Stanton, et al., 2009a). SLN fruits may be eaten by sheep from mid-January to late April if there is a low supply of alternative pasture (Heap & Honan, 1993). Seeds can stay in the digestive system for 31 days, while most seeds will be excreted 7-9 days after ingestion (Heap & Honan, 1993). McKenzie (1975) detected 85% germination of excreted seeds compared to 77% germination of non- ingested seeds, while Heap and Honan (1993) reported a lower germination percentage of excreted seeds, possibly due to dormancy. A high proportion

(11 and 23%) of excreted seeds can be recovered (Heap & Honan, 1993;

McKenzie, 1975) and these seeds germinate faster than non-ingested seeds

(21% compared to 1% germination after 2 days) (McKenzie, 1975).

Therefore, subsequent transportation of contaminated livestock will contribute to spread.

Irrigation water and organic manure are also considered as effective seed dissemination vectors (Brunel, 2011; Taleb et al., 2007). Accidental transport of seeds through agricultural products, such as hay or grain contamination, can be another important mode of long-distance dispersal

(Mekki, 2007). In addition, mature fruits can remain attached to dead stems for months and subsequent transport of these stems by machinery, wind and water can also contribute to distribution (Mekki, 2007).

Compared to the seedbank, the soil rootbank plays a major role in maintaining infestations (Wapshere, 1988). SLN spread within farms can be attributed to cultivation, machinery and the expansion of the root system

29 (Stanton, et al., 2009a). However, the expansion of the root system highly depends on soil moisture. Under favourable conditions, the diameter of SLN patches can increase 3.9 m annually, while patches can decrease in size during dry seasons (McKenzie, 1980).

In conclusion, SLN is a widespread perennial weed worldwide. It is one of the worst weeds in Australian agricultural systems. Due to inadequate knowledge of SLN, management strategies are very limited, especially for dense and large infestations (Wassermann et al., 1988).

Chemical control is useful, however efficacy is impacted by many factors, such as morphological variation (Brewer, et al., 1991), genetic diversity

(Marshall & Moss, 2008) and growth stage (Greenfield, 2003). Better understanding of morphological variation, genetic diversity and phenological development is required for better management of SLN.

30 CHAPTER 3 PUBLISHED PAPER

As an initial step, this chapter is a large scale study of SLN morphological variation and highlights the potential correlation between rainfall and phenotypes.

Zhu, X. C., Wu, H. W., Stanton, R., Burrows, G. E., Lemerle, D., & Raman,

H. (2013). Morphological variation of silverleaf nightshade (Solanum elaeagnifolium Cav.) in south-eastern Australia. Weed Research. doi:

10.1111/wre.12032.

31 DOI: 10.1111/wre.12032

Morphological variation of Solanum elaeagnifolium in south-eastern Australia

X C ZHU*†,HWWU*‡,RSTANTON*†, G E BURROWS*†, D LEMERLE*† &HRAMAN*‡ *EH Graham Centre for Agricultural Innovation (an alliance between NSW Department of Primary Industries and Charles Sturt University), Wagga Wagga, NSW, Australia, †School of Agricultural and Wine Sciences, Charles Sturt University, Wagga Wagga, NSW, Australia, and ‡Wagga Wagga Agricultural Institute, PMB, Wagga Wagga, NSW, Australia

Received 13 December 2012 Revised version accepted 16 April 2013 Subject Editor: David Clements, Trinity Western University, Canada

possible correlation between rainfall and morphology. Summary Scanning electron microscopy comparison of leaf Solanum elaeagnifolium (silverleaf nightshade) is an surfaces showed lower trichome and stomatal densities invasive perennial weed in Australia, with aerial growth on the adaxial surface (67.0 Æ 3.3 trichomes mmÀ2 commencing in spring from either the perennial root and 603.4 Æ 29.2 stomata mmÀ2 respectively) than on system or the soil seedbank, with senescence occurring the abaxial surface (131.9 Æ 7.2 trichomes mmÀ2 and in autumn. A total of 642 S. elaeagnifolium individuals 813.7 Æ 30.5 stomata mmÀ2 respectively). The mor- were collected at ﬂowering from 92 locations in south- phological plasticity of S. elaeagnifolium highlighted in eastern Australia to study morphological variation and this study could probably contribute to its adaptability its implications for management. Large morphological and partly explain its establishment and continuing variation was found between individuals from different expansion in Australia. locations. Leaf length, width and area ranged from Keywords: invasive weed, growth, silverleaf nightshade, 1.44 to 10.6 cm, 0.39 to 4.09 cm and 0.41 to 25.8 cm2 size variability, stomata, trichomes. respectively. Plants from higher rainfall regions were signiﬁcantly taller and had larger leaves, suggesting a

ZHU XC, WU HW, STANTON R, BURROWS GE, LEMERLE D&RAMAN H (2013). Morphological variation of Solanum elaeagnifolium in south-eastern Australia. Weed Research.

& Cuthbertson, 2001). Solanum elaeagnifolium repro- Introduction duces sexually and vegetatively. In Australia, the above- Solanum elaeagnifolium Cav. (silverleaf nightshade) is a ground shoots emerge from perennial root systems or weed of worldwide signiﬁcance that originated from soil seedbank from September (spring) and the shoots south-western USA and northern Mexico (Stanton senesce in May (autumn) (Stanton et al., 2009). et al., 2009). It infests at least 0.35 million hectares in Morphological variation of S. elaeagnifolium such as Australia, covering various climatic zones and has the prickle density, leaf size and shape has been reported potential to infest 398 million hectares (Feuerherdt, previously in the USA (Bryson et al., 2012), Greece (En- 2009). The species occurs over the southern cereal comidou & Yannitsaros, 1975) and Australia, based on cropping zone of Australia (Stanton et al., 2009), limited herbarium samples (Symon, 1981; Bean, 2004). which has an annual rainfall of 250–600 mm (Parsons Solanum elaeagnifolium plants are usually between 10

Correspondence: X C Zhu, School of Agricultural and Wine Sciences, Charles Sturt University, Wagga Wagga, NSW 2678, Australia. Tel: (+61) 2 6933 2749; Fax: (+61) 2 6938 1861; E-mail: [email protected]

© 2013 European Weed Research Society 32 2 X C Zhu et al. and 100 cm high (Encomidou & Yannitsaros, 1975), production (95–3420 per plant) under higher water with leaf length and width ranging from 3 to 6.5 cm and availability (Travlos, 2013). 0.8 to 1.4 cm respectively (Bean, 2004). Prickles are Variations of leaf morphology, such as leaf size absent or present on the stem, leaf and calyx. Prickle (Richburg et al., 1994), trichome density (Huangfu density on S. elaeagnifolium stems may increase under et al., 2009) and stomatal density (Ricotta & Masiunas, dry conditions (Encomidou & Yannitsaros, 1975). 1992), can affect foliar herbicide uptake. Richburg et al. Under glasshouse condition with reliable water supply, (1994) reported that small leaf size of Cyperus spp. plants can grow 50–120 cm high with prickles present (nutsedge) affected herbicide droplet coverage and on calyx, leaf and stem (Bryson et al., 2012). retention and reduced foliar herbicide efficacy. In addi- Morphological variation has been highlighted in tion, dense trichomes can form a water repellent surface many Solanum species, including eggplant (Solanum mel- that blocks herbicide uptake (Brewer et al., 1991). Hua- ongena L.) (Prohens et al., 2005), Solanum bahamense L. ngfu et al. (2009) observed that glyphosate uptake in (canker berry), Solanum capsicoides All. (cockroach Brassica juncea was negatively correlated (r = À0.65) berry) and Solanum carolinense L. (carolina horsenettle) with trichome density on the adaxial leaf surface. (Bryson et al., 2012). Prohens et al. (2005) compared the However, herbicides can penetrate into leaves through morphological characters of 28 Spanish eggplant the guard cells of stomata (Wang & Liu, 2007). For cultivars and found considerable variation, especially example, Ricotta and Masiunas (1992) detected a high regarding fruit length, ranging from 9.3 to 25.8 cm. negative correlation between acifluorfen tolerance and Bryson et al. (2012) observed morphological characters stomatal density (r = À0.895) in different tomato and variations of 18 weedy and non-weedy prickly genotypes. nightshades (Solanum spp.) in the south-eastern USA Effective management strategies for S. elaeagnifoli- and provided diagnostic features to identify these um are very limited, especially for large and dense infes- species. tations, and are often herbicide based. A large-scale Morphological variation has been studied in many study of the morphological variation of S. elaeagnifoli- weed species under controlled environment conditions, um across Australia is required for improving the including Oryza sativa L. (weedy rice) (Fogliatto et al., identification, understanding and management of this 2010) and Brassica juncea (L.) Czern. (Indian mustard) weed. The objective of this study was to assess the (Huangfu et al., 2009), or in field surveys, such as of morphological variation of S. elaeagnifolium in south- Solidago canadensis L. (Canada goldenrod) (Weber, eastern Australia and to identify possible relationships 1997). Oryza sativa from north-west Italy were divided between morphology and abiotic factors such as into three groups: awnless, mucronate and awned rainfall. (Fogliatto et al., 2010). The flag leaf length of awnless O. sativa was significantly shorter than the awned one, Materials and methods while the awned and awnless populations were significantly taller than the mucronate population (Fogliatto Plant material et al., 2010). Weber (1997) sampled 379 S. canadensis field individuals from Europe and found that leaf A total of 642 individuals were collected from 92 loca- length and width ranged from 6.5 to 20.7 cm and 0.8 tions across New South Wales (NSW), South Australia to 35 cm respectively. (SA), Victoria (VIC) and Queensland (QLD) in Febru- These morphological variations may be attributed ary 2010 (Appendix 1). Sampling locations and rainfall to genetic (Clements & Ditommaso, 2011) and/or areas are shown in Fig. 1. All individuals were phenotypic plasticity in response to abiotic factors collected within 3 weeks, to minimize the environmen- such as light (Xu et al., 2012), habitats (Sharma & tal impacts on morphology. One to 12 individuals at Esler, 2008) and nutrition (Hejcman et al., 2012). Phe- flowering stage were randomly collected at each loca- notypic plasticity increases the adaptability of invasive tion, depending on the level of infestation. One indi- species and contributes to their success (Clements & vidual was sampled if there was only one patch of Ditommaso, 2011). For example, habitats significantly S. elaeagnifolium present, while at locations with large impacted on morphology of Echium plantagineum L. populations (more than 1 ha), up to 12 individuals (Paterson’s Curse), including plant height, seed size were sampled and spaced at least 50 m apart from and seed weight (Sharma & Esler, 2008). Such mor- each other to reduce the probability of sampling clonal phological plasticity was also reported in S. elaeagnifo- individuals. Individuals were chosen from open fields lium, with a significant increase in plant height (around with few trees or buildings to avoid any shading 30–90 cm), total leaf area (around 80–200 cm2), effects. An above-ground shoot from each individual biomass (around 200–500 g per plant) and seed was cut, placed in a zip-lock plastic bag and kept in an

Fig. 1 Solanum elaeagnifolium sampling points and total rainfall between Septem- ber 2009 and February 2010; rainfall data obtained from Bureau of Meteorology, 0 100 200 300 Australia. Kilometres insulated container for transport. Digital photographs ally assessed as low, medium and high density (<30, of the sixth, seventh and eighth leaves from the shoot 30–50 and >50 prickles/calyx respectively). Leaf prick- apex were taken within 24 h after sampling for subse- les were classified as absent, present on abaxial leaf quent measurement of leaf size and shape. These leaves surface only, or present on both surfaces. were chosen because they are at similar maturity stage and fully expanded. Additionally, the chosen leaves Trichome and stomatal densities assessed by SEM would have to be available from the shortest plants and have to be clean (not too close to the ground). Scanning electron microscopy (JCM 5000 NeoScope, The shoots were then pressed and dried for scanning JEOL, Japan) images of 77 individuals from 33 loca- electron microscopy (SEM) examination. tions were used to study trichome density (Appendix 1). Three mature leaves were randomly chosen from each individual. For each leaf, small areas were cut from the Morphological evaluation adaxial and abaxial surfaces, adhered to 12-mm carbon All individuals were measured for 10 morphological tabs (ProSciTech, Australia) and observed by SEM. The traits, as follows: plant height, leaf length, width, area adaxial trichome density was counted from three SEM and roundness, trichome density on the adaxial and images (0.2–3.7 mm2, according to trichome density) per abaxial leaf surfaces, and prickle density on stem, leaf leaf (Fig. 2A). Due to the presence of multiple trichome and calyx. layers on the abaxial surface (Christodoulakis et al., Plant height was measured in the field from the 2009), trichomes were shaved using a scalpel under a ground to the highest shoot apex. Digital photographs dissecting microscope, and the trichome basal cell den- of the sixth, seventh and eighth leaves were processed sity was counted on a single SEM image (0.2–3.7 mm2, using Image J software (Schneider et al., 2012) to according to trichome density) per leaf (Fig. 2B). measure leaf length, width and area. Leaf roundness Stomatal density was determined from a subset of 41 was also calculated according to formula: 4 9 leaf individuals from 23 locations. Adaxial and abaxial tric- area/p(major axis)2 using the same software, with a homes were shaved, and adaxial stomata were counted value of 1.0 indicating a perfect circle, where major from three SEM images (0.02–0.15 mm2, according to axis indicates the long axis of the best fitting ellipse. stomatal density) per leaf (Fig. 2C), while the abaxial Trichome density was visually assessed on fresh stomata were counted from a single image (0.02– leaves on the adaxial and abaxial surfaces and classi- 0.15 mm2, according to stomatal density) per leaf fied into three groups: low, medium and high densities (Fig. 2D). according to the degree of silvery-white colour on the leaf surface. Preliminary investigations suggested that a Statistical analysis visual rating of the ‘medium’ category equated to a trichome density from 40 to 80 trichomes mmÀ2. This Coefficients of variation and a histogram plot were visual rating was verified by trichome counts on calculated for each morphological character. Mean, SEM images from 77 randomly selected individuals standard error of mean, and correlation matrix were (described in the next section). calculated for quantitative traits. Plant height was Stem prickles were visually assessed on a 5 cm square root transformed, while the other quantitative length of the mid-stem and classified into three density traits were log-transformed before analysis to normal- levels: low (<40 prickles), medium (40–60 prickles) and ize variances. Individuals from the same rainfall areas high (more than 60 prickles). Calyx prickles were visu- (Fig. 1) were considered as a population and subjected

Fig. 2 Scanning electron microscopy images of trichome and stomatal densities in Solanum elaeagnifolium; (A): trichomes on the adaxial leaf surface; (B): trichome basal cells on shaved abaxial leaf surface; (C): stomata on shaved adaxial leaf surface; and (D): stomata on shaved abaxial leaf surface.

Fig. 3 Variation of Solanum elaeagnifolium leaf size and shape. All leaves were the seventh leaf from the shoot apex of ﬂowering plants, showing the adaxial surface only. to unbalanced analysis of variance (ANOVA). Means leaf width from 0.39 to 4.09 cm (Figs 3 and 4). Leaf were separated by Fisher’s LSD at the 5% level. Prin- area had the largest coefﬁcient of variation at 69.5%. ciple component analysis (PCA) was performed for The largest leaf area was 25.8 cm2, which was 63 times quantitative traits using the multivariate analysis greater than the smallest area (0.41 cm2). Most individ- model. The relation between trichome and stomatal uals (88.2%) had a leaf area of 0.41–12 cm2. Leaf densities was tested by linear regression analysis. All roundness ranged from suborbicular (0.61) to narrowly analyses were performed using Genstat 14th edition strap shaped (0.15). Plant height varied from 7 to (Payne et al., 2011). 73 cm, with an average of 33.6 cm. There were 510 individuals (79.4%) between 20 and 50 cm in height. Results High correlations were found among leaf length, width and area (Table 1). Leaf roundness showed Morphological variation weak correlation with leaf width and area (r = 0.43 Considerable morphological variation was present in and 0.22 respectively). In addition, a weak, negative quantitative traits of the 642 individuals of S. elaeag- correlation was found between plant height and leaf nifolium. Leaf length ranged from 1.44 to 10.6 cm and roundness (r = À0.24).

250 160 400 200 200 200 120 300 150 150 150 80 200 100 100 100

40 50 50 100 50 Number of individuals 0 0 0 0 0 1 283 4 596 7 10 0 1.0 2.0 3.0 4.0 0 3 6 9 12 15 18 21 24 27 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0510 2030 400 60 70 80 Leaf length (cm) Leaf width (cm) Leaf area (cm2) Leaf roundness Plant height (cm)

350 400 250 300 600 300 500 300 250 200 250 200 400 150 200 200 150 300 150 100 100 200 100 100 50 50 100

Number of individuals 50 0 0 0 0 0 Absent Low Medium High Absent Abaxial Both Absent Low Medium High Low Medium High Low Medium High Prickle density on stem Prickle presence on leaf Prickle density on calyx Trichome density on adaxial surface Trichome density on abaxial surface

Fig. 4 Frequency distributions of the morphological traits of Solanum elaeagnifolium.

Table 1 Correlation matrix for quantitative traits of Solanum not the leaves. The remaining 289 individuals (45.0%) elaeagnifolium had prickles on both the stem and the leaves. By Leaf Leaf Leaf Leaf contrast, no obvious correlation was found between Traits length width area roundness calyx prickle density and other morphological traits. The majority of individuals (91.6%) had a high Leaf – length trichome density on the abaxial leaf surface (more than À2 Leaf 0.855*** – 80 trichomes mm ). However, 77, 250 and 315 width individuals were classified with low, medium and high Leaf area 0.923*** 0.943*** – adaxial leaf trichome densities respectively. À – Leaf 0.030 0.434*** 0.223*** Plant height and leaf shape characteristics were roundness significantly different between rainfall areas (Table 2). Plant 0.057 À0.109** À0.004 À0.240*** height In areas of greater than 300 mm of rainfall, individuals had significantly (P < 0.001) larger leaf length, width Correlation is significant at **: P<0.01 level and ***: P ≤ 0.001 and area (5.81 cm, 1.58 cm and 6.88 cm2 respectively) level. as compared to plants from low (<200 mm) rainfall areas (4.55 cm, 1.34 cm and 4.71 cm2 respectively). In High variation was found in qualitative traits the higher rainfall areas, plants were taller and had (Fig. 4), with the coefficient of variation of prickle den- more pointed leaves. In addition, plants from high sity on the leaf, stem and calyx at 45.6%, 46.1% and rainfall areas tended to have fewer stem prickles 29.5% respectively. Individuals with no or low stem (Fig. 5). Plants from low (42.9%) and medium prickle density represented 64.5% of the 642 individu- (36.6%) rainfall areas had a larger proportion of als, while similar numbers of individuals had medium medium and high prickle density individuals than those and high stem prickle densities (109 and 117 individuals from high rainfall areas (19.5%). respectively). More than half of the individuals (54.1%) Individuals with a larger leaf area had lower did not have prickles on the leaves, while 96 (15%) indi- (P < 0.001) adaxial trichome density. The average leaf viduals had prickles on both leaf surfaces. Prickles were area of the individuals with low adaxial trichome 2 always present on the abaxial leaf surface if they were density (TAL) was 7.99 Æ 0.64 cm , almost twice the present on the adaxial surface. The majority of individ- size of individuals with high trichome density (TAH, 2 uals (65%) had a medium level of prickles on the calyx 4.10 Æ 0.16 cm ). Most of the TAH individuals (73%) (30–50 prickles). had a relatively small leaf area between 0.41 and 2 There were 164 individuals (25.6%) without prickles 6cm, while the majority (78%) of TAL individuals on either the stems or leaves. Six individuals (0.9%) had a leaf area between 2 and 12 cm2, with 8% of had prickles on the leaves but not on the stem, and individuals with a leaf area of more than 18 cm2 183 individuals (28.5%) had prickles on the stem but (Fig. 6).

Table 2 Relationship between growing season rainfall areas and morphological traits of Solanum elaeagnifolium

Rainfall September 2009 – February 2010 (mm) <200 200–300 >300 Traits Mean SE Mean SE Mean SE

Leaf length (cm) 4.55a 0.098 4.68a 0.092 5.81b 0.120 Leaf width (cm) 1.34a 0.032 1.38a 0.035 1.58b 0.043 Leaf area (cm2) 4.71a 0.247 5.05a 0.229 6.88b 0.321 Leaf roundness 0.28a 0.004 0.27b 0.004 0.25c 0.004 Plant height (cm) 29.3a 0.734 34.2b 0.708 39.0c 0.871

SE, standard error of mean. Values sharing the same letters within each row are not signiﬁcantly different according to Fisher’s LSD (P = 0.05).

60% dispersed, and no well-separated groups were identiﬁed Absent Low (Fig. 7). 50% Medium High

40% Trichome and stomatal densities observed by SEM 30% Trichome density varied between individuals, with a coefficient of variation of 42.3% and 47.9% for the adaxial 20% each rainfall area and abaxial leaf surfaces respectively (Fig. 8). Trichome 10% density on the adaxial leaf surface ranged from 21.9 to Percentage of individuals from 196.0 trichomes mmÀ2, with a mean of 67.0 Æ 3.3 tric- 0 À2 <200 (N = 219) 200-300 (N = 274) >300 (N = 149) homes mm , which is lower than the density on the Rainfall September 2009 - February 2010 (mm) abaxial surface (57.4–395.3 trichomes mmÀ2,witha mean of 131.9 Æ 7.2 trichomes mmÀ2). Fig. 5 Frequency distributions of stem prickle density for Solanum elaeagnifolium individuals from different rainfall areas. Numbers The coefficient of variation for stomatal density was in the parentheses indicated the sample size in the particular 29.5% and 24.0% for the adaxial and abaxial leaf sur- rainfall area. faces respectively. Stomatal density ranged from 284 to 942 (mean of 603.4 Æ 29.2) and 455 to 1519 (mean of 813.7 Æ 30.5 stomata mmÀ2) on the adaxial and abax- Principal component analysis ial leaf surfaces respectively (Fig. 8). The first two components of PCA explained 83.3% of There was a weak positive correlation in trichome the total variation (Fig. 7). The first principal compo- density between the adaxial and abaxial leaf surfaces nent accounted for 57.8% of the variation and was (r = 0.43). A weak positive correlation was also found mainly contributed by length, width and area of the on stomatal density between the adaxial and abaxial leaf leaves (Table 3). The second component explained surfaces (r = 0.59). For all individuals, trichome and 25.5% of the variation with dominance of plant stomatal densities were always higher on the abaxial height and leaf length. A negative correlation was surface. No correlation was found between trichome found with leaf roundness. Individuals were randomly and stomatal densities.

35%

TAL population 30% TAM population

TAH population 25%

20%

15%

10% Fig. 6 Frequency distributions of leaf Percentage of individuals 5% area (cm2) for Solanum elaeagnifolium

0% individuals with low (TAL, n = 77), 0 6 12 18 24 0 6 12 18 24 0 6 12 18 24 medium (TAM, n = 250) and high (TAH, 2 Leaf area (cm ) n = 315) adaxial trichome densities.

The reliability of the visual assessment method of than 80 trichomes mmÀ2, respectively, with an error trichome density based on the degree of the silvery- rate of about 6.5% (five out of 77). white colour on leaf surface was confirmed with the direct measurement of trichome densities from SEM Discussion observation. The SEM observation indicated that those individuals visually assessed as having low, medium This is the first large-scale morphological study of and high trichome densities were successful validated S. elaeagnifolium in south-eastern Australia. High mor- to have a trichome density of <40, 40–80 and more phological variation was found for all traits, except trichome density on the abaxial leaf surface. The ranges in S. elaeagnifolium leaf length and width reported 4 here were greater than reported previously in Australia (3.0–6.5 cm in length and 0.8–1.4 cm in width) (Bean, 2004), but similar to those reported from Greece 2 (2–19 cm in length and 0.5–5.2 cm in width) (Encomi- dou & Yannitsaros, 1975). The morphological difference may be associated with genetic diversity of this 0 weed (Zhu et al., 2012) or edaphic and climatic difference between locations. Individuals collected from –2 higher rainfall areas are often taller and have larger leaves, indicating a possible correlation between phenotypic plasticity and water availability. A recently study –4 has shown that plant height and total leaf area of S. elaeagnifolium reduced from around 90 to 30 cm 2 Second principal component axis (25.53%) Individuals from high rainfall area and 200 to 80 cm in response to dry conditions –6 Individuals from medium rainfall area Individuals from low rainfall area respectively (Travlos, 2013). Phenotypic plasticity has been reported in response to habitats for Echium –2 0 2 4 6 8 plantagineum (Sharma & Esler, 2008), light for Alter- First principal component axis (57.79%) nanthera philoxeroides (Mart.) Griseb. (alligator weed) (Xu et al., 2012) and nutrition for Rumex crispus L. Fig. 7 Principle component analysis for 642 Solanum elaeagnifolium individuals from Australia. (curled dock) (Hejcman et al., 2012). Such phenotypic plasticity plays an important role in weed establishment and increases the adaptability of invasive species Table 3 Principal component analysis of 642 Solanum elaeagnifo- to novel environments (Dawson et al., 2012). lium individuals from Australia; showing the percentage of variation accounted by the first two principle components In addition, this study indicated that individuals from low rainfall areas tended to have more prickles Variables PC1 PC2 on the main stem. Encomidou and Yannitsaros (1975) Leaf length 0.53922 0.28436 also reported that S. elaeagnifolium growing under dry Leaf width 0.57962 À0.08693 conditions in Greece were more likely to have more Leaf area 0.57622 0.10267 prickles. Prickles are structural defence adaptations À Leaf roundness 0.19835 0.66912 that help protect plants from herbivores and are more Plant height À0.04365 0.67329 likely to be developed under suboptimal conditions,

AB400

350 1400

) ) 300

2 2 1200

250 1000 200 800

(trichome/mm 150 (stomata1/mm

Abaxial trichome density Fig. 8 Relationship between trichome (A) 100 Abaxial stomata1 density 600 and stomata (B) densities on Solanum ela- 50 eagnifolium adaxial and abaxial leaf sur- 200175150125100755025 900800700600500400300 faces observed by scanning electron Adaxial trichome density Adaxial stomatal density microscopy. (trichome/mm2) (stomatal/mm2)

© 2013 European Weed Research Society 38 8 X C Zhu et al. such as low rainfall, because the drier conditions may Masiunas (1992) reported that guard cells are more reduce overall plant growth, leading to increased permeable than other leaf cells, due to a thinner cuticle resources being available from photosynthesis for and better connection with subjacent cells. Further development of trichomes and prickles (Hanley et al., research may focus on suitable adjuvants and concen- 2007). trations to improve foliar herbicide efficacy on S. ela- Scanning electron microscopy observation indicated eagnifolium, through increasing leaf penetrability. The that trichome densities ranged from 22 to 196 tric- potential difficulties in uptake of foliar herbicide also homes mmÀ2 on the adaxial surface and from 57 to suggest that root absorbed residual herbicides should 395 trichomes mmÀ2 on the abaxial surface, which is be used in conjunction with the foliar applied herbi- much higher than other investigated Solanum species cides to improve control of S. elaeagnifolium. (0.84–7.13 trichomes mmÀ2), including six species of In conclusion, high morphological variation was S. arboreum group and two species of S. deflexiflorum found in S. elaeagnifolium from south-eastern Austra- group (de Rojas & Ferrarotto, 2009) and eggplant lia. The ranges of the leaf size parameters were larger (Leite et al., 2003). Recently, Blonder et al. (2012) than previous records in Australia. In addition, showed that most leaves shrink 10–30% when dried; individuals from high rainfall areas had a larger leaf thus, the densities reported here may be higher than area than those from low rainfall areas, suggesting a that in fresh leaves. High trichome density might help possible relationship between rainfall and phenotypic protect leaves from high summer temperatures and plasticity. However, the relationship between leaf size reduce transpiration rates (Perez-Estrada et al., 2000), and trichome/stomatal density is not well understood. solar radiation (Jordan et al., 2005) and herbivores Further studies are needed using clonal material (Levin, 1973). Jordan et al. (2005) highlighted that under controlled environments to determine the open vegetation usually associates with dense impact of abiotic factors on any such correlation. trichomes or papillae in Proteaceae species, which indi- Variation in leaf size and trichome and stomatal den- cated the photoprotection function of trichomes. Tric- sities highlighted in this study could potentially homes on S. elaeagnifolium create a silver-white surface, impact the retention and uptake of foliar applied her- which could reflect and reduce the light that reaches leaf bicides, thus influencing the control of this weed. surfaces and protects leaves from damage due to high High trichome and stomata densities on both leaf sur- summer temperatures and solar radiation. Trichomes faces suggested that effective management of S. ela- also play an important role against herbivores including eagnifolium may require the use of soil-applied molluscs and leaf chewing and sap-sucking insects (Han- herbicide through root uptake and the use of appro- ley et al., 2007), which probably explains the very lim- priate adjuvants to improve the efficacy of foliar ited insect damage found on S. elaeagnifolium leaves applied herbicide. during our sample collection. The stomatal density on both leaf surfaces detected Acknowledgements in this study is extremely high compared with other studies (e.g. Beaulieu et al., 2008). A high density of We acknowledge Charles Sturt University, Australia stomata on both leaf surfaces is probably associated for funding this research, Dr. Neil Coombes (NSW with increasing CO2 conductance and photosynthetic Department of Primary Industries) for his assistance capacity under optimal conditions (Beaulieu et al., on statistical analysis, Craig Poynter (Spatial Data 2008). Analysis Network, Charles Sturt University) and the The morphological variation of leaf area and National Climate Centre, Bureau of Meteorology for trichome and stomatal densities highlighted in this study providing Australian rainfall information and a num- could potentially impact herbicide control of S. elaeag- ber of state and local organizations across Australia nifolium. Plants with larger leaves may intercept more for their assistance in field sampling. herbicide droplets than those of smaller leaves (Rich- burg et al., 1994), thus increasing the amount of herbi- References cide available for uptake. Dense trichomes can form a hydrophobic barrier on the leaf surface (Brewer et al., BEAN AR (2004) The taxonomy and ecology of Solanum 1991), reduce droplet retention, create air pockets and subg. Leptostemonum (Dunal) Bitter (Solanaceae) in block herbicide uptake (Kraemer et al., 2009), hence Queensland and far north-eastern New South Wales. Austrobaileya 6, 639–816. may restrict herbicide efficacy on S. elaeagnifolium. BEAULIEU JM, LEITCH IJ, PATEL S, PENDHARKAR A&KNIGHT However, this study also showed high stomatal densi- CA (2008) Genome size is a strong predictor of cell size ties on both leaf surfaces, which may aid in chemical and stomatal density in angiosperms. The New Phytologist management of S. elaeagnifolium, as Ricotta and 179, 975–986.

BLONDER B, BUZZARD V, SIMOVA I et al. (2012) The leaf-area LEVIN DA (1973) The role of trichomes in plant defense. The shrinkage effect can bias paleoclimate and ecology Quarterly Review of Biology 48,3–15. research. American Journal of Botany 99, 1756–1763. PARSONS WT & CUTHBERTSON EG (2001) Noxious Weeds of BREWER CA, SMITH WK & VOGELMANN TC (1991) Functional Australia, 2nd edn. CSIRO Publishing, Collingwood, interaction between leaf trichomes, leaf wettability and the Australia. optical properties of water droplets. Plant, Cell and PAYNE RW, MURRAY DA, HARDING SA, BAIRD DB & Environment 14, 955–962. SOUTAR DM (2011) An Introduction to GenStat for BRYSON CT, REDDY KN & BYRD JD Jr (2012) Growth, Windows, 14th edn. VSN International, Hemel development, and morphological differences among native Hempstead, UK. and nonnative prickly nightshades (Solanum spp.) of the PEREZ-ESTRADA LB, CANO-SANTANA Z&OYAMA K (2000) southeastern United States. Invasive Plant Science and Variation in leaf trichomes of Wigandia urens: Management 5, 341–352. environmental factors and physiological consequences. CHRISTODOULAKIS NS, LAMPRI PN & FASSEAS C (2009) Tree Physiology 20, 629–632. Structural and cytochemical investigation of the leaf of PROHENS J, BLANCA JM & NUEZ F (2005) Morphological and silverleaf nightshade (Solanum elaeagnifolium), a drought- molecular variation in a collection of eggplants from a resistant alien weed of the Greek flora. Australian Journal secondary center of diversity: implications for conservation of Botany 57, 432–438. and breeding. Journal of the American Society for CLEMENTS DR & DITOMMASO A (2011) Climate change and Horticultural Science 130,54–63. weed adaptation: can evolution of invasive plants lead to RICHBURG JS, WILCUT JW & WEHTJE GR (1994) Toxicity of greater range expansion than forecasted? Weed Research AC 263,222 to purple (Cyperus rotundus) and yellow 51, 227–240. nutsedge (C. esculentus). Weed Science 42, 398–402. DAWSON W, ROHR RP, VAN KLEUNEN M&FISCHER M (2012) RICOTTA JA & MASIUNAS JB (1992) Relationship of leaf Alien plant species with a wider global distribution are surface characteristics to acifluorfen tolerance in tomato better able to capitalize on increased resource availability. (Lycopersicon esculentum) and related species. Weed The New Phytologist 194, 859–867. Science 40, 402–407. ENCOMIDOU E&YANNITSAROS A (1975) Research on the DE ROJAS CEB & FERRAROTTO M (2009) Morphology of weed flora of Greece. III. Morphology, development and foliar epidermis in two groups of Solanum section phenology of Solanum elaeagnifolium Cav. Candollea 30, Geminata. Caldasia 31,31–40. 29–41. SCHNEIDER CA, RASBAND WS & ELICEIRI KW (2012) NIH FEUERHERDT L (2009) Overcoming a deep rooted perennial Image to ImageJ: 25 years of image analysis. Nature problem–silverleaf nightshade (Solanum elaeagnifolium)in Methods 9, 671–675. South Australia. Plant Protection Quarterly 24, 123–124. SHARMA GP & ESLER KJ (2008) Phenotypic plasticity among FOGLIATTO S, VIDOTTO F&FERRERO A (2010) Morphological Echium plantagineum populations in different habitats of characterisation of Italian weedy rice (Oryza sativa) Western Cape, South Africa. South African Journal of populations. Weed Research 52,60–69. Botany 74, 746–749. HANLEY ME, LAMONT BB, FAIRBANKS MM & RAFFERTY CM STANTON R, HEAP JW, CARTER RJ & WU H (2009) Solanum (2007) Plant structural traits and their role in anti-herbivore elaeagnifolium. In: The Biology of Australian Weeds, Vol. 3 defence. Perspectives in Plant Ecology, Evolution and (ed. FD PANETTA), 274–293. R. G. and F. J. Richardson, Systematics 8, 157–178. Melbourne. HEJCMAN M, KRISTALOVA V, CERVENA K, HRDLICKOVA J& SYMON DE (1981) A revision of the genus Solanum in PAVLU V (2012) Effect of nitrogen, phosphorus and Australia. Journal of the Adelaide Botanic Gardens 4,1–367. potassium availability on mother plant size, seed TRAVLOS I (2013) Responses of invasive silverleaf nightshade production and germination ability of Rumex crispus. (Solanum elaeagnifolium) populations to varying soil water Weed Research 52, 260–268. availability. Phytoparasitica 41,41–48. HUANGFU CH, SONG XL & QIANG S (2009) Morphological WANG CJ & LIU ZQ (2007) Foliar uptake of pesticides– disparities in the epidermal and anatomical features of the present status and future challenge. Pesticide Biochemistry leaf among wild Brassica juncea populations. Weed Biology and Physiology 87,1–8. and Management 9, 234–242. WEBER E (1997) Morphological variation of the introduced JORDAN GJ, DILLON RA & WESTON PH (2005) Solar perennial Solidago canadensis L. sensu lato (Asteraceae) in radiation as a factor in the evolution of scleromorphic leaf Europe. Botanical Journal of the Linnean Society 123,197–210. anatomy in Proteaceae. American Journal of Botany 92, XU CY, SCHOOLER SS & VAN KLINKEN RD (2012) 789–796. Differential influence of clonal integration on KRAEMER T, HUNSCHE M&NOGA G (2009) Surfactant- morphological and growth responses to light in two induced deposit structures in relation to the biological invasive herbs. PLoS ONE 7, e35873, DOI: 10.1371/ efficacy of glyphosate on easy- and difficult-to-wet weed journal.pone.0035873. species. Pest Management Science 65, 844–850. ZHU XC, WU HW, RAMAN H et al. (2012) Evaluation of LEITE GLD, PICANCO M, ZANUNCIO JC & MARQUINI F (2003) simple sequence repeat (SSR) markers from Solanum crop Factors affecting mite herbivory on eggplants in Brazil. species for Solanum elaeagnifolium. Weed Research 52, Experimental and Applied Acarology 31, 243–252. 217–223.

Appendix 1 The total rainfall of growing season, sampling size and locations of Solanum elaeagnifolium collected from different sate of Australia: New South Wales (NSW), South Australia (SA), Victoria (VIC), Queensland (QLD) and Western Australia (WA). A number of samples used for trichome density observation (TO) and for stomata observation (SO) were also included. Rainfall in the growing season from September 2009 to February 2010: H = total rainfall > 300mm, M = total rainfall between 200 and 300, and L = total rainfall < 200 (See Fig. 1).

Location State Rainfall Longitude/Latitude Sample size TO SO

Adelaide SA L À34°40/138°41 11 0 0 Angas Valley SA L À34°44/139°19 10 0 0 Annadale SA L À34°24/139°21 2 3 1 Appila 1 SA M À33°01/138°26 6 1 1 Appila 2 SA M À33°00/138°28 2 1 1 Avon SA L À34°15/138°20 10 0 0 Balranald NSW M À34°56/143°28 2 0 0 Bingara 1 NSW H À29°52/150°33 9 0 0 Bingara 2 NSW H À29°48/150°32 4 0 0 Bingara 3 NSW H À29°49/150°32 10 0 0 Blyth SA M À33°50/138°30 1 1 1 Boree Creek NSW L À35°08/146°27 4 0 0 Bridgewater VIC M À36°38/143°54 10 3 3 Burra SA M À33°41/138°55 4 2 2 Calivil 1 VIC M À36°21/144°07 5 1 1 Calivil 2 VIC M À36°17/144°05 12 1 1 Cambrai SA L À34°39/139°15 2 0 0 Cartwrights Hill NSW M À34°56/147°25 4 0 0 Carwarp VIC L À34°28/142°10 5 0 0 Clare SA M À33°43/138°37 9 2 2 Coonabarabran NSW H À31°05/149°33 8 0 0 Corowa NSW M À35°53/146°18 10 3 3 Crystal Brook SA M À33°19/138°12 6 0 0 Culcairn NSW M À35°41/146°58 10 3 1 Delungra NSW H À29°45/150°42 6 0 0 Dimboola VIC L À36°25/142°00 3 0 0 Dookie 1 VIC M À36°13/145°40 4 2 0 Dookie 2 VIC M À36°12/145°42 6 2 1 Dubbo NSW H À32°11/148°48 10 0 0 Dunedoo NSW H À31°58/149°30 12 0 0 Echuca VIC M À36°07/144°52 10 4 3 Eudunda SA L À34°11/139°05 9 0 0 Finley NSW M À35°37/145°35 10 0 0 Ganmain NSW M À34°53/146°59 6 0 0 Gilgandra NSW H À31°40/148°42 4 0 0 Grifﬁth NSW L À34°26/146°11 6 0 0 Gulgong NSW H À32°23/149°36 10 0 0 Hay NSW L À34°29/145°17 3 0 0 Hopetoun 1 VIC L À35°36/142°26 10 0 0 Hopetoun 2 VIC L À35°31/142°22 5 0 0 Inglewood QLD H À29°05/151°17 2 3 3 Inverell NSW H À29°39/151°12 10 0 0 Jarklin 1 VIC M À36°16/143°58 3 0 0 Jarklin 2 VIC M À36°14/143°56 8 4 0 Keith 1 SA L À36°06/140°16 10 0 0 Keith 2 SA L À36°04/140°17 10 0 0 Keith 3 SA L À36°06/140°21 3 0 0 Koonoona SA M À33°49/138°56 10 0 0 Lake Boga VIC M À35°28/143°39 10 4 0 Langhorne Creek SA L À35°19/139°00 8 0 0 Leeton NSW L À34°27/146°22 5 0 0 Lochiel SA L À33°57/138°10 2 0 0 Longerenong VIC M À36°40/142°18 10 0 0 Loxton 1 SA L À34°28/140°37 10 0 0

Appendix 1 (Continued)

Location State Rainfall Longitude/Latitude Sample size TO SO

Loxton 2 SA L À34°38/140°41 10 0 0 Mangalo SA L À33°29/136°31 5 0 0 Mannum SA L À35°00/139°14 10 1 1 Mitchelville SA L À33°35/137°04 5 0 0 Morven NSW M À35°35/147°09 10 3 3 Mount Priscilla SA L À33°46/136°24 5 0 0 Mudgee NSW H À32°31/149°33 10 0 0 Murray Bridge SA L À35°04/139°13 10 0 0 Nanneella VIC M À36°20/144°49 3 2 0 Narrandera NSW L À34°46/146°25 7 0 0 Nhill 1 VIC M À36°24/141°27 1 0 0 Nhill 2 VIC L À36°24/141°49 9 2 0 Parkes NSW H À33°13/148°13 11 2 2 Port Pirie SA M À33°16/138°09 9 3 0 Red Cliffs VIC L À34°24/142°00 7 0 0 Rochester VIC M À36°23/144°46 9 2 2 Scone NSW H À31°58/150°51 7 0 0 Sedan SA L À34°33/139°18 3 1 1 Serpentine VIC M À36°24/143°58 10 3 0 Shepparton VIC M À36°25/145°27 10 5 3 Snowtown SA M À33°44/138°05 10 0 0 Spalding SA M À33°19/138°35 2 1 0 Swan Hill VIC M À35°19/143°31 2 1 0 Tamworth NSW H À31°03/150°51 6 0 0 Tarlee SA M À34°12/138°43 1 0 0 Temora NSW M À34°24/147°36 10 0 0 Ungarie 1 NSW M À33°39/146°59 12 0 0 Ungarie 2 NSW M À33°38/146°58 5 0 0 Ungarie 3 NSW M À33°36/146°55 10 0 0 Walpeup VIC L À35°09/142°03 5 0 0 Wellington NSW H À32°31/148°48 10 0 0 West Wyalong NSW M À34°00/147°15 2 0 0 Wirrabara SA H À33°02/138°16 9 4 3 Wunghnu VIC M À36°10/145°28 10 3 1 Wunkar SA L À34°29/140°12 10 1 1 Yanco 1 NSW L À34°38/146°25 2 0 0 Yanco 2 NSW L À34°34/146°23 3 0 0 Young NSW H À34°27/148°19 11 3 0 Total 642 77 41

The SLN morphological variation detected in Chapter 3 suggested the requirement for studies on genetic diversity. Therefore, Chapters 4 and 5 aimed to develop genetic markers for investigating SLN genetic diversity.

Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., Stanton, R., & Burrows, G.

E. (2012). Evaluation of simple sequence repeat (SSR) markers from

Solanum crop species for Solanum elaeagnifolium. Weed Research, 52(3),

217-223. doi: 10.1111/j.1365-3180.2012.00908.x.

43 DOI: 10.1111/j.1365-3180.2012.00908.x

Evaluation of simple sequence repeat (SSR) markers from Solanum crop species for Solanum elaeagnifolium

XCZHU* ,HWWU*à,HRAMAN*à, D LEMERLE* ,RSTANTON* & G E BURROWS* *EH Graham Centre for Agricultural Innovation (An Alliance Between NSW Department of Primary Industries and Charles Sturt University), School of Agricultural and Wine Sciences, Charles Sturt University, Wagga Wagga, NSW, Australia, and àWagga Wagga Agricultural Institute, PMB, Wagga Wagga, NSW, Australia

Received 20 April 2011 Revised version accepted 24 January 2012 Subject Editor: Stephen Novak, Boise, USA

Summary and eggplant to S. elaeagnifolium was 20%, 40% and 46% respectively. SSR analysis revealed high level of Molecular markers speciﬁc for Solanum elaeagnifolium genetic diversity among 40 individuals collected within a (silverleaf nightshade) are currently not available. A total paddock. Highly polymorphic and transferable cross- of 35 simple sequence repeat (SSR) primer pairs from species SSR markers would be useful for determining potato (Solanum tuberosum), tomato (S. lycopersicum) the extent of genetic diversity in S. elaeagnifolium and eggplant (S. melongena) were tested for cross-species populations. transferability in S. elaeagnifolium. Among them, 13 primer pairs successfully produced alleles (bands). The Keywords: silverleaf nightshade, invasive weed, cross- polymorphism information content ranged from 0 to species SSR, microsatellite, genetic diversity. 0.84. The transferable rate of SSR from potato, tomato

ZHU XC, WU HW, RAMAN H, LEMERLE D, STANTON R & BURROWS GE (2012). Evaluation of simple sequence repeat (SSR) markers from Solanum crop species for Solanum elaeagnifolium. Weed Research 52, 217–223.

Introduction competes with pastures and other crops for soil water and nutrients and causes up to 77% reduction in Solanum elaeagnifolium Cav. (silverleaf nightshade) is a cereal crop yields (Stanton et al., 2009). Current diploid (2n = 2x = 24), deep-rooted, summer growing management strategies are ineffective and unreliable, perennial that originated in south-western United States especially for dense and large infestations (Wasser- and northern Me´xico (Stanton et al., 2009). It repro- mann et al., 1988). Improved management of this weed duces both sexually (obligate outcrossing) and vegeta- would require a better understanding of genetic tively (Hardin et al., 1972). Solanum elaeagnifolium has diversity in S. elaeagnifolium, because genetically been introduced around the world (Stanton et al., 2009). diverse weed species will affect the choice of appro- It was first reported in Australia in 1901 and gradually priate control strategies, such as the selection of spread over New South Wales (NSW), Victoria (VIC) biocontrol agents (Dekker, 1997). and South Australia (SA) (Stanton et al., 2009). This Molecular markers such as simple sequence repeat invasive weed currently infests at least 350 000 hectares (SSR), amplified fragment length polymorphism in Australia, with the potential to infest 398 million (AFLP) and random amplified polymorphic DNA hectares (Feuerherdt, 2009). Solanum elaeagnifolium (RAPD) have been widely used in the assessment of

Correspondence: Xiaocheng Zhu, School of Agricultural and Wine Sciences, Charles Sturt University, Wagga Wagga, NSW 2678, Australia. Tel: (+61) 2 6933 2749; Fax: (+61) 2 6938 1861; E-mail: [email protected]

2012 The Authors 44 218 X C Zhu et al. genetic diversity in weeds (OÕHanlon et al., 2000). GTT GTA AAA CGA C-3¢) was tailed to the 5¢ end of the However, there is limited information available on the forward primer of each SSR primer pair (Raman et al., level of genetic variation in S. elaeagnifolium. Hawker 2005). et al. (2006) employed RAPD markers to investigate PCR amplification was carried out in 12 lLof genetic diversity of S. elaeagnifolium populations in SA reaction mixture consisting of 50–100 ng of template and found high levels of variation. SSR markers have DNA, 1.2 lLof10· buffer, 0.24 lLof25mM MgCl2, )1 been shown to be much more polymorphic and repro- 1.2 lLof2mM dNTPÕs, 0.08 lL of Taq (5 units lL ) ducible between laboratories than RAPD markers (Promega, Australia), 0.05 lL of forward primer )1 )1 (McGregor et al., 2000). Molecular marker resources (3 lM lL ), 0.1 lL of reverse primer (3 lM lL ), )1 specific to S. elaeagnifolium have not been developed 0.15 lL of M13 (2 lM lL ) labelled with one of the worldwide. However, SSR primers sequences have fluorescent dyes (D2, D3 or D4; Beckman Coulter, Brea, been published and applied in genetic diversity analy- CA, USA). After an initial denaturation of 4 min at ses of some members of Solanum: potato (Solanum 94C, 30 cycles of 1 min at 94C, 1 min at 45–58C tuberosum L.), tomato (S. lycopersicum L.) and eggplant (depending on the primers, Tables 1 and S1) and 1 min (S. melongena L.). at 72C were performed, followed by a final extension of In this study, the transferability of cross-species SSR 10 min at 72C. The amplifications were carried out in a markers derived from potato, tomato and eggplant to Gene Amp PCR System 2700 (Applied Biosystems, S. elaeagnifolium was evaluated, and the suitability of Singapore). PCR products were separated on a CEQ these markers for genetic diversity analysis validated 8000 genetic analysis system (Beckman Coulter Inc.) as using S. elaeagnifolium plants collected from a single described previously (Raman et al., 2005). Individuals paddock in NSW. These SSR markers could be used in with null alleles were confirmed by at least three different the future to investigate the genetic diversity of PCR amplifications. In addition, null alleles detected for S. elaeagnifolium throughout Australia and elsewhere single-locus SSRs were further checked using the soft- around the world. ware program Micro-Checker, set at the 95% confidence interval (Van Oosterhout et al., 2004). Materials and methods Data analysis Plant material Scoring of SSRs was based on their amplified fragment Young leaf samples from 40 individuals were collected size (base pairs). The PIC value of each SSR locus was following a zigzag pattern (Menchari et al., 2007) from a calculated by PowerMarker V3.0 software (Liu & Muse, paddock heavily infested with S. elaeagnifolium at 4–5 2005), while the observed and expected heterozygosity plants m)2 in Ungarie, central west NSW [latitude (N): were calculated using POPGENE version 1.32 (Yeh & )3536¢, longitude (E): 14655¢]. Individuals were sam- Boyle, 1997). For each fragment size, binary scores (1 pled at least 50 m apart, to reduce the probability of for present, 0 for absent) were assigned for each allele to sampling clonal material. Leaves were collected from each calculate genetic similarity matrices. plant and placed in 1.5 mL Eppendorf tubes and then A similarity matrix for all test samples was calculated stored at )80C in the laboratory until DNA extraction. by JaccardÕs coefficient using Similarity for Qualitative data (SIMQUAL) in NTSYS-pc 2.1 (Rohlf, 2000). A dendrogram based on this matrix was computed to DNA extraction illustrate genetic relationships among individuals. This Genomic DNA was isolated from leaf material using a was performed by the Unweighted Pair Group Method standard phenol ⁄ chloroform method (Sambrook et al., with Arithmetic mean (UPGMA) using the SAHN 1989). Quality of DNA was checked by electrophoresis procedures of the same software. The relationship (MI-DEAR, SYS-MD 120, USA) in a 1.0% agarose gel between the similarity and cophenetic (ultrametric) at 200 V for 10 min. matrices was determined by Mantel test using the matrix comparison plot based on the product-moment correlation, r, and 1000 permutations were used in the Mantel PCR and SSR analysis test. Thirty-five SSR primer pairs (synthesised by Sigma Aldrich, Sydney, Australia) were selected on the basis of Results their high polymorphism information content (PIC) values in the source species (Tables 1 and S1). A Thirteen of the 35 tested markers (37%) generated bands nineteen-nucleotide-long M13 sequence (5¢-CAC GAC (Table 1). Primers from eggplant had the highest levels of

2012 The Authors 45 SSR marker selection for Solanum elaeagnifolium 219

Table 1 Cross-species simple sequence repeat (SSR) primers that generated amplicons in S. elaeagnifolium, including eight producing results consistent with single-locus SSRs and ﬁve consistent with multi-locus SSRs

Repeat Source

Locus motif species PICSO TA (C) Primer sequence (5¢–3¢) Reference

Single-locus ampliﬁcation

STI001 (AAT)n Potato 0.69 58 F: CAGCAAAATCAGAACCCGAT Ghislain et al., R: GGATCATCAAATTCACCGCT (2009) CA158 (GA)32 Tomato 0.85 55 F: CATGCACGTACAACCTGTTT Martins et al., R: TAGTTCCCTTGCTGCAGTAA (2006) SSR63 (AT)39 Tomato 0.80 50 F: CCACAAACAATTCCATCTCA Kwon et al., R: GCTTCCGCCATACTGATACG (2009) SSR111 (TC)6(TCTG)6 Tomato 0.88 50 F: TTCTTCCCTTCCATCAGTTCT Kwon et al., R: TTTGCTGCTATACTGCTGACA (2009) TSR2 (AT)15 Tomato 0.81 50 F: TCAAGTGAGTTTATCTGCCCAC Yi et al., R: GCTCATCCTACACATTCATGCTC (2008) EM135 (CA)11(GA)20 Eggplant 0.75 58 F: ATCCTGTTGCTGCTCATTTTCCTC Nunome R: AGGAGGATCCAAGAGGTTTGTTGA et al., (2003) EM140 (AC)4GC(AC)5T(AC) Eggplant 0.52 53–48 F: CCAAAACAATTTCCAGTGACTGTGC Munoz-Falcon 3ATGC(AC)4AT(AC) R: GACCAGAATGCCCCTCAAATTAAA et al., (2009) 6(AT)5G(TA)13 EM141 (AT)16(GT)19 Eggplant 0.83 50 F: TCTGCATCGAATGTCTACACCAAA Nunome et al., R: AAAAGCGCTTGCACTACACTGAAT (2003) Multi-locus ampliﬁcation

STG10 (TG)n Potato 0.69 55–50 F: CGATCTCTGCTTTGCAGGTA Ghislain et al., R: GTTCATCACTACCGCCGACT (2009) EM117 (AC)19(AT)11 Eggplant 0.74 55 F: GATCATCACTGGTTTGGGCTACAA Nunome et al., R: AGGGGAGAGGAAACTTGATTGGAC (2003) EM127 (AC)13(AT)13 Eggplant 0.60 55–50 F: CAGACACAACTGCTGAGCCAAAAT Munoz-Falcon R: CGGTTTAATCATAGCGGTGACCTT et al., (2009) EM155 (CT)38 Eggplant 0.64 50–45 F: CAAAAGATAAAAAGCTGCCGGATG Munoz-Falcon R: CATGCGTGAGTTTTGGAGAGAGAG et al., (2009) ESM3 (TA)9(GA)8 Eggplant 0.51 55–50 F: ATTGAAAGTTGCTCTGCTTCAC Munoz-Falcon R: ACATCGTTCCGCCTCTATTG et al., (2009)

PICSO, polymorphism information content (PIC) in source species; TA, annealing temperature used in this study. transferability (46%, 7 ⁄ 15) to S. elaeagnifolium, followed Both product-moment correlation, r (0.75), and the by tomato (40%, 4 ⁄ 10) and potato (20%, 2 ⁄ 10) Mantel test statistic, Z (t = 9.55), were highly signifi- (Table S1). A total of 88 bands were amplified from these cant (P < 0.01) for the clustering shown in Fig. 1. 13 SSR loci, ranging from one for the SSRs TSR2 According to the UPGMA dendrogram, two main and EM141 to 21 for the SSR EM117 (Table 2). Eight clusters were defined (Fig. 1). Seven individuals were of these SSR primer pairs produced 1 or 2 discrete grouped into Custer A and the remaining 33 individuals bands and were considered as single-locus markers. formed Cluster B. Among them, two (TSR2 and EM141) were monomorphic. Amplification of five primer pairs (STG10, EM117, Discussion EM127, EM155 and ESM3) resulted in the amplification of three or more bands, consistent with multilocus gene Transferability of SSR markers expression; therefore, these bands were scored as present or absent. The observed heterozygosity ranged from 0 to This study is the first report evaluating cross-species 0.85 and the expected heterozygosity from 0 to 0.87. The SSR markers in S. elaeagnifolium. A total of 13 cross- PIC value, a measure of allelic diversity, varied from 0 for species SSR markers amplified DNA fragments in this the primers TSR2 and EM141 to 0.84 for the primer weed. The rate of SSR transferability to S. elaeagni- CA158. folium ranged from 20% to 46%, depending on the JaccardÕs genetic similarity coefficients among the 40 source Solanum species. Higher transferability of SSR individuals analysed in this study ranged from 0.15 to markers was observed between S. elaeagnifolium and 0.79, with the mean value of 0.37 (Table S2), indicating eggplant and tomato, as compared with potato. It high genetic diversity among the individuals within this is possible that closely related species may have paddock. similar primer binding sites (Rossetto, 2001). Both

2012 The Authors 46 220 X C Zhu et al.

Table 2 Comparison of information provided and band size of 13 simple sequence repeat (SSR) markers between S. elaeagnifolium and source species, including eight producing results consistent with single-locus SSRs and ﬁve consistent with multi-locus SSRs

Fluorescent

Locus NA ⁄ NB HO HE PICSE FSSO FSSE Dye used

Single-locus ampliﬁcation STI001 2 0.23 0.20 0.18 185–208 205–211 D3 CA158 13 0.43 0.87 0.84 198–250 217–249 D3 SSR63 2 0 0.18 0.16 250 Null–183 D2 SSR111 4 0 0.54 0.49 188 Null–179 D3 TSR 2 1 0 0 0 219–301 286 D3 EM135 11 0.85 0.75 0.70 260 233–262 D4 EM140 5 0 0.62 0.54 277–290 Null–223 D4 EM141 1 0 0 0 228 184 D4

Multi-locus ampliﬁcation STG10 5 – – – 175–192 177–271 D4 EM117 21 – – – 123 120–172 D2 EM127 12 – – – 200–210 Null–294 D4 EM155 4 – – – 232–264 113–298 D3 ESM3 7 – – – 230–243 Null–351 D4 Mean 6.77

NA, number of alleles in S. elaeagnifolium;NB, number of bands in S. elaeagnifolium; observed (HO) and expected (HE) heterozygosity; PICSE, average PIC of each marker in S. elaeagnifolium;FSSO, fragment size of source species; FSSE, fragment size of S. elaeagnifolium.

1 38 2 35 37 3 29 15 32 31 21 24

13 B Cluster 22 4 23 6 33 17 34 7 18 14 19 26 36 39 8 28 5 16 25 30

9 A Cluster 10 12 11 40 20 27 0.29 0.39 0.49 0.59 0.70 0.80 0.90 1.00 Jaccard's coefficient

Fig. 1 Unweighted pair group method with arithmetic mean (UPGMA) based dendrogram showing the genetic variation of 40 individual samples of S. elaeagnifolium collected from Ungarie, NSW, Australia.

S. elaeagnifolium and eggplant are located in the et al. (2008) found a low (27%) transferability rate of Leptostemonum clade of Solanum, while tomato and SSR markers from potato to naranjilla (Solanum quito- potato are located in the Potato clade, based on ndhF, ense Lam.). However, the within-subgenus transferable trnTF and waxy DNA sequence data (Weese & Bohs, rate of SSR markers in these two Solanum studies is 2007). Varying rates of cross-species SSR transferability much lower than in other species, such as in Magnolia have been reported previously. For example, Torres (90.9%) and Vitis (93.5%) (Rossetto, 2001). Solanum,as

2012 The Authors 47 SSR marker selection for Solanum elaeagnifolium 221 one of the largest genera of flowering plants, might have of 0.37. Two main groups were clustered according to a undergone a long evolutionary process or have a much UPGMA dendrogram (Fig. 1). The predominant ÔClus- higher speciation rate than other species (Whalen & ter BÕ contains 82.5% (33) of the individuals, distributed Caruso, 1983). These factors may cause genetic diver- across the paddock. This may be a reflection that the gence among Solanum species, which may be one of the spread of S. elaeagnifolium within a paddock can be reasons leading to poor amplification when using cross- assisted by cultivation practices and grazing animals species genetic markers (Whalen & Caruso, 1983; Torres (Stanton et al., 2009). ÔCluster AÕ consisted seven indi- et al., 2008). The results presented here suggest that viduals which were also distributed across the paddock. eggplant and tomato are reliable sources of cross-species No specific grouping of individual samples was observed SSRs for S. elaeagnifolium. according to GPS location (data not shown). ÔCluster AÕ may be indicative of the range of genetic diversity among individuals in this paddock. Alternatively, these Information provided by markers individuals may have been introduced into the paddock Differences in the level of SSR polymorphism (0–0.84) from elsewhere. Dispersal of S. elaeagnifolium from one were observed (Table 2). This may be attributed to gene paddock to another may occur through seed-contami- conservation between source species and S. elaeagnifo- nated fodder, agricultural produce and ⁄ or farm machin- lium, source of the SSRs (genomic and EST-SSR) ery (Stanton et al., 2009). Differentiating between these and ⁄ or nature of the SSRs (nucleotide repeat unit, such two alternative scenarios requires further genetic anal- as di-, tri- and tetra). These results show that the high ysis of additional paddocks. The high level of genetic PIC values of SSR markers in source species (Tables 1 variation identified in this study is consistent with the and S1) were not preserved in S. elaeagnifolium. For report of Hawker et al. (2006). It is believed that a high example, the SSR markers of high PIC values in source level of genetic diversity will contribute to the adapta- species, such as STI001, SSR63, TSR 2 and EM141, did tion of weed species to various environments and not possess similarly high PIC values in S. elaeagnifo- contribute to the capacity of weeds to respond to lium (Tables 1 and 2). Similarly, it has also been selection pressures, reducing the effectiveness of weed reported that many SSR markers when transferred from management (Dekker, 1997). For example, differential Cirsium acaule (L.) Scop. and Zostera marina L. to other herbicide responses have been reported in many weed Cirsium and Zostera species, respectively, did not species, such as Alopecurus myosuroides Huds. (black- produce similar levels of polymorphism (Reusch, 2000; grass) (Marshall & Moss, 2008) and Amaranthus retro- Jump et al., 2002). This phenomenon suggests that a flexus L. (redroot pigweed) (Scarabel et al., 2007). A highly informative marker from a source species does genetically diverse weed species could also limit the not necessarily lead to high levels of polymorphism in a effectiveness of biocontrol. The biocontrol agent, Puc- test species. In addition, some of these 13 SSR loci did cinia chondrillina, showed differential pathogenicity not produce similar sized alleles in this investigation between genotypes of Chondrilla juncea L. (Burdon compared with those in the source species (SSR63, et al., 1984). Therefore, the high genetic variation of EM140, EM141, EM155, EM127 and ESM3) (Table 2). S. elaeagnifolium may represent a challenge to effectively Differences in the size of alleles in the source species and managing this weed using biological control. the individuals analysed in this study may be attributed In conclusion, this research demonstrates the trans- to chromosomal rearrangements during the evolution of ferability of 13 cross-species SSR markers for genetic the S. elaeagnifolium genome, or strand slippage during research in S. elaeagnifolium. The highly polymorphic DNA replication. SSRs identified in this study can be used for genetic We considered five SSRs that generated multiple mapping, genetic diversity analysis and molecular bands as multilocus markers. Amplification of multiple evolution studies. These SSR markers are currently alleles might be attributed to the divergence and ⁄ or being employed to determine the genetic diversity of duplication of genomic regions (Senthilvel et al., 2008), S. elaeagnifolium collected from different regions of which has been detected in many plants such as Australia. To this end, amplicons from multi-locus pearl millet [Pennisetum glaucum (L.) R.Br.] and tea SSRs could be sequenced and then the flanking (Camellia sinensis L.) (Senthilvel et al., 2008; Sharma sequence of SSRs could be used to generate single- et al., 2009). locus markers.

Genetic variation among Solanum elaeagnifolium Acknowledgements A high level of genetic variation was identiﬁed in the We would like to acknowledge the Charles Sturt population studied, having an average genetic similarity University, Australia, for funding this research.

2012 The Authors 48 222 X C Zhu et al.

We thank Mr Robert Thompson of NSW Department weed of winter cereal fields. Molecular Ecology 16, 3161– of Primary Industries for his assistance in field sampling. 3172. MUNOZ-FALCON JE, PROHENS J, VILANOVA S et al. (2009) Distinguishing a protected geographical indication References vegetable (Almagro eggplant) from closely related varieties with selected morphological traits and molecular BURDON JJ, GROVES RH, KAYE PE & SPEER SS (1984) markers. Journal of the Science of Food and Agriculture 89, Competition in mixtures of susceptible and resistant geno- 320–328. types of Chondrilla juncea differentially infected with rust. NUNOME T, SUWABE K, IKETANI H&HIRAI M (2003) Oecologia 64, 199–203. Identification and characterization of microsatellites in DEKKER J (1997) Weed diversity and weed management. Weed eggplant. Plant Breeding 122, 256–262. Science 45, 357–363. OÕHANLON PC, PEAKALL R&BRIESE DT (2000) A review of FEUERHERDT L (2009) Overcoming a deep rooted perennial new PCR-based genetic markers and their utility to weed problem – silverleaf nightshade (Solanum elaeagnifolium) ecology. Weed Research 40, 239–254. in South Australia. Plant Protection Quarterly 24, 123– PROVAN J, POWELL W&WAUGH R (1996) Microsatellite 124. analysis of relationships within cultivated potato (Solanum GARCIA-MARTINEZ S, ANDREANI L, GARCIA-GUSANO M, tuberosum). Theoretical and Applied Genetics 92, 1078–1084. R R J GEUNA F&RUIZ JJ (2006) Evaluation of amplified fragment AMAN R, AMAN H, OHNSTONE K et al. (2005) Genetic and in length polymorphism and simple sequence repeats for silico comparative mapping of the polyphenol oxidase gene tomato germplasm fingerprinting: utility for grouping closely in bread wheat (Triticum aestivum L.). Functional & Inte- related traditional cultivars. Genome 49, 648–656. grative Genomics 5, 185–200. R GHISLAIN M, NUNEZ J, HERRERA MR et al. (2009) Robust and EUSCH TBH (2000) Five microsatellite loci in eelgrass Zostera highly informative microsatellite-based genetic identity kit marina and a test of cross-species amplification in Z. noltii for potato. Molecular Breeding 23, 377–388. and Z. japonica. Molecular Ecology 9, 371–373. R HARDIN JW, DOERKSON G, HERNDON D, HOBSON M&THOMAS OHLF FJ (2000) NTSYS-pc: Numerical Taxonomy and Mul- F (1972) Pollination and floral biology of four weedy tivariate Analysis System. Version 2.1. Exeter Publications, species in southern Oklahoma. Southwest Naturalist 16, New York, USA. 403–412. ROSSETTO M (2001) Sourcing of SSR markers from related HAWKER V, PRESTON C&BAKER J (2006) Genetic variation plant species. In: Plant Genotyping: The DNA Fingerprinting within and among silverleaf nightshade (Solanum elaeag- of Plants, Vol. 1 (ed RJ HENRY), 211–224. CAB Interna- nifolium Cav.) populations in South Australia. In: 15th tional, Wallingford, UK. S F M Australian Weeds Confere (eds C PRESTON,JHWATTS &ND AMBROOK J, RITSCH EF & ANIATIS T eds. (1989) Molecular CROSSMAN), 176. Adelaide Conference Centre, Adelaide, Cloning: A Laboratory Manual, 2nd edn. Cold Spring South Australia. Harbor Laboratory, New York, USA. S V S JUMP AS, DAWSON DA, JAMES CM, WOODWARD FI & BURKE T CARABEL L, AROTTO S& ATTIN M (2007) A European (2002) Isolation of polymorphic microsatellites in the stem- biotype of Amaranthus retroflexus cross-resistant to ALS less thistle (Cirsium acaule) and their utility in other Cirsium inhibitors and response to alternative herbicides. Weed species. Molecular Ecology Notes 2, 589–592. Research 47, 527–533. S J M KWON YS, PARK SG & YI SI (2009) Assessment of genetic ENTHILVEL S, AYASHREE B, AHALAKSHMI V et al. (2008) variation among commercial tomato (Solanum lycopersicum Development and mapping of simple sequence repeat L.) varieties using SSR markers and morphological charac- markers for pearl millet from data mining of expressed teristics. Genes & Genomics 31, 1–10. sequence tags. Bmc Plant Biology 8, 119. S B N M A LIU KJ & MUSE SV (2005) PowerMarker: an integrated HARMA RK, HARDWAJ P, EGI R, OHAPATRA T& HUJA analysis environment for genetic marker analysis. Bioinfor- PS (2009) Identification, characterization and utilization of matics 21, 2128–2129. unigene derived microsatellite markers in tea (Camellia MARSHALL R&MOSS SR (2008) Characterisation and molec- sinensis L.). Bmc Plant Biology 9, 53. ular basis of ALS inhibitor resistance in the grass weed STAGEL A, PORTIS E, TOPPINO L, ROTINO GL & LANTERI S Alopecurus myosuroides. Weed Research 48, 439–447. (2008) Gene-based microsatellite development for mapping MARTINS K, CHAVES LJ, BUSO GSC & KAGEYAMA PY (2006) and phylogeny studies in eggplant. BMC Genomics 9, 357– Mating system and fine-scale spatial genetic structure of 370. Solanum lycocarpum St.Hil. (Solanaceae) in the Brazilian STANTON RA, HEAP JW, CARTER RJ & WU H (2009) Solanum Cerrado. Conservation Genetics 7, 957–969. elaeagnifolium. In: The Biology of Australian weeds, Vol. 3. MCGREGOR CE, LAMBERT CA, GREYLING MM, LOUW JH & (ed. FD PANETTA), 274–293. R. G. and F. J. Richardson, WARNICH L (2000) A comparative assessment of DNA Melbourne. fingerprinting techniques (RAPD, ISSR, AFLP and SSR) in TORRES AF, ARIAS AS, ARAHANA V&TORRES ML (2008) tetraploid potato (Solanum tuberosum L.) germplasm. Eu- Preliminary assessment of genetic diversity and phenetic phytica 113, 135–144. relations for section Lasiocarpa by means of heterologous MENCHARI Y, DELYE C&LE CORRE V (2007) Genetic variation SSR markers. Crop Science 48, 2289–2297. and population structure in black-grass (Alopecurus myos- VAN OOSTERHOUT C, HUTCHINSON WF, WILLS DPM & uroides Huds.), a successful, herbicide-resistant, annual grass SHIPLEY P (2004) Micro-checker: software for identifying

2012 The Authors 49 SSR marker selection for Solanum elaeagnifolium 223

and correcting genotyping errors in microsatellite data. Supporting Information Molecular Ecology Notes 4, 535–538. WASSERMANN VD, ZIMMERMANN HG & NESER S (1988) The Additional supporting information may be found in the Weed Silverleaf Bitter Apple (ÔSatansbosÕ) (Solanum online version of this article. elaeagnifolium Cav.) with Special Reference to its Status in Table S1 Cross-species SSR primers not transferable to South Africa. Plant Protection Research Institute, Pretoria. S. elaeagnifolium. PIC , polymorphism information W B SO EESE TL & OHS L (2007) A three-gene phylogeny of content (PIC) in source species; T , annealing temper- the genus Solanum (Solanaceae). Systematic Botany 32, A ature used in this study. 445–463. WHALEN MD & CARUSO EE (1983) Phylogeny in Solanum sect. Table S2 Matrix of genetic similarity between individ- Lasiocarpa (Solanaceae): congruence of morphological and uals based upon JaccardÕs coefﬁcient. molecular data. Systematic Botany 8, 369–380. Please note: Wiley-Blackwell is not responsible for YEH FC & BOYLE TJB (1997) Population genetic analysis of co- the content or functionality of any supporting materials dominant and dominant markers and quantitative traits. supplied by the authors. Any queries (other than missing Belgian Journal of Botany 129, 157. material) should be directed to the corresponding author YI SS, JATOI SA, FUJIMURA T et al. (2008) Potential loss of for the article. unique genetic diversity in tomato landraces by genetic colonization of modern cultivars at a non-center of origin. Plant Breeding 127, 189–196.

2012 The Authors 50

CHAPTER 5 PUBLISHED PAPER

Zhu, X., Raman, H., Wu, H., Lemerle, D., Burrows, G., & Stanton, R.

(2013). Development of SSR markers for genetic analysis of silverleaf nightshade (Solanum elaeagnifolium) and related species. Plant Molecular

Biology Reporter, 31(1), 248-254. doi: 10.1007/s11105-012-0473-z.

51 Plant Mol Biol Rep (2013) 31:248–254 DOI 10.1007/s11105-012-0473-z

BRIEF COMMUNICATION

Development of SSR Markers for Genetic Analysis of Silverleaf Nightshade (Solanum elaeagnifolium) and Related Species

Xiao Cheng Zhu & Harsh Raman & Hanwen Wu & Deirdre Lemerle & Geoffrey E. Burrows & Rex Stanton

Published online: 31 May 2012 # Springer-Verlag 2012

Abstract Silverleaf nightshade (2n02x024) is a serious continent. It occurs in many countries including the USA, weed in the Solanaceae, for which no specific molecular Australia, Argentina, Brazil, Chile, India, Israel, Greece, markers are currently available. In order to investigate the Morocco, South Africa and Spain (Stanton et al. 2009). This extent and distribution of genetic diversity among accessions invasive weed reproduces both sexually (outcrossing) and of silverleaf nightshade, we developed 23 simple sequence vegetatively. In Australia, silverleaf nightshade can cause up repeat (SSR) markers from publicly available nucleotide and to 77 % yield lost in cereal crops (Stanton et al. 2009). EST databases for silverleaf nightshade. Eleven of them were However, little is known about its pollination biology and single-locus polymorphic markers. The number of alleles genetic diversity. Improved management of this weed among these loci ranged from 2 to 4. The observed and requires a better understanding of genetic diversity in silver- expected heterozygosity ranged from 0 to 0.97 and 0.07 to leaf nightshade since genetically diverse weed species will 0.64, respectively. Fourteen SSR markers enabled to amplify affect the choice of appropriate control strategies, such as alleles in morphologically similar species quena. These results the selection of biocontrol agents (Dekker 1997). proved that the SSR markers that we developed could be useful Silverleaf nightshade is often morphologically confused for (1) determining genetic diversity and structure among with an Australian native species quena (Solanum esuriale natural populations of silverleaf nightshade and (2) identifying Lindl.), and microscopic examination is required to distin- silverleaf nightshade and quena ecotypes. This is the first set of guish these two species (Bean 2004; Zhu et al. 2011). Quena species-specific SSR markers identified in silverleaf night- is noninvasive and easier to control than silverleaf nightshade; shade, which could contribute to the better understanding of thus, correct identification is critical for silverleaf nightshade genetic diversity of silverleaf nightshade and related species. management (Johnson et al. 2006). Currently, there is limited genomic resource available for silverleaf nightshade. Several Keywords EST-SSR . Genomic SSR . Weed . Silverleaf marker systems such as restriction fragment length polymor- nightshade . Quena phism, amplified fragment length polymorphism, random amplified polymorphism DNA, Diversity Array Technology, simple sequence repeat (SSR) and single-nucleotide polymor- Introduction phism have been utilised for genetic analysis of various plant species (Zhou 2005). SSRs have been the marker of choice Silverleaf nightshade (Solanum elaeagnifolium Cav.) is a due to their abundance, high rate of polymorphism and repro- summer-growing perennial weed native to the American ducibility, high transferability across species, codominance, genetic stability and suitability for higher throughput analysis * : : : : using highly parallel automated systems (Ellis and Burke X. C. Zhu ( ): H. Raman H. Wu D. Lemerle G. E. Burrows R. Stanton 2007;O’Hanlon et al. 2000; Swapna et al. 2011). EH Graham Centre for Agricultural Innovation (an alliance Many SSR markers have been developed and widely used between Charles Sturt University and NSW Department of in Solanum species for genetic diversity studies (Kwon et al. Primary Industries), Wagga Wagga, NSW 2650, Australia 2009) and species and variety identification, especially in the e-mail: [email protected] three main Solanum crops: eggplant (Solanum melongena L.),

52 Plant Mol Biol Rep (2013) 31:248–254 249 potato (Solanum tuberosum L.) and tomato (Solanum lycoper- database (www.ncbi.nlm.nih.gov). Vector sequences (if sicum L.) (Ghislain et al. 2009). Some of these SSRs were any) were trimmed using VecScreen (www.ncbi.nlm.nih. transferred to other Solanaceae species such as naranjilla gov/VecScreen/VecScreen.html). ESTs were then assembled (Solanum quitoense Lam.) and bush tomato (Solanum cen- to eliminate redundancy using CD-HIT Suite with a 90 % trale J. M. Black) for genetic diversity analysis (Torres et al. sequence similarity threshold (Huang et al. 2010), while the 2008; Waycott et al. 2011). However, only a few cross-species nucleotide sequences were compared and integrated using SSR markers were available for the molecular characterization BlastN (blast.ncbi.nlm.nih.gov). The generated non- of silverleaf nightshade ecotypes (Zhu et al. 2012). In the redundant sequences were further used to detect the SSR public databases, 181 nucleotide and expressed sequence tag motifs using SSRIT software (Temnykh et al. 2001) with the (EST) sequences of silverleaf nightshade are available from criteria of at least four and three repeat units for di-, tri- and National Center for Biotechnology Information (NCBI) Gen- higher order nucleotides, respectively. bank (www.ncbi.nlm.nih.gov), which can be exploited to develop molecular markers for genetic analysis. Primer Design In the present study, we developed a suite of SSR markers from publicly available ESTand nucleotide sequences in order Primer pairs were designed based on the flanking sequences to determine the extent of natural population diversity in a of the detected SSR motifs using Primer Premier 5.0 subset of silverleaf nightshade accessions. These markers will (www.premierbiosoft.com) with a length of 18–30 bp, am- provide a valuable tool to understand the genetic diversity and plification product size of 100–350 bp and melting temper- structure of silverleaf nightshade and to assist in the identifi- ature ranged from 55 to 60°C (Table 2). The 5′ end of the cation of silverleaf nightshade and quena ecotypes. forward primer of each SSR primer pair was tailed with M13 sequence (Raman et al. 2005), which allows an inex- pensive way to perform high throughput fragment analysis Materials and Methods (Rampling et al. 2001; Schuelke 2000).

Plant Material and DNA Extraction PCR Amplification and SSR Analysis

Thirty-nine samples of silverleaf nightshade were collected from The PCR protocol was modified from Raman et al. (2005). nine locations across southeastern Australia (Table 1). In addi- Amplification was carried out in 12 μLofreactionmixture tion, two samples of quena were also collected to investigate the consisting of 50–100 ng of template DNA, 1.2 μLof10× transferability of primer pairs between these two morphologi- buffer (containing Tris·Cl, KCl, (NH4)2SO4 and15mM cally similar Solanum species. Genomic DNA from each sample MgCl2), 6 mM MgCl2, 240 mM of each dNTP, 0.4 unit of was extracted from the frozen leaf material using the standard HotStar Taq (Qiagen, Australia), 0.15 μM of forward primer phenol/chloroform method (Sambrook et al. 1989). (Sigma-Aldrich, Australia), 0.3 μM of reverse primer and 0.3 μM of M13 primer (D4, D3 or D2; Beckman Coulter, Identification of Microsatellites USA). After an initial denaturation of 4 min at 94°C, 30 cycles for 30 s at 94°C, 30 s at 55–60°C (depending on the primers, A total of 169 ESTs and 12 nucleotide sequences derived Table 2) and 30 s at 72°C were performed, followed by a final from S. elaeagnifolium were sourced from the NCBI extension of 10 min at 72°C. Then, 0.6, 0.7 and 0.8 μLofPCR

Table 1 Silverleaf nightshade and quena samples used in Species Location Abbreviation Samples GPS (latitude/longitude) this study Silverleaf nightshade Narrandera Nar 5 −34°46′/146°25′ Temora Tem 4 −34°24′/147°36′ Ungarie Ung 5 −33°35′/146°55′ Loxton Lox 4 −34°38′/140°41′ Wirrabara Wir 4 −33°02′/138°16′ Keith Kei 4 −36°06′/140°16′ Hopetoun Hop 5 −35°36′/142°26′ Serpentine Ser 4 −36°24′/143°58′ Jarklin Jar 4 −36°14′/143°56′ Quena Wagga Wagga Q-W 1 −35°07′/147°20′ Jarklin Q-J 1 −36°14′/143°56′

53 250 Plant Mol Biol Rep (2013) 31:248–254

Table 2 Characteristics of 26 SSR primers developed in S. elaeagnifolium, including locus name, forward (F) and reverse (R) primer sequence, repeat motif, annealing temperature (Ta), allele size range including 19 bp of M13-tail and GenBank accession number

Locus Primer sequence (5′–3′) Repeat motif Ta (°C) Size range (bp) GenBank

SLNZ1 F: ACTAATACCTTACCCCGTTCATCT (TTC)4 55 308 EU983576 R: ATTCGTTCAAGAAGGGCTCC

SLNZ2 F: ATAGTACACTCAGCATCCATCATAAG (AT)4…(TA)4 55 221–232 GO496323 R: ACAGGAGGAACAGCAAGGC

SLNZ3 F: TCACACCACTAAAGGGGGGAT (TTA)3 50–60 – GO496323 R: ATCAACAGGAGGAACAGCAAGG

SLNZ4 F: ATGTAGGGACTAGTGCTCGAGTT (TCC)3 55 328–331 GO496325 R: AATAAAGCAAGGGCAATAGGTC

SLNZ5 F: TATGGGGCACATGGGAGAG (CTTCT)3 58 196–204 GO496328 R: AACCCCCATTCTAAATCCTTGT

SLNZ6 F: CTTTGTTCGGAGTTGTTGACC (GA)5 58 256–278 GO496340 R: CCTCCATCGCAAAACCATC

SLNZ7 F: AGAGTGGAGAGGAGAAGTAGAAGG (AAG)3 58 226–259 GO496340 R: GGTAAATTGAGGATCTTGGGTG

SLNZ8 F: GGAATTAAGGGTCCAAGGC (ATG)3…(TTA)3 58 195–202 GO496341 R: CTCACAAGTTACTCGGGCTCT

SLNZ9 F: TTCATAAATGAGAACTTACACGGAC (GTG)3 58 226–268 GO496344 R: TCTTAGCAGCGAACTGGGAC

SLNZ10 F: CCAAGCGAGGAAATAGCACT (ATG)3 58 213 GO496346 R: GTGCTTCCGATTTCTCCAAC

SLNZ11 F: GGTGTTTGTTGGAGAAATCGG (CAA)3 60 231 GO496350 R: TCTTCTACGATTTCCTTGGTGC

SLNZ12 F: GAAATGAAAGTCCCATCTCC (TTTTAT)4 55 328 GO496350 R: TGACTTCAGAACCAGTTACTCCT

SLNZ13 F: CAATCACAGTAGAAAGGGTCGCT (TG)4 50–60 – GO496355 R: TTACCATTCCCTATGTTGATCCAG

SLNZ 14 F: GCGAACGAATAATTGACCACC (TG)4 60 299 GO496355 R: AGTCGCCAAACTCCACATCTC

SLNZ15 F: TCATCACGCAAACGCTTACTC (AAG)4 55 174–186 GO496359 R: ATTTAACTATGTGCTAATTGTTATCGC

SLNZ16 F: CAAAGATACGGACCGCACCT (AT)4 50–60 – GO496359 R: GGTAAACGCCAGACGAACAAG

SLNZ17 F: CCAAGGCTCGGAAGAACC (AG)4 58 162–174 GO496370 R: CCACGAAAACACAACCTAACTAAC

SLNZ18 F: GGCTAAGTGACTAAACAAAAATGG (CA)5 55 185 GO496384 R: AGCAGTGGTATCAATTTGTGTCG

SLNZ19 F: TGGTAGAGGCGAAGGCAT (AG)4 58 216 GO496385 R: GCATCTTCAGGTCCCAACTT

SLNZ20 F: CACTTGCCCCTATTCCTGTCAT (CA)4 58 218–242 GO496403 R: CTTGTATCCTTCTCGCTACCTTTC

SLNZ21 F: GCTGCTACTCCCAATCCTAACTG (TA)4 58 245–289 GO496403 R: AAATCTCCGACGAAAGCTACTACT

SLNZ22 F: GCAGAATCCCGTGAACCATC (CG)5 55 202–257 AY996508 R: CGCCGAGAGAGTTGGGTTAC

SLNZ23 F: ATTGGTTGGGCTGTGTTCCT (TTA)3 55 294 AF224067 R: TGGGCGGATTTAGCAACTG

SLNZ24 F: TTTAGCCTATTCCACAATGTCTCA (ATT)3 58 353 AF224067 R: TGGCGAATACAACCAACTATCAT

SLNZ25 F: TCACTATCTCTATGGGGTAAAAACG (AAT)3 58 224 AF224067 R: GCATAGTATTGTCCGATTCATAAGG

SLNZ26 F: GGCATTGGAAATACTTTTTATTAC (TC)4 55 123–160 DQ180399 R: CCTAAAAGCGGAGGAATGTC

– no amplification)

54 Plant Mol Biol Rep (2013) 31:248–254 251 products generated by D4, D3 and D2 (Beckman Coulter, remaining five SSR motifs (data not shown) were inappro- USA), respectively, were mixed with 0.4 μL of DNA size priate for primer design because of insufficient flanking standard kit 400 (Beckman Coulter, USA) and 28.6 μLof sequence of the SSR loci. All 26 primer pairs were further loading solution (Beckman Coulter, USA), and separated on a used to amplify genomic DNA of 39 silverleaf nightshade CEQ 8000 genetic analysis system (Beckman Coulter, Inc.) as samples. Of these primer pairs examined, 3 (SLNZ 3, SLNZ described previously (Raman et al. 2005). 13 and SLNZ 16) did not produce any amplification products, while the other 23 primer pairs produced repeatable Data Analysis and reliable alleles (Table 3). Twenty-one of them produced one or two discrete fragments in each individual and were Only single-locus markers were further used for calculating considered as single-locus markers. Among these 21 primer observed and expected heterozygosity and genetic similari- pairs, 10 were monomorphic, while the other 11 SSRs ty. The observed and expected heterozygosity were calcu- (comprising ten EST-SSRs and one genomic SSR) were lated on the polymorphic primers using POPGENE software polymorphic. By contrast, two primer pairs SLNZ 7 and version 1.32 (Yeh and Boyle 1997). A similarity matrix was SLNZ 22 amplified multiple bands (three or more). Ampli- calculated based on Jaccard’s coefficient using Similarity for fication of multiple alleles might be caused by the duplica- Qualitative data in NTSYS-pc 2.1 (Rohlf 2000). An Un- tion of genomic regions (Senthilvel et al. 2008), which is weighted Pair Group Method with Arithmetic Mean common in plants (Yu et al. 2004; Sharma et al. 2009). (UPGMA) dendrogram was calculated using the same soft- The EST- and genomic SSRs reported here is the first set ware to illustrate genetic relationships between the samples. of specific SSR markers for silverleaf nightshade. The number of alleles detected by single-locus markers ranged from

Results and Discussion Table 3 Results of initial primer screening in S. elaeagnifolium, including the source of SSR for each primer pair, number of alleles SSR Motif Frequency and Distribution (NA), observed (HO) and expected (HE) heterozygosity and cross- species amplification in S. esuriale A total of 63 non-redundant sequences were identified Locus Source NA HO HE Cross-transferability through CD-HIT and BlastN analysis, including 56 ESTs and seven nucleotide sequences. Seventeen of the 56 EST SLNZ1 Nucleotide 1 0.00 0.00 + sequences contained 25 SSR motifs, with about 30 % of the SLNZ2 EST 2 0.00 0.10 + non-redundant ESTs containing at least 1 SSR motif. One in SLNZ4 EST 2 0.00 0.23 − 3.3 non-redundant ESTs contained at least one SSR. The SLNZ5 EST 3 0.31 0.30 − number of di- and tri- EST-SSR motif was 12 and 11, SLNZ6 EST 3 0.10 0.50 − respectively. The remaining two motifs consisted of one SLNZ7a EST 7 ––− penta- and one hexa-nucleotide motifs. In addition, four of SLNZ8 EST 2 0.08 0.07 + the seven nucleotide sequences contained six SSR motifs SLNZ9 EST 2 0.00 0.10 + (data not shown). SLNZ10 EST 1 0.00 0.00 + The number of dinucleotide (48 %) and trinucleotide SLNZ11 EST 1 0.00 0.00 − (44 %) repeats in ESTs was similar in this study, and there SLNZ12 EST 1 0.00 0.00 − were no obviously abundant motifs found in di- and tri- SLNZ14 EST 1 0.00 0.00 − repeats. However, compared to other EST-SSR studies in SLNZ15 EST 3 0.97 0.64 + other Solanaceae species, trinucleotides were the most fre- SLNZ17 EST 3 0.05 0.37 + quent motifs; with AT and GA, and AAG and AAT were the SLNZ18 EST 1 0.00 0.00 − most common di- and trinucleotide motifs, respectively SLNZ19 EST 1 0.00 0.00 − (Stagel et al. 2008; Nunome et al. 2009; Feingold et al. SLNZ20 EST 3 0.13 0.25 + 2005). The density and frequency of SSRs estimated in this SLNZ21 EST 2 0.00 0.10 + study might be affected by the limited ESTs available in SLNZ22a Nucleotide 10 ––+ silverleaf nightshade expressed genome. SLNZ23 Nucleotide 1 0.00 0.00 + SLNZ24 Nucleotide 1 0.00 0.00 + SSR Amplification and Polymorphism SLNZ25 Nucleotide 1 0.00 0.00 + SLNZ26 Nucleotide 4 0.03 0.17 + Out of the 25 EST-SSR motifs and 6 nucleotide SSR motifs, a total of 26 primer pairs were designed that included 20 + amplified, − not amplified) EST-SSR and 6 genomic SSR primer pairs (Table 2). The a SLNZ marker with multiple band amplifications

55 252 Plant Mol Biol Rep (2013) 31:248–254 two to four with an average of 2.6 per locus. The observed achieved 27 % transferable rate when they transferred SSR heterozygosity ranged from 0 to 0.97, and the expected markers from potato to naranjilla. When SSR markers from heterozygosity ranged from 0.07 to 0.64, with average of tomato and eggplant were transferred to bush tomato, a 0.15 and 0.26, respectively (Table 3). The level of polymor- transferable rate of 60 % was detected (Waycott et al. phism detected in this study is much lower than in our 2011). Previously, we reported an overall 37 % (13/35) previous cross-species SSR study (average expected hetero- cross-transferability of SSR markers from potato, tomato zygosity at 0.53) in silverleaf nightshade (Zhu et al. 2012). and eggplant to silverleaf nightshade (Zhu et al. 2012). This is probably due to the conserved nature of EST-SSR Solanum is one of the largest genera (approximately 1,600 markers. Most of the SSRs developed here are EST-SSR, species) of flowering plants. Compared to others, Solana- which is usually less polymorphic than genomic SSRs (Cho ceae species might have experienced a longer evolutionary et al. 2000). However, EST-SSR markers derived from process or higher speciation rate (Whalen and Caruso 1983). cDNA provide a valuable resource for identification and These factors may lead to great genetic divergence among developing gene-associated SSR markers. Solanum species and therefore resulted in low transferability (Whalen and Caruso 1983; Torres et al. 2008). Cross-Species Transferability of SSRs Genetic Diversity Among Silverleaf Nightshade The 23 markers that produced bands were further tested for cross-transferability in quena. Fourteen of them (61 %) were According to the UPGMA dendrogram, quena and silverleaf successfully transferred, including eight EST-SSRs and six nightshade were clearly separated at similarity level of 0.13 genomic SSRs (Table 3), with four of them (SLNZ 8, SLNZ based on Jaccard’s coefficient (Fig. 1), which indicated the 10, SLNZ 15 and SLNZ 22) polymorphic. This within great genetic divergence between the two species. Genetic subgenus SSR transferable ratio is lower than cross-species similarity among silverleaf nightshade individuals ranged SSR investigation in other species. Rossetto (2001) from 0.4 to 1.0, with an average genetic similarity of 0.79 reviewed that SSR primer pairs showed an average 89.8 % (data not shown). Two main subgroups were observed success rate when applied within subgenera (such as sub- among silverleaf nightshade individuals: one contained genera within Magnolia and Vitis). However, the transfer- two silverleaf nightshade individuals (Lox 2 and Lox 3) ability of SSR primer pairs within Solanaceae species is from Loxton, South Australia and the other included all usually much lower. For instance, Torres et al. (2008) the remaining 37 silverleaf nightshade samples. In this

Nar1 Nar5 Tem2 Jar4 Wir3 Nar2 Kei4 Ser4 Hop4 Nar3 Tem3 Ung3 Ung4 Hop5 Ung5 Kei3 Lox4 Ser1 Wir1 Jar1 Wir2 Wir4 Ung2 Hop3 Tem4 Ser2 Kei2 Ser3 Nar4 Tem1 Lox1 Ung1 Hop2 Jar2 Kei1 Hop1 Jar3 Lox2 Lox3 Q-W Q-J 0.13 0.24 0.35 0.46 0.57 0.67 0.78 0.89 1.00 Jaccard's Coefficient

Fig. 1 UPGMA dendrogram calculated by Jaccard’s coefficient showing three clusters. Sample details are included in Table 1

56 Plant Mol Biol Rep (2013) 31:248–254 253 study, we found less diversity among silverleaf nightshade Acknowledgments The authors gratefully acknowledge the financial accessions as compared to a previous genetic diversity study support of Charles Sturt University and NSW Department of Primary Industries (WWAI) for providing facilities to conduct this research. that was based on cross-species SSR markers (with an average genetic similarity of 0.37; Zhu et al. 2012). This is probably due to the conserved nature of EST-SSR markers References (Cho et al. 2000). Genetic diversity of silverleaf nightshade may be attrib- Bean AR (2004) The taxonomy and ecology of Solanum subg. Lep- uted to its propagation systems and multiple introductions. tostemonum (Dunal) Bitter (Solanaceae) in Queensland and far Silverleaf nightshade propagates both sexually (self-incom- north-eastern New South Wales. Austrobaileya 6:639–816 patibility) and asexually (through root fragments). Cross- Cho YG, Ishii T, Temnykh S, Chen X, Lipovich L, McCouch SR, Park pollinating species usually have a higher level of genetic WD, Ayres N, Cartinhour S (2000) Diversity of microsatellites derived from genomic libraries and GenBank sequences in rice diversity than those reproduced clonally or by self- (Oryza sativa L.). Theor Appl Genet 100:713–722. doi:10.1007/ pollination (Ward and Jasieniuk 2009). Silverleaf night- s001220051343 shade samples used in this study were collected from geo- Cuthbertson EG, Leys AR, McMaster G (1976) Silverleaf nightshade—a graphically diverse area in NSW, SA and VIC and therefore potential threat to agriculture. Agric Gaz of New South Wales 87:11–13 may represent genetically diverse genotypes. In addition, Dekker J (1997) Weed diversity and weed management. Weed Sci the possibility of multiple introductions in these states has 45:357–363 been highlighted by many researchers (Cuthbertson et al. Ellis JR, Burke JM (2007) EST-SSRs as a resource for population genetic – 1976; Stanton et al. 2009), which will also contribute to analyses. Heredity 99:125 132. doi:10.1038/sj.hdy.6801001 Feingold S, Lloyd J, Norero N, Bonierbale M, Lorenzen J (2005) genetic diversity of silverleaf nightshade in Australia. Mapping and characterization of new EST-derived microsatellites In this study, five species-specific markers were identified for potato (Solanum tuberosum L.). Theor Appl Genet 111:456– in silverleaf nightshade (primer pair SLNZ 11, SLNZ 12, 466. doi:10.1007/s00122-005-2028-2 SLNZ 14, SLNZ 18 and SLNZ 19 amplifying 231-, 328-, Ghislain M, Nunez J, Herrera MR, Pignataro J, Guzman F, Bonierbale M, Spooner DM (2009) Robust and highly informative 299-, 185- and 216-bp fragments, respectively) and four in microsatellite-based genetic identity kit for potato. Mol Breed quena (primer pair SLNZ 2, SLNZ 8, SLNZ 20 and SLNZ 21 23:377–388. doi:10.1007/s11032-008-9240-0 amplifying 223-, 161-, 220- and 285-bp, respectively). Some Huang Y, Niu B, Gao Y, Fu L, Li W (2010) CD-HIT Suite: a web diagnostic micromorphological distinguishing features were server for clustering and comparing biological sequences. Bioin- forma 26:680–682 reported previously, such as the trichomes of silverleaf night- Johnson A, Brooke G, Thompson R, Roberts K, Hertel K, Border N, shade having an intrusive base (Bean 2004; Zhu et al. 2011). McNee T, Sullivan P (2006) Weed control for cropping and pasture However, currently only three unique alleles (85-, 222- and in central west NSW. http://www.dpi.nsw.gov.au/agriculture/ 249-bp corresponding to EM 117, EM 135 and ESM 3 pests-weeds/weeds/publications/central-west.Accessed28May 2011 primer pairs, respectively) are available to distinguish these Kwon YS, Park SG, Yi SI (2009) Assessment of genetic variation two species (Zhu et al. 2011). In conjunction with these three among commercial tomato (Solanum lycopersicum L.) varieties unique alleles, nine new species-specific SSR markers will using SSR markers and morphological characteristics. Genes – enable to distinguish both quena and silverleaf nightshade. Genom 31:1 10 Nissen SJ, Masters RA, Lee DJ, Rowe ML (1995) DNA-based marker Reliable identification of invasive species is required for systems to determine genetic diversity of weedy species and their selection of biocontrol agents (Nissen et al. 1995). In addi- application to biocontrol. Weed Sci 43:504–513 tion, correct identification can also assist the selection of Nunome T, Negoro S, Kono I, Kanamori H, Miyatake K, Yamaguchi herbicide and management strategies as quena is not a seri- H, Ohyama A, Fukuoka H (2009) Development of SSR markers derived from SSR-enriched genomic library of eggplant (Solanum ous problem in Australian crop and pasture system and is melongena L.). Theor Appl Genet 119:1143–1153. doi:10.1007/ easier to control than silverleaf nightshade (Johnson et al. s00122-009-1116-0 2006). O’Hanlon PC, Peakall R, Briese DT (2000) A review of new PCR- In conclusion, we have developed the first set of species- based genetic markers and their utility to weed ecology. Weed Res 40:239–254 specific SSR primers for silverleaf nightshade. These markers Raman R, Raman H, Johnstone K, Lisle C, Smith A, Martin P, Allen H could be useful in determining natural population diversity (2005) Genetic and in silico comparative mapping of the poly- and structure and distinguishing silverleaf nightshade from phenol oxidase gene in bread wheat (Triticum aestivum L.). Func – quena. EST-SSR markers are often preferable for comparative Integr Genomics 5:185 200. doi:10.1007/s10142-005-0144-3 Rampling LR, Harker N, Shariflou MR, Morell MK (2001) Detection genomic analysis to other markers such as RFLP, AFLP and and analysis systems for microsatellite markers in wheat. Aust J RAPD as they show higher similarity across different species Agric Res 52:1131–1141. doi:10.1071/ar01027 than intergenic chromosomal regions. Therefore, the marker Rohlf FJ (2000) NTSYS-pc: numerical taxonomy and multivariate set that we have developed is anticipated to be directly appli- analysis system, version 2.1. Exeter Publications, New York Rossetto M (2001) Sourcing of SSR markers from related plant spe- cable for comparative genomics and evolution studies on cies. In: Henry RJ (ed) Plant genotyping: the DNA fingerprinting silverleaf nightshade and related species. of plants, vol. 1. CAB International, Wallingford, pp 211–224

57 254 Plant Mol Biol Rep (2013) 31:248–254

Sambrook J, Fritsch EF, Maniatis T (eds) (1989) Molecular cloning: a Torres AF, Arias AS, Arahana V, Torres ML (2008) Preliminary laboratory manual, 2nd edn. Cold Spring Harbor Laboratory, New assessment of genetic diversity and phenetic relations for section York Lasiocarpa by means of heterologous SSR markers. Crop Sci Schuelke M (2000) An economic method for the fluorescent labeling 48:2289–2297. doi:10.2135/cropsci2007.11.0623 of PCR fragments. Nat Biotechnol 18:233–234 Ward SM, Jasieniuk M (2009) Review: sampling weedy and invasive Senthilvel S, Jayashree B, Mahalakshmi V, Kumar PS, Nakka S, plant populations for genetic diversity analysis. Weed Sci 57:593– Nepolean T, Hash CT (2008) Development and mapping of 602. doi:10.1614/ws-09-082.1 simple sequence repeat markers for pearl millet from data mining Waycott M, Jones BL, Van Dijk JK, Robson HLA, Calladine A (2011) of expressed sequence tags. BMC Plant Biol 8:119. doi:10.1186/ Microsatellite markers in the Australian desert plant, Solanum 1471-2229-8-119 centrale (Solanaceae). Am J Bot 98:E81–E83. doi:10.3732/ Sharma R, Bhardwaj P, Negi R, Mohapatra T, Ahuja P (2009) Identi- ajb.1000356 fication, characterization and utilization of unigene derived micro- Whalen MD, Caruso EE (1983) Phylogeny in Solanum sect. Lasio- satellite markers in tea (Camellia sinensis L.). BMC Plant Biol carpa (Solanaceae): congruence of morphological and molecular 9:53. doi:10.1186/1471-2229-9-53 data. Syst Bot 8:369–380. doi:10.2307/2418356 Stagel A, Portis E, Toppino L, Rotino GL, Lanteri S (2008) Gene- Yeh FC, Boyle TJB (1997) Population genetic analysis of co-dominant based microsatellite development for mapping and phylogeny and dominant markers and quantitative traits. Belgian J Bot studies in eggplant. BMC Genomics 9:357–370. doi:10.1186/ 129:157 1471-2164-9-357 Yu JK, La Rota M, Kantety RV, Sorrells ME (2004) EST derived SSR Stanton RA, Heap JW, Carter RJ, Wu H (2009) Solanum elaeagnifo- markers for comparative mapping in wheat and rice. Mol Genet lium. In: Panetta FD (ed) The biology of Australian weeds, vol 3. Genomics 271:742–751. doi:10.1007/s00438-004-1027-3 R. G. and F. J. Richardson, Melbourne, pp 274–293 Zhou YQ (2005) Technology of DNA molecular markers utilized on Swapna M, Sivaraju K, Sharma RK, Singh NK, Mohapatra T (2011) plants researches. Biological laboratory series. Chemical Industry Single-strand conformational polymorphism of EST-SSRs: a po- Press, Beijing tential tool for diversity analysis and varietal identification in Zhu XC, Burrows G, Wu H, Raman H, Stanton R, Lemerle D (2011) sugarcane. Plant Mol Biol Rep 29:505–513. doi:10.1007/ Identification of silverleaf nightshade using microsatellite markers s11105-010-0254-5 and microstructure. Paper presented at the 23rd Asian-Pacific Temnykh S, DeClerck G, Lukashova A, Lipovich L, Cartinhour S, Weed Science Society Conference, Cairns, Australia McCouch S (2001) Computational and experimental analysis of Zhu XC, Wu HW, Raman H, Lemerle D, Stanton R, Burrows GE microsatellites in rice (Oryza sativa L.): frequency, length varia- (2012) Evaluation of simple sequence repeat (SSR) markers from tion, transposon associations, and genetic marker potential. Ge- Solanum crop species for Solanum elaeagnifolium.WeedRes nome Res 11:1441–1452 52:217–223. doi:10.1111/j.1365-3180.2012.00908.x

58 CHAPTER 6 ACCEPTED PAPER

A total of 36 SSR markers are described in Chapters 4 and 5. In

Chapter 6, 19 highly polymorphic SSR primer pairs were selected to study the genetic diversity of SLN in Australia. A subset of individuals was further evaluated using AFLP markers in Chapter 7.

Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., Stanton, R., & Burrows, G.

E. (2013). SSR marker analysis to determine the genetic variation in

Solanum elaeagnifolium in Australia. Plant Protection Quarterly. (accepted).

59 GENETIC VARIATION IN SOLANUM ELAEAGNIFOLIUM IN AUSTRALIA USING SSR MARKER Xiaocheng Zhu1, 2, Hanwen Wu1, 3, Harsh Raman1, 3, Deirdre Lemerle1, 2, Rex Stanton1, 2 and Geoffrey E. Burrows1, 2 1 EH Graham Centre for Agricultural Innovation (an alliance between NSW Department of Primary Industries and Charles Sturt University), NSW. Email: [email protected] 2School of Agricultural and Wine Sciences, Charles Sturt University, Wagga Wagga, NSW. 3 Wagga Wagga Agricultural Institute, PMB, Wagga Wagga, NSW, Australia

SUMMARY Silverleaf nightshade (Solanum elaeagnifolium Cav.) is a problematic summer-growing perennial weed in Australia. The genetic diversity of silverleaf nightshade is poorly understood. Nine silverleaf nightshade specific and 10 cross-species simple sequence repeat (SSR) primer pairs were utilized to investigate the genetic variations among 94 silverleaf nightshade populations collected in Australia. High genetic diversity was found within silverleaf nightshade populations, with an average genetic similarity at 0.43. The Unweighted Pair Group Method with Arithmetic mean based dendrogram indicated the presence of genetically diverse silverleaf nightshade populations in Australia. However, no well support genetic structure was found. The Mantel test indicated that there is no significant correlation between genetic variation and geographic distance. These result suggested lack of geographic structure of genetic diversity, which is probably due to the long distant spread of seeds of silverleaf nightshade. The high genetic diversity of silverleaf nightshade could contribute to the inconsistency in control efficacy between populations. Keywords: silverleaf nightshade; invasive weed; cross-species SSR; microsatellites; genetic diversity

INTRODUCTION Silverleaf nightshade (Solanum elaeagnifolium Cav.) is a summer-growing perennial weed and is a Weed of National Significance in Australia (Australian Weeds Committee 2012). It is widely distributed in the cereal cropping zone in southern Australia, with potential to infest 398 million hectares (Kwong 2006). It has been reported that silverleaf nightshade can cause up to 77% yield loss in cereals (Heap et al. 1997).

To date, classical biological control has not been implemented in Australia despite being declared a target for biological control in 1985 (Kwong 2006). In addition, the high regenerative ability of the root system limited the efficacy of mechanical managements (Stanton et al. 2011), thus chemical control is the only useful option for silverleaf nightshade management in Australia. However, herbicide efficacy can be influenced by many factors including plant genetic variation (Marshall and Moss 2008). Therefore, effective management of silverleaf nightshade requires comprehensive assessment of genetic diversity (Dekker 1997, Holt and Hochberg 1997).

Genetic diversity studies have been conducted in many weed species, such as bitter vine (Mikania micrantha) (Wang et al. 2012) and false helleborine (Veratrum album) (Treier and Muller-Scharer 2011). These studies contribute to the understanding of weed genetic diversity, evolution and invasion. Silverleaf nightshade was found to be genetically diverse in South Australia by random amplification of polymorphic DNA (RAPD) (Hawker et al. 2006). However, genetic diversity of silverleaf nightshade populations

60 growing across Australia is largely unknown. The genetic diversity studies using genetic markers will contribute to the management of silverleaf nightshade (Dekker 1997).

Simple sequence repeat (SSR) markers are usually co-dominant, and more informative than dominant markers such as RAPDs because they are capable of differentiating homozygous individuals from heterozygous plants (McGregor et al. 2000, Peakall 1997). Furthermore SSRs are more reproducible, easily scored and analysed on high throughput genotyping platforms. Thirty six SSR markers have recently been developed for silverleaf nightshade (Zhu et al. 2012, 2013).

In this study, a subset of 19 high polymorphic primer-pairs was applied to study genetic diversity of 94 populations of silverleaf nightshade collected across Australia.

MATERIALS AND METHODS Plant material A total of 670 silverleaf nightshade individuals were collected from 94 locations (populations) in New South Wales (NSW), South Australia (SA), Victoria (VIC), Queensland (QLD) (Figure 1) and Katanning, Western Australia (WA). One to ten individuals were collected from each location, depending on the level of infestation. Sampled individuals were at least 50 m apart to reduce the probability of collecting clonal plants. In addition, five field samples of quena (S. esuriale a native Solanum species), and five commercial samples of eggplant (S. melongena L.; Hortico, Australia), were included for comparison. About 1 g of fresh, undamaged leaf material was collected from each individual, placed in a 1.5 mL eppendorf tube, and then stored at -80ºC in the laboratory until DNA isolation.

DNA isolation Genomic DNA was isolated individually and the quality and concentration was checked as described previously (Zhu et al. 2013). The individual DNA concentration was then adjusted to 20 ng/µL. Equal amounts of DNA from individuals representing the same population were bulked for PCR amplification, as a cost efficient method of population analysis (Arunyawat et al. 2007, Eschholz et al. 2008).

PCR reaction and SSR analysis Nineteen SSR primer-pairs (Table 1) were selected to investigate genetic diversity between populations, on the basis of their high expected heterozygosity value (HE). The details of these primer-pairs have been described previously (Zhu et al. 2012, 2013). The 5’ end of the forward primer of each SSR primer-pair was tailed with a M13 sequence to perform high throughput fragment analysis (Raman et al. 2005). PCR amplification and detection of the amplification products were carried out as described (Zhu et al. 2013).

Data analysis Binary data, as the presence or absence (1 or 0) of bands of each locus for each population, was scored to construct a similarity matrix by Jaccard’s coefficient (Jaccard 1908) using NTSYS-pc 2.1 (Rohlf 2000). The Unweighted Pair Group Method with Arithmetic mean (UPGMA) was calculated using the SAHN procedures of the same software to construct a dendrogram of population genetic relationships. Non-parametric bootstrapping (N=1,000 replicates) was used to estimate statistical support at detected clades, using Paleontological Statistics Software Package (PAST) (Hammer et al. 2001). Correlation between genetic

61 and geographical distance among all pair-wise population comparisons was tested by Mantel test using NTSYS, with 1, 000 random permutations.

Figure 1 Sampling locations of silverleaf nightshade in New South Wales, South Australia, Victoria and Queensland (Western Australian population 94 not shown).

RESULTS AND DISCUSSIONS The genetic diversity in 94 populations was assessed according to allele frequency. The SSR analysis illustrated a high level of genetic variation between silverleaf nightshade populations, with a total of 182 polymorphic bands (alleles) detected. The number of polymorphic bands varied from two (primer pairs SLNZ 8, SLNZ 17 and SLNZ 20) to 32 (primer pair CA 158), with an average of 9.6 polymorphic bands per locus (Table 1). Mean Jaccard’s genetic similarity between populations was 0.43, varying from 0.21 to 0.76.

Bulk DNA analysis was used in this study. The reliability of bulk DNA analysis has been checked using a subset of individuals from nine locations (Zhu et al. 2013), which achieved similar results (average Jaccard similarity: 0.73 and 0.79 for bulk and individual analysis, respectively). This method can lead to the loss of the co-dominant feature of SSR analysis and does not allow estimates of the heterozygosity within a population. However, individual genotype information was not essential for estimating between population genetic diversity (Dubreuil et al. 1999). The DNA bulking method is highly repeatable and reliable for population genetic studies, such as in maize (Zea mays) (Eschholz et al. 2008) and wild tomatoes (Solanum peruvianum and S. chilense) (Arunyawat et al. 2007).

The 19 SSR markers were successfully used to assess the genetic variation among 94 populations of silverleaf nightshade collected from different states of Australia. The present study detected a high level of genetic polymorphism among silverleaf nightshade populations within Australia, with a mean genetic similarly of 0.43. The high level of genetic variation in Australia might be attributable to multiple introductions (Cuthbertson

62 et al. 1976), the heterogeneous nature of the initial introduction(s) and/or the self- incompatibility in silverleaf nightshade. Obligate outcrossing species usually have a higher level of genetic diversity than clonally or self pollinated species (Ward and Jasieniuk 2009).

Table 1 Details of SSR primers used to investigate genetic diversity between Solanum species, including the number of bands detected for each primer-pair and the corresponding allele sizes, including 19 bp of M13-tailed sequence.1 Estimate allele Size (bp) Primer ID Band Number S. elaeagnifolium S. esuriale S. melongena SLNZ5 4 184 - 203 Fail2 188 SLNZ 6 4 256 - 279 Fail Fail SLNZ 7 7 226 - 255 Fail 248 SLNZ 8 2 196 - 202 160 - 196 160 - 197 SLNZ 15 4 174 - 187 174 - 187 183 SLNZ 17 2 162 - 164 174 164 SLNZ 20 2 236 - 238 218 - 220 241 - 249 SLNZ 22 14 174 - 246 242 242 - 248 SLNZ26 3 123 -147 160 166 CA158 32 217 - 264 277 - 239 249 - 255 ESM3 22 Null - 355 249 - 258 264 - 268 EM117 30 110 - 178 85 - 178 96 - 116 EM127 9 Null - 222 168 - 268 Fail EM135 16 Null - 267 222 - 288 283 EM140 6 Null - 231 216 - 235 230 - 233 EM155 7 Null - 334 144 - 171 126 - 295 SSR111 7 Null - 180 174 - 176 183 STI001 4 205 - 217 205 213 STG0010 7 177 - 271 178 178 1Detailed sequence information is described previously (Zhu et al. 2012, 2013); 2Fail: No amplification detected.

The UPGMA dendrogram based on the Jaccard’s coefficient clearly separated quena and eggplant from silverleaf nightshade populations and supported by high bootstrap value, suggesting the genetic variability among related species (Figure 2). The 94 populations of silverleaf nightshade were clustered into two main groups with low bootstrap support (<70%), which indicated no well supported structure was found in Australia (Figure 2). In addition, no significant correlation was found between genetic and geographical distance among populations (r = -0.03, t = -091 and p = 0.17). The UPGMA

Figure 2 UPGMA dendrogram of Jaccard’s coefficient from dominant scored alleles of 94 Silverleaf nightshade accessions, quena (95) and eggplant (96) from Australia. Only bootstrap >70% are indicated.

64 dendrogram and the Mantel test suggested that there is no geographical structure of genetic variation in Australian silverleaf nightshade populations. Similar results were found in other invasive species such as Flaveria bidentis (Ma et al. 2011) and Parthenium hysterophorus (Tang et al. 2009). Long distance distribution of silverleaf nightshade is aided by the spread of fruits (seeds). Transportation of the contaminated livestock or fodder contributes to the dispersal of silverleaf nightshade populations. Such long-distance dispersal may explain the lack of geographic structure of silverleaf nightshade in Australia.Seeds can generate new plants and hybridize with other genotypes, leading to gene flow and may have contributed to the genetic diversity.

Weeds with high genetic diversity are more likely to develop new phenotypes in response to natural selection pressures, which allow better adaption to the environment or management practices (Dekker 1997). Therefore, genetic diversity in silverleaf nightshade highlights the challenge of successful management of this weed in Australia. Similarly, this study also shows the important role of seed spread in silverleaf nightshade infestation. Attention should be paid to stopping seed set and minimizing the movement of agricultural products, livestock, and machinery.

ACKNOWLEDGEMENTS We would like to acknowledge the Charles Sturt University, Australia for funding this research. We thank a number of state and local organisations across Australia for their assistance in field sampling.

REFERENCES Arunyawat, U., Stephan, W., Stadler, T. (2007). Using multilocus sequence data to assess population structure, natural selection, and linkage disequilibrium in wild tomatoes. Molecular Biology and Evolution 24(10), 2310-22. Australian Weeds Committee (2012). Weeds of National Significance: silverleaf nightshade (Solanum elaeagnifolium) draft strategic plan. Australian Weeds Committee, Canberra. Cuthbertson, E.G., Leys, A.R., McMaster, G. (1976). Silverleaf nightshade - a potential threat to agriculture. Agricultural Gazette of New South Wales 87(6), 11-3. Dekker, J. (1997) Weed diversity and weed management. Weed Science 45(3),357-63. Dubreuil, P., Rebourg, C., Merlino, M., Charcosset, A. (1999). Evaluation of a DNA pooled-sampling strategy for estimating the RFLP diversity of maize populations. Plant Molecular Biology Reporter 17(2), 123-38. Eschholz, T.W., Peter, R., Stamp, P., Hund, A. (2008). Genetic diversity of Swiss maize (Zea mays L. ssp mays) assessed with individuals and bulks on agarose gels. Genetic Resources and Crop Evolution 55(7), 971-83. Hammer, Ø., Harper, D.A.T., Ryan, P.D. (2001). PAST: Paleontological Statistics Software Package for education and data analysis. Palaeontologia Electronica 4(1), 1-9. Hawker, V., Preston, C., Baker, J. (2006). Genetic variation within and among silverleaf nightshade (Solanum elaeagnifolium Cav.) populations in South Australia. In: C. Preston, Watts JH, Crossman ND (eds) 15th Australian Weeds Conferences, Adelaide p 176. Heap, J.W., Smith, E., Honan, I. (1997). Silverleaf nightshade: a technical handbook for animal and plant control boards in South Australia. Primary Industries South Australia Animal and Plant Control Commission, pp. 42, Adelaide. Holt, R.D., Hochberg, M.E. (1997). When is biological control evolutionarily stable (or is it)? Ecology 78(6), 1673-83. Jaccard, P. (1908). Nouvelles recherches sur la distribution florale. Bulletin de la Societe Vaudoise des Sciences Naturelles 44,223-70. Kwong, R. (2006). Feasibility of biological control of solanaceous weeds of temperate Australia. North Sydney. pp. 150. Ma, J.W., Geng, S.L., Wang, S.B., Zhang, G.L., Fu, W.D., Shu, B. (2011). Genetic diversity of the newly invasive weed Flaveria bidentis (Asteraceae) reveals consequences of its rapid range expansion in northern China. Weed Research 51(4), 363-72. Marshall, R., Moss, S.R. (2008). Characterisation and molecular basis of ALS inhibitor resistance in the grass weed Alopecurus myosuroides. Weed Research (5), 48439-47.

65 McGregor, C.E., Lambert, C.A., Greyling, M.M., Louw, J.H., Warnich, L. (2000). A comparative assessment of DNA fingerprinting techniques (RAPD, ISSR, AFLP and SSR) in tetraploid potato (Solanum tuberosum L.) germplasm. Euphytica 113(2), 135-44. McKenzie, D.N. (1975). Silverleaf nightshade - one method of dispersal. Australian Weeds Research Newsletter 22, 13-5. Peakall, R. (1997). PCR-based genetic markers and their applications to turfgrass breeding. International Turfgrass Society Research Journal 8, 243-59. Raman, R., Raman, H., Johnstone, K., Lisle, C., Smith, A., Martin, P., Allen, H. (2005). Genetic and in silico comparative mapping of the polyphenol oxidase gene in bread wheat (Triticum aestivum L.). Functional & Integrative Genomics 5(4), 185-200. Rohlf, F.J. (2000). NTSYS-pc: Numerical Taxonomy and Multivariate Analysis System. Version 2.1 Exeter Publications. New York, USA. Stanton, R., Wu, H., Lemerle, D. (2011). Root regenerative ability of silverleaf nightshade (Solanum elaeagnifolium Cav.) in the glasshouse. Plant Protection Quarterly, 26(2), 54-6. Tang, S.Q., Wei, F., Zeng, L.Y., Li, X.K., Tang, S.C., Zhong, Y., Geng, Y.P. (2009). Multiple introductions are responsible for the disjunct distributions of invasive Parthenium hysterophorus in China: evidence from nuclear and chloroplast DNA. Weed Research 49(4), 373-80. Treier, U.A., Muller-Scharer, H. (2011). Differential effects of historical migration, glaciations and human impact on the genetic structure and diversity of the mountain pasture weed Veratrum album L. Journal of Biogeography 38(9), 1776-91. Wang, T., Chen, G.P., Zan, Q.J., Wang, C.B., Su, Y.J. (2012). AFLP genome scan to detect genetic structure and candidate loci under selection for local adaptation of the invasive weed Mikania micrantha. Plos One 7(7), e41310. Ward, S.M., Jasieniuk, M. (2009). Review: sampling weedy and invasive plant populations for genetic diversity analysis. Weed Science 57(6), 593-602. Zhu, X., Raman, H., Wu, H., Lemerle, D., Burrows, G., Stanton, R. (2013). Development of SSR markers for genetic analysis of silverleaf nightshade (Solanum elaeagnifolium) and related species. Plant Molecular Biology Reporter 31(1), 248-54. Zhu, X.C., Wu, H.W., Raman, H., Lemerle, D., Stanton, R., Burrows, G.E. (2012). Evaluation of simple sequence repeat (SSR) markers from Solanum crop species for Solanum elaeagnifolium. Weed Research 52(3), 217-23.

CHAPTER 7 PUBLISHED PAPER

Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., Stanton, R., & Burrows, G.

E. (2013). Genetic variation and structure of Solanum elaeagnifolium in

Australia analysed by AFLP markers. Weed Research. doi:

10.1111/wre.12029.

67 DOI: 10.1111/wre.12029

Genetic variation and structure of Solanum elaeagnifolium in Australia analysed by ampliﬁed fragment length polymorphism markers

X C ZHU*†,HWWU*‡, H RAMAN*‡,DLEMERLE*†,RSTANTON*& G E BURROWS*† *EH Graham Centre for Agricultural Innovation (an alliance between NSW Department of Primary Industries and Charles Sturt University), Wagga Wagga, NSW, Australia, †School of Agricultural and Wine Sciences, Charles Sturt University, Wagga Wagga, NSW, Australia, and ‡Wagga Wagga Agricultural Institute, PMB, Wagga Wagga, NSW, Australia

Received 3 November 2012 Revised 16 January 2013 Subject Editor: Pietro Iannetta, James Hutton Institute, UK

suggested that Australian S. elaeagnifolium may have Summary originated from two distinct gene pools. These results Solanum elaeagnifolium is a weed of national signifi- were further supported by principal co-ordinates anal- cance in Australia. However, the genetic diversity of ysis. Large spatial groups of individuals assigning to S. elaeagnifolium is poorly understood. Four amplified these two gene pools were found in western Victoria fragment length polymorphism primer combinations and south-western New South Wales (NSW) and were utilised to investigate the genetic variation and northern NSW, which correlated well with the early structure of 187 S. elaeagnifolium individuals collected records of S. elaeagnifolium in both regions. The high from 94 locations in Australia. High genetic diversity genetic diversity found here could add difficulties to was found, with an average Jaccard’s genetic similarity effective control of S. elaeagnifolium across regions. at 0.26. Individuals were assigned to two genetic Keywords: silverleaf nightshade, invasive weed, AFLP, clusters or considered as admixed according to their genetic diversity, genetic structure. membership coefficient value (q) calculated by Bayes- ian model-based genetic structure analysis. This

ZHU XC, WU HW, RAMAN H, LEMERLE D, STANTON R&BURROWS GE (2013). Genetic variation and structure of Solanum elaeagnifolium in Australia analysed by ampliﬁed fragment length polymorphism markers. Weed Research.

were reported in Victoria (VIC) in 1909 and South Introduction Australia (SA) in 1914, suggesting the possibility of Solanum elaeagnifolium Cav. (silverleaf nightshade) multiple introductions (Cuthbertson et al., 1976). belongs to the Leptostemonum subgenus in the Sola- Isolated infestations also occur in Western Australia num genus (Levin et al., 2006) and is believed to be a (WA) and Queensland (QLD). As a weed of national native of Central America (Stanton et al., 2009). It signiﬁcance, S. elaeagnifolium infests at least 0.35 mil- was ﬁrst recorded in Australia in 1901 at Bingara, lion hectares in Australia and has the potential to New South Wales (NSW). Subsequent infestations infest 398 million hectares (Feuerherdt, 2009).

Correspondence: X Zhu, School of Agricultural and Wine Sciences, Charles Sturt University, Wagga Wagga, NSW 2678, Australia. Tel: (+61) 2 6933 2749; Fax: (+61) 2 6938 1861; E-mail: [email protected]

Solanum elaeagnifolium is a deep-rooted, summer- Materials and methods growing perennial that reproduces both sexually as a self-incompatible outcrosser and vegetatively from Plant material adventitious buds in the root system (Hardin et al., 1972; Petanidou et al., 2012). Sexual reproduction is A total of 187 individuals were collected from 94 loca- important for long-distance seed dispersal. By con- tions across NSW, SA, VIC, WA and QLD (Fig. 1 trast, root systems can generate multiple adventitious and Appendix 1). One to three individuals were shoots in a single season and contribute to rapid randomly selected at each location, depending on the population increases or new infestations through levels of infestation. Each individual was spaced at movement of viable root fragments (Stanton et al., least 50 m apart from each other to reduce the proba- 2011). bility of sampling identical clones. Leaf materials were Effective weed management strategies are limited collected and stored as previously described (Zhu for S. elaeagnifolium, especially for large and dense et al., 2012). infestations. Chemical control is often expensive and may have a residual effect that can damage sensitive DNA isolation crop or pasture species sown in subsequent years (Stanton et al., 2009). Biocontrol proved to be unre- Genomic DNA was isolated individually from leaf liable in Australia due to harsh climatic conditions materials using the standard phenol/chloroform and the strong plant regenerative ability of the root method with three extractions (Sambrook et al., 1989). system (Stanton et al., 2009). Comprehensive assess- DNA quality was determined by electrophoresis ment of the genetic diversity in S. elaeagnifolium is (MI-DEAR, SYS-MD 120, USA) in a 1.0% agarose required for effective management, particularly in gel at 200 V for 10 min. The concentration of DNA À developing appropriate control strategies (Dekker, was then adjusted to 20 ng lL 1 for further analysis. 1997). Molecular markers have been widely used to assess AFLP analysis genetic diversity in many weed species, such as wild oats (Avena fatua;Liet al., 2007) and weedy red rice DNA digestion, ligation and pre-selective amplification â (Oryza sativa; Shivrain et al., 2010). Dominant ran- were performed as described in the manual of AFLP dom amplified polymorphic DNA (RAPD) markers Analysis System I (Invitrogen, Australia). Four primer detected a high level of genetic diversity in S. elaeag- combinations (E-ACC/M-CAA, E-ACC/M-CTT, nifolium in SA (Hawker et al., 2006). In addition, E-AGC/M-CAA and E-AGC/M-CTT) which showed high level of genetic diversity was also detected in 10 high polymorphism in preliminary screening (data not locations in Australia using 36 simple sequence repeat shown) were chosen for selective amplification. PCR SSR markers (Zhu et al., 2012, 2013). However, amplification (12 lL) was carried out in a Gene Amp information is scarce on the genetic diversity of PCR System 2700 (Applied Biosystems, Singapore), S. elaeagnifolium populations growing across containing 0.4 U Taq (Promega, Australia), 1.2 lL Australia. 10 9 buffer, 6 mM MgCl2, 240 mM of each dNTP’s, Amplified fragment length polymorphism (AFLP) 0.3 lM WellRED D4-PA labelled forward primer markers are more reproducible and informative than (Sigma Aldrich, Australia), 0.3 lM reverse primer RAPD markers (Jones et al., 1997; McGregor et al., (Sigma Aldrich) and 2 lL of a 1/20 dilution of 2000). Compared with single locus specific SSR pre-amplified template. markers, AFLPs are multilocus markers generated PCR products were separated using a CEQ8000 by digestion and specific amplification of fragments (Beckman Coulter Inc.). Data collection and fragment representing the entire genome (Vos et al., 1995) and analysis were performed using CEQTM 8000, version reveal the greatest amount of genetic diversity in 8.0. Peak criteria were 5% slope threshold, 5% relative several crops including potato (Solanum tuberosum L.; peak height and 95% size estimation confidence. McGregor et al., 2000). AFLPs provide higher resolu- The 600 internal size standard (Beckman Coulter Inc.) tion at the individual level and are suitable for deter- was used to calibrate allele sizes. mining genetic structure (Van der Wurff et al., 2003). Fragments were scored according to Stodart et al. In this study, AFLP markers were used to investigate (2005). Fragments that ranged from 60 to 600 bp were genetic diversity and structure of 187 individuals of recorded and binned into two nucleotide differences. A S. elaeagnifolium and to provide a comprehensive anal- binary matrix was obtained using the CEQ 8000 soft- ysis of the distribution of genetic variation in this weed ware. Only fragments with more than 10% frequency in Australia. were used for further analysis to ensure that fragment

Fig. 1 Distribution of sampled locations across south-eastern Australia (individual from WA not shown). artefacts were not considered as polymorphism Results (Stodart et al., 2005). Four AFLP primer combinations amplified 532 polymorphic fragments among the 187 S. elaeagnifolium Genetic diversity and structure analysis individuals. High genetic diversity was detected, with The binary matrix was used to calculate the Jaccard’s the average Jaccard’s genetic similarity at 0.26, ranging genetic similarity matrix using PAleontological STatis- from 0.07 to 0.69. tics software package (PAST), version 2.02 (Hammer The genetic structure of S. elaeagnifolium was et al., 2001).The Bayesian model-based structure analy- inferred using STRUCTURE (Fig. 2). The LnProb sis was obtained by STRUCTURE, version 2.3 to infer (D) value showed an incremental increase, which is the genetic structure of S. elaeagnifolium in Australia, common in STRUCTURE analysis (Evanno et al., using the admixture model (Falush et al., 2007). In this 2005). However, the DK value clearly suggested K = 2 model, K clusters are assumed and characterised by (Fig. 3). At the threshold of q ≥ 0.8, 53 (28.3%) and allele frequencies. A burn-in period of 30 000 was 51 (27.3%), individuals were assigned to clusters while applied, followed by 300 000 steps of Markov Chain the other 83 (44.4%) individuals could not be assigned Monte Carlo simulations. K was set to vary from 1 to to any clusters and were considered admixed. 20 with five iterations. A DK value was calculated and The first two PCoA axes explained 15.4% of the used to identify the number of clusters that best total variation (Fig. 4). Individuals are widely spread explained the data (Evanno et al., 2005). Individuals along the axes, reflecting the high level of genetic were assigned to clusters according to their membership diversity. The PCoA achieved similar result with coefficient (q) value. The threshold q ≥ 0.8 was applied STRUCTURE analysis. Individuals that were assigned as described by Menchari et al. (2007). The outputs of to Cluster 1 or Cluster 2 in the STRUCTURE analysis the estimated K were plotted using Distruct 1.1 (Rosen- were clearly separated from each other and those that berg, 2004). In addition, principal co-ordinates analysis were admixed were placed in the middle of Cluster 1 (PCoA) was calculated based on the Jaccard’s genetic and Cluster 2 individuals (Fig. 4). similarity matrix using the same software to investigate Several spatial groups were defined (Fig. 5). Within the relationship between individuals. each spatial group, more than 70% of individuals

Fig. 2 Genetic structure of Solanum elaeagnifolium in Australia inferred by STRUCTURE according to AFLP data.

3000 locations or the spatial isolation from other locations with similar genotypes (such as the location 93 in 2500 QLD and 94 in WA). There was a large group of individuals located in western VIC and south-western 2000 NSW that was closely associated with Cluster 1, and a concentrated group in northern NSW and SA that was 1500

ΔK closely associated with Cluster 2. In addition, groups of admixed individuals were usually found spatially 1000 close to groups of Cluster 1 and 2 individuals (Fig. 5).

500 Discussion 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 K Genetic diversity of Solanum elaeagnifolium

Fig. 3 Plot of the second order rate of change of lnProb (D), Four AFLP primer combinations were successfully DK, as a function of the number of genetic clusters or gene pools, used to assess the genetic variation among 187 indi- K, based on STRUCTURE analysis of AFLP data, showing that viduals of S. elaeagnifolium collected from across DK value peaked at K = 2. Australia. High level of genetic diversity was detected in S. elaeagnifolium from Australia, with a mean Jaccard’s genetic similarity of all individuals at 0.26. 0.4 Combined with a previous study (Hawker et al., 2006), a high genetic diversity in S. elaeagnifolium was deter- 0.3 mined in Australia at both state and national levels. The high level of genetic variation in Australia might 0.2 be attributed to multiple introductions (Cuthbertson et al., 1976), heterogeneous and/or heterozygous nat- 0.1 ure of the initial introduction(s) and/or inter- and intra- species hybridisation events. Solanum elaeagnifo-

PCoA 2 (3.5%) PCoA 0 lium is a xenogamous species with the potential to hybridise with other Solanaceae species (Hardin et al.,

–0.1 Cluster 1 1972). Solanum esuriale, a native species often occupy- Cluster 2 ing the same habitat as S. elaeagnifolium, could poten- admixed –0.2 tially hybridise with S. elaeagnifolium. This native –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 0.3 species also belongs to the Leptostemonum subgenus PCoA 1 (11.9%) and shares the same chromosome number (n = 12) Fig. 4 Principal co-ordinates analysis (PCoA) of 187 Solanum with S. elaeagnifolium (Randell & Symon, 1976) and is elaeagnifolium individuals using four AFLP primer combinations noted to be morphologically very similar to S. elaeag- based on Jaccard’s genetic similarity, showing that individuals nifolium (Bean, 2004). Many intermediate forms that were assigned to Cluster 1 and Cluster 2 or were admixed between S. elaeagnifolium and S. esuriale were found were well separated. during our sampling trips (data not shown). Artificial hybridisation between these two Solanum species is assigned to a single cluster (Cluster 1, Cluster 2) or underway to determine whether cross-species were admixed. A total of 127 individuals from 63 loca- hybridisation occurs in the field. Preliminary results tions formed such spatial groups, which left 60 individ- showed that crosses between the two species can form uals from 31 locations that failed to form any spatial berries. The viability and identity of the suspected F1 structure, either because of mixed genotypes in the seeds are to be further investigated. Once confirmed,

Fig. 5 Spatial groups of individuals assigning to the two genetic clusters inferred by STRUCTURE analysis of AFLP data. Within each spatial group, more than 70% individuals were assigned to a single cluster (Cluster 1: dotted line ellipse, Cluster 2: solid line ellipse) or were admixed (broken line ellipse). A total of 60 individuals from 31 locations failed to form any spatial group either because of the mixed genotypes in locations or the spatial isolation from other locations (such as the location 93 in QLD). this potential interspecies hybridisation could increase by STRUCTURE). In VIC, one of the earliest records the genetic diversity in local populations of S. elaeag- of S. elaeagnifolium was in Hopetoun, where 75% indi- nifolium. viduals were assigned to Cluster 1 (location ID: 75 and 76). This result indicated the possible sites of first establishment of S. elaeagnifolium in Australia. In Genetic structure of Solanum elaeagnifolium addition, small spatially isolated groups of these two Bayesian model-based STRUCTURE analysis and genetic demes were also found in south-eastern SA and PCoA analysis achieved similar results. Individuals southern NSW, indicating that multiple introductions that were assigned to different clusters or were (Cuthbertson et al., 1976) were likely in those areas. admixed in the STRUCTURE analysis were separated Solanum elaeagnifolium fruits can be eaten by livestock by PCoA (Fig. 4). These results suggested that there (Heap & Honan, 1993). Subsequent long-distance were probably two main distinct genetic clusters in transport and trade of these livestock could have led Australia. Spatially concentrated individuals from the to the spread of contaminated faeces, thus causing the same genetic demes were found (Fig. 5). AFLP analy- spatially isolated groups of genetic demes. Sexual sis indicated that Cluster 2 was the main genetic deme crossing between these two genetic clusters may also in northern NSW, while a large spatial group of have happened as groups of admixed individuals were Cluster 1 individuals was found in western VIC and usually spatially close to groups of Cluster 1 and 2 south-western NSW. The spatial distribution of these individuals (Fig. 5). two genetic clusters correlated with the early record of The high level of genetic diversity of S. elaeagnifolium S. elaeagnifolium in Australia. Solanum elaeagnifolium identified in this study suggests that successful manage- was first detected in Bingara, NSW, where most indi- ment of this weed may be a challenging task. Weeds of viduals were assigned to Cluster 2 (location ID: 18, 19, higher genetic diversity were considered to be more dif- 20; six individuals studied in this area, with five ficult to manage than those of lower genetic diversity grouped into Cluster 2 and one identified as admixed (Dekker, 1997). Control strategies suitable for one

weed population in a given area might not be effective HEAP JW & HONAN I (1993) Weed seed excretion by sheep - for a genetically distinct population in other areas. In temporal patterns and germinability. In: Proceedings of the addition, survival rate of biocontrol agents may be dif- 10th Australian Weeds Conference and 14th Asian Pacific Weed Science Society Conference (eds. WILSON BJ & ferent between different weed genotypes. For example, SWARBRICK JT), 431–434. Primary Industries (SA), larvae survival rate of stem-mining midge, a biocontrol Brisbane, Australia. agent for Hydrilla verticillata (L.f.) Royle, was signifi- JONES CJ, EDWARDS KJ, CASTAGLIONE S et al. (1997) cantly different among the genotypes of this weed Reproducibility testing of RAPD, AFLP and SSR markers (Schmid et al., 2010). Therefore, management strate- in plants by a network of European laboratories. gies might need to be modified between genetically Molecular Breeding 3, 381–390. distinct populations. Genetic structure analysis indi- LEVIN RA, MYERS NR & BOHS L (2006) Phylogenetic cated that S. elaeagnifolium from different gene pools relationships among the “spiny solanums” (Solanum subgenus Leptostemonum, Solanaceae). American Journal was distributed across south-eastern Australia. Thus, it of Botany 93, 157–169. is important to prevent seed contamination from vari- LI RZ, WANG SW, DUAN LS et al. (2007) Genetic diversity ous sources, as seed is the main contributor to long- of wild oat (Avena fatua) populations from China and the distance dispersal. United States. Weed Science 55,95–101. MCGREGOR CE, LAMBERT CA, GREYLING MM, LOUW JH & WARNICH L (2000) A comparative assessment of DNA Acknowledgements fingerprinting techniques (RAPD, ISSR, AFLP and SSR) in tetraploid potato (Solanum tuberosum L.) germplasm. We would like to acknowledge Charles Sturt Univer- Euphytica 113, 135–144. sity, Australia, for funding this research. We thank MENCHARI Y, DELYE C&LE CORRE V (2007) Genetic a number of state and local organisations across variation and population structure in black-grass Australia for their assistance in field sampling. (Alopecurus myosuroides Huds.), a successful, herbicide- resistant, annual grass weed of winter cereal fields. Molecular Ecology 16, 3161–3172. References PETANIDOU T, GODFREE RC, SONG DS et al. (2012) Self-

BEAN AR (2004) The taxonomy and ecology of Solanum compatibility and plant invasiveness: comparing species in subg. Leptostemonum (Dunal) Bitter (Solanaceae) in native and invasive ranges. Perspectives in Plant Ecology – Queensland and far north-eastern New South Wales. Evolution and Systematics 14,3 12. Austrobaileya 6, 639–816. RANDELL BR & SYMON DE (1976) Chromosome numbers in CUTHBERTSON EG, LEYS AR & MCMASTER G (1976) Australian Solanum species. Australian Journal of Botany – Silverleaf nightshade - a potential threat to agriculture. 24, 369 379. Agricultural Gazette of New South Wales 87,11–13. ROSENBERG NA (2004) DISTRUCT: a program for the DEKKER J (1997) Weed diversity and weed management. graphical display of population structure. Molecular – Weed Science 45, 357–363. Ecology Notes 4, 137 138. EVANNO G, REGNAUT S&GOUDET J (2005) Detecting the SAMBROOK J, FRITSCH EF & MANIATIS T eds (1989) Molecular number of clusters of individuals using the software Cloning: A Laboratory Manual, 2nd edn. Cold Spring STRUCTURE: a simulation study. Molecular Ecology 14, Harbor Laboratory, New York, NY. 2611–2620. SCHMID TA, CUDA JP, MACDONALD GE & GILLMORE JL FALUSH D, STEPHENS M&PRITCHARD JK (2007) Inference of (2010) Performance of two established biological control population structure using multilocus genotype data: agents on Hydrilla genotypes susceptible and resistant to dominant markers and null alleles. Molecular Ecology ﬂuridone herbicide. Journal of Aquatic Plant Management – Notes 7, 574–578. 48, 102 105. FEUERHERDT L (2009) Overcoming a deep rooted perennial SHIVRAIN VK, BURGOS NR, AGRAMA HA et al. (2010) Genetic problem - silverleaf nightshade (Solanum elaeagnifolium)in diversity of weedy red rice (Oryza sativa) in Arkansas, – South Australia. Plant Protection Quarterly 24, 123–124. USA. Weed Research 50, 289 302. HAMMER Ø, HARPER DAT & RYAN PD (2001) PAST: STANTON RA, HEAP JW, CARTER RJ & WU H (2009) Solanum Paleontological Statistics Software Package for education elaeagnifolium. In: The Biology of Australian Weeds, Vol. – and data analysis. Palaeontologia Electronica 4,1–9. 3 (ed. FD Panetta), 274 293. R. G. and F. J. Richardson, HARDIN JW, DOERKSON G, HERNDON D, HOBSON M& Melbourne, Vic. THOMAS F (1972) Pollination and ﬂoral biology of four STANTON R, WU H&LEMERLE D (2011) Root regenerative weedy species in southern Oklahoma. Southwest Naturalist ability of silverleaf nightshade (Solanum elaeagnifolium Cav.) – 16, 403–412. in the glasshouse. Plant Protection Quarterly 26,54 56. HAWKER V, PRESTON C&BAKER J (2006) Genetic variation STODART BJ, MACKAY M&RAMAN H (2005) AFLP and within and among silverleaf nightshade (Solanum SSR analysis of genetic diversity among landraces of bread elaeagnifolium Cav.) populations in South Australia. In: wheat (Triticum aestivum L. em. Thell) from different 15th Australian Weeds Conferences (eds C PRESTON,JH geographic regions. Australian Journal of Agricultural – WATTS &NDCROSSMAN), 176. Adelaide Conference Centre, Research 56, 691 697. Adelaide, SA.

VAN DER WURFF AWG, ISAAKS JA, ERNSTING G&VAN Appendix 1 (Continued) STRAALEN NM (2003) Population substructures in the soil invertebrate Orchesella cincta, as revealed by microsatellite Location Sample Longitude/ and TE-AFLP markers. Molecular Ecology 12, 1349–1359. ID Location State size Latitude VOS P, HOGERS R, BLEEKER M et al. (1995) AFLP – a new 39 Wunkar SA 3 À34°29/140°12 technique for DNA-ﬁngerprinting. Nucleic Acids Research 40 Angas Valley SA 3 À34°44/139°19 – 23, 4407 4414. 41 Cambrai SA 1 À34°39/139°15 ZHU XC, WU HW, RAMAN H et al. (2012) Evaluation of 42 Sedan SA 1 À34°33/139°18 simple sequence repeat (SSR) markers from Solanum crop 43 Annadale SA 1 À34°24/139°21 species for Solanum elaeagnifolium. Weed Research 52, 44 Eudunda SA 1 À34°11/139°05 217–223. 45 Koonoona SA 3 À33°49/138°56 ZHU X, RAMAN H, WU H et al. (2013) Development of SSR 46 Burra SA 3 À33°41/138°55 markers for genetic analysis of silverleaf nightshade 47 Clare SA 3 À33°43/138°37 (Solanum elaeagnifolium) and related species. Plant 48 Blyth SA 1 À33°50/138°30 Molecular Biology Reporter 31, 248–254. 49 Avon SA 3 À34°15/138°20 50 Lochiel SA 1 À33°57/138°10 51 Snowtown SA 3 À33°44/138°05 Appendix 1 The Location ID, sample size and locations of 52 Crystal Brook SA 3 À33°19/138°12 Solanum elaeagnifolium collected from different states of 53 Port Pirie SA 2 À33°16/138°09 Australia: New South Wales (NSW), South Australia (SA), 54 Wirrabara SA 2 À33°02/138°16 Victoria (VIC), Queensland (QLD) and Western Australia (WA) 55 Appila 1 SA 1 À33°01/138°26 56 Appila 2 SA 1 À33°00/138°28 Location Sample Longitude/ 57 Spalding SA 1 À33°19/138°35 ID Location State size Latitude 58 Tarlee SA 1 À34°12/138°43 59 Adelaide SA 3 À34°40/138°41 1 Leeton NSW 2 À34°27/146°22 60 Murray Bridge SA 2 À35°04/139°13 2 Narrandera NSW 2 À34°46/146°25 61 Mannum SA 3 À35°00/139°14 3 Ganmain NSW 2 À34°53/146°59 62 Langhorne Creek SA 3 À35°19/139°00 4 Boree Creek NSW 2 À35°08/146°27 63 Keith 1 SA 2 À36°06/140°16 5 Yanco 1 NSW 2 À34°38/146°25 64 Keith 2 SA 2 À36°04/140°17 6 Yanco 2 NSW 2 À34°34/146°23 65 Keith 3 SA 2 À36°06/140°21 7 Grifﬁth NSW 1 À34°26/146°11 66 Mount Priscilla SA 2 À33°46/136°24 8 Cartwrights Hill NSW 2 À34°56/147°25 67 Mangalo SA 2 À33°29/136°31 9 Temora NSW 2 À34°24/147°36 68 Mitchellville SA 2 À33°35/137°04 10 West Wyalong 1 NSW 1 À34°00/147°15 69 Carwarp VIC 2 À34°28/142°10 11 West Wyalong 2 NSW 2 70 Red Cliffs VIC 2 À34°24/142°00 12 Ungarie 1 NSW 2 À33°39/146°59 71 Nhill 1 VIC 1 À36°24/141°27 13 Ungarie 2 NSW 1 À33°38/146°58 72 Nhill 2 VIC 3 À36°24/141°49 14 Ungarie 3 NSW 3 À33°36/146°55 73 Dimboola VIC 2 À36°25/142°00 15 Dubbo NSW 2 À32°11/148°48 74 Longerenong VIC 2 À36°40/142°18 16 Gilgandra NSW 2 À31°40/148°42 75 Hopetoun 1 VIC 2 À35°36/142°26 17 Coonabarabran NSW 2 À31°05/149°33 76 Hopetoun 2 VIC 1 À35°31/142°22 18 Bingara 1 NSW 2 À29°52/150°33 77 Walpeup VIC 2 À35°09/142°03 19 Bingara 2 NSW 2 À29°48/150°32 78 Echuca VIC 2 À36°07/144°52 20 Bingara 3 NSW 2 À29°49/150°32 79 Nanneella VIC 2 À36°20/144°49 21 Delungra NSW 2 À29°45/150°42 80 Rochester VIC 2 À36°23/144°46 22 Inverell NSW 2 À29°39/151°12 81 Serpentine VIC 2 À36°24/143°58 23 Tamworth NSW 2 À31°03/150°51 82 Calivil 1 VIC 2 À36°21/144°07 24 Scone NSW 2 À31°58/150°51 83 Calivil 2 VIC 2 À36°17/144°05 25 Dunedoo NSW 2 À31°58/149°30 84 Jarklin 1 VIC 1 À36°16/143°58 26 Gulgong NSW 2 À32°23/149°36 85 Jarklin 2 VIC 2 À36°14/143°56 27 Mudgee NSW 2 À32°31/149°33 86 Swan Hill VIC 2 À35°19/143°31 28 Wellington NSW 2 À32°31/148°48 87 Lake Boga VIC 2 À35°28/143°39 29 Parkes NSW 2 À33°13/148°13 88 Bridgewater VIC 2 À36°38/143°54 30 Young NSW 2 À34°27/148°19 89 Shepparton VIC 3 À36°25/145°27 31 Hay NSW 1 À34°29/145°17 90 Wunghnu VIC 3 À36°10/145°28 32 Balranald NSW 2 À34°56/143°28 91 Dookie 1 VIC 1 À36°13/145°40 33 Finley NSW 2 À35°37/145°35 92 Dookie 2 VIC 3 À36°12/145°42 34 Corowa NSW 2 À35°53/146°18 93 Inglewood QLD 3 À29°05/151°17 35 Culcairn NSW 2 À35°41/146°58 94 Katanning WA 1 À33°39/117°44 36 Morven NSW 2 À35°35/147°09 Total 187 37 Loxton 1 SA 3 À34°28/140°37 38 Loxton 2 SA 2 À34°38/140°41

Some of the markers described in Chapters 4-7, together with some micro-morphological features, were used to improve the identification of

SLN in this chapter.

Zhu, X. C., Burrows, G., Wu, H., Raman, H., Stanton, R., & Lemerle, D.

(2011). Identification of silverleaf nightshade using microsatellite markers and microstructure. Proceedings of the 23rd Asian-Pacific Weed Science

Society Conference, Cairns, Australia, pp. 604-609.

75 23rd Asian-Pacific Weed Science Society Conference

26 – 29 September 2011 The Sebel Cairns

“Weed Management in a Changing World”

Conference Proceedings Volume 1

ISBN Number: 978-0-9871961-0-1

76 23rd Asian-Pacific Weed Science Society Conference The Sebel Cairns, 26-29 September 2011 IDENTIFICATION OF SILVERLEAF NIGHTSHADE USING MICROSATELLITE MARKERS AND MICROSTRUCTURE

Xiao Cheng Zhu1 2, Geoffrey Burrows1 2, Hanwen Wu1 3, Harsh Raman1 3, Rex Stanton1 2, and Deirdre Lemerle 1 2 1EH Graham Centre for Agricultural Innovation (an alliance between Industry & Investment NSW and Charles Sturt University) NSW 2650, Australia. 2School of Agriculture and Wine Science, Charles Sturt University, Wagga Wagga, NSW 2678, Australia, 3Wagga Wagga Agricultural Institute, PMB, Wagga Wagga, NSW 2650, Australia.

ABSTRACT Silverleaf nightshade (Solanum elaeagnifolium Cav.) originated in America and is a serious summer-growing perennial weed in Australia. It is often confused with the native Solanaceae species quena (S. esuriale Lindl.). Both belong to the ―Leptostemonum‖ subclass in the Solanaceae and are remarkably similar in their morphological traits. Correct identification is critical for the successful management of S. elaeagnifolium, as different biotypes could vary significantly in their response to control measures, such as herbicides and biocontrol agents. In order to improve the identification of S. elaeagnifolium, DNA polymorphism and microstructure of S. elaeagnifolium and S. esuriale were compared. Thirteen cross-species simple sequence repeat (SSR) primer pairs were utilized to investigate the polymorphism between S. elaeagnifolium and S. esuriale. SSR markers clearly separated the two species. Three unique SSR alleles were present in S. esuriale but not in S. elaeagnifolium, which could be used to distinguish the two species. Scanning electron microscope (SEM) examination of the microstructure of the leaf surface of these two species showed that the complex stellate trichomes on the upper leaf surface of S. elaeagnifolium had a deep ―root‖ structure penetrating into the palisade mesophyll, while this structure was not found in S. esuriale. Combination of molecular phylogeny and SEM will considerably assist in the correct identification of S. elaeagnifolium.

Key Words: silverleaf nightshade, quena, perennial weed, SSR, SEM, trichome

INTRODUCTION

Silverleaf nightshade (Solanum elaeagnifolium Cav.) is a deep-rooted, summer-growing perennial weed which originated in America (Heiser and Whitaker 1948; Stanton et al. 2009). This invasive weed infests at least 350,000 ha in Australia and with the potential to infest 398 million ha (Kwong 2006; Feuerherdt 2009).

Correct identification of silverleaf nightshade is required for selection of herbicides and biocontrol agents (Nissen et al. 1995; Lopez-Martinez et al. 1999). However, silverleaf nightshade often confused with an Australian native Solanaceae species, quena (S. esuriale Lindl.). The misidentification resulted in delays to control of silverleaf nightshade in South Australia (Hosking et al. 2000). Currently differentiation between the two species is based on morphological characteristic such as stamens length, spine density or fruit shape (Kidston et al. 2006). However, these morphological traits vary considerably in

ISBN Number: 978-0-9871961-0-1 604

77 23rd Asian-Pacific Weed Science Society Conference The Sebel Cairns, 26-29 September 2011

Australian silverleaf nightshade populations (Stanton, et al. 2009). Identification based solely on morphological traits is unreliable.

Molecular markers have been widely used in Solanum species to delineate species and cultivars (Chimote et al. 2004) . For example, Chimote, et al. (2004) used simple sequence repeat (SSR) markers to differentiate 32 Indian cultivars of potato (S. tuberosum). In addition, previous studies have used molecular markers to identify the main clades and phylogenetic relationships within Solanum (Bohs 2005; Weese and Bohs 2007). The Australian native species quena was not included in these previous studies.

Micro-morphological parameters, such as leaf trichomes were considered as some of the most distinguishing features in Solanum (Roe 1971). Bean (2004) noted that silverleaf nightshade was so similar in morphology to quena that microscopic examination was usually required for identification.

In this study, micro-morphological traits and SSR markers were used to differentiate silverleaf nightshade from the Australian native Solanum species quena.

MATERIAL AND METHODS

Molecular Analysis

Silverleaf nightshade leaf samples were collected from Jarklin, Shepparton and Lake Boga (Victoria), and Corowa, Morven and Gulgong (New South Wales). Quena samples were collected from Jarklin (Victoria) and Ungarie, Wellington and Wagga Wagga (New South Wales). Genomic DNA was extracted from leaf material using the standard phenol/ chloroform method (Sambrook et al. 1989). The DNA samples of each species from the same location were bulked for PCR amplification. Thirteen SSR primer-pairs from other Solanum species were used in this study. Primer details were those mentioned in previous study of Zhu et al. (2011). The 5‘ end of the forward primer of each SSR primer- pair was tailed with M13 sequence and PCR amplification and detection of the amplification products were carried out as described by Raman et al. (2005). The alleles were scored in a binary form as the presence or absence (1 or 0) of bands of each SSR primer-pairs for each population and data were analysed using PAST (Hammer et al. 2001).

Micro-morphological Analysis

Seven populations of silverleaf nightshade from Jarklin, Shepparton, Lake Boga, Corowa, Morven, Gulgong and Wagga Wagga and four quena populations from Jarklin, Ungarie, Wellington and Wagga Wagga were observed in this study. The 4th leaf from the shoot apex was collected from each individual. Only the adaxial surface of each leaf was examined in this study. Trichomes were removed with forceps from fresh leaves and placed on a 12 mm carbon tab (ProSciTech, Australia). A total of 195 and 74 trichomes were observed for silverleaf nightshade and quena, respectively. Images of these trichomes were obtained using a scanning electron microscope (SEM) (JEOL JCM 5000 NeoScope, Japan). The length of the ―root‖ structure of each trichome was measured using software ImageJ (Ferreira and Rasband 2010). Data were analyzed using unpaired two sample t-test of GenStat 13.0 (Buysse et al. 2004).

ISBN Number: 978-0-9871961-0-1 605

78 23rd Asian-Pacific Weed Science Society Conference The Sebel Cairns, 26-29 September 2011

RESULTS

Quena populations were clearly differentiated from silverleaf nightshade populations by the unweighted pair group method with arithmetic mean (UPGMA) dendrogram based on the Jaccard‘s similarity indicating the genetic variability between related species (Figure 1). In addition, three unique alleles were amplified in quena: fragment 85 bp with primer-pair EM 117, 222 bp with EM 135 and 249 bp with ESM 3. These unique alleles could be utilized to distinguish quena from silverleaf nightshade.

Figure 1. UPGMA dendrogram clearly separated quena (Q) and silverleaf nightshade (S) populations.

Significant difference on the length of trichome ―root‖ structure (P < 0.001) was found between these two species (Figure 2). The average lengths of trichomes of silverleaf nightshade and quena were 134 and 33 μm, respectively.

ISBN Number: 978-0-9871961-0-1 606

79 23rd Asian-Pacific Weed Science Society Conference The Sebel Cairns, 26-29 September 2011

Figure 2. A comparison of trichome structure from adaxial leaf surface between quena and silverleaf nightshade. A: quena; B: silverleaf nightshade. Arrows indicate the root structure under epidermal level.

DISCUSSION

Silverleaf nightshade and quena were separated at a level of about 10% similarity (Figure 1), indicating the great genetic divergence between these two species. Furthermore, the unique alleles found in quena will provide a reliable method to distinguish these two species.

This is the first report on the examined morphological characteristics of quena. Silverleaf nightshade had a much longer ―root‖ structure of the trichome than quena, which has been highlighted by previous studies (Bruno et al. 1999; Christodoulakis et al. 2009). This ―root‖ deeply penetrates into the palisade mesophyll making it very difficult to pull off from the leaf. The significant difference on trichome structures between quena and silverleaf nightshade found in this study can be considered diagnostic in order to discriminate these two species. This possibility of the impacts of this structure on herbicide uptake needs to be further tested.

Correct identification of silverleaf nightshade will improve the weed management. Generally, quena is easier to manage than silverleaf nightshade (Johnson et al. 2006), therefore correct identification will help the herbicide selection and management strategies. In addition, reliable identification is also required for biocontrol, as agent/ weed compatibility has a significant influence on agent success (Nissen, et al. 1995).

ACKNOWLEDGEMENTS

We would like to acknowledge the Charles Sturt University, Australia for funding this research and to EH Graham Centre for Agricultural Innovation for the financial support to attend the conference.

REFERENCES

Bean, A. R. (2004). The taxonomy and ecology of Solanum subg. Leptostemonum (Dunal) Bitter (Solanaceae) in Queensland and far north-eastern New South Wales. Austrobaileya 6(4): 639-816.

ISBN Number: 978-0-9871961-0-1 607

80 23rd Asian-Pacific Weed Science Society Conference The Sebel Cairns, 26-29 September 2011

Bohs, L. (2005). Major clades in Solanum based in ndhF sequences In R. C. Keating, V. C. Hollowell and T. B. Croat (eds.), A festschrift for William G. D’Arcy: the legacy of a taxonomist. Monographs in Systematic Botany from the Missouri Botanical Garden. Missouri Botanical Garden 104. St. Louis: Missouri Botanical Garden. Vol. pp. 27-49.

Bruno, G., Cosa, M. T., and Dottori, N. (1999). Ontogenetic development of stellate trichomes in Solanum elaeagnifolium (Solanaceae). Kurtziana 27(1): 169-172.

Buysse, W., Stern, R., and Coe, R. (2004). GenStat discovery edition for everyday use. ICRAF Nairobi, Kenya.

Chimote, V. P., Chakrabarti, S. K., Pattanayak, D., and Naik, P. S. (2004). Semi- automated simple sequence repeat analysis reveals narrow genetic base in Indian potato cultivars. Biologia Plantarum 48(4): 517-522.

Christodoulakis, N. S., Lampri, P. N., and Fasseas, C. (2009). Structural and cytochemical investigation of the leaf of silverleaf nightshade (Solanum elaeagnifolium), a drought- resistant alien weed of the Greek flora. [Article]. Australian Journal of Botany 57(5): 432- 438.

Ferreira, T., and Rasband, W. (2010). The ImageJ User Guide —Version 1.43. http://rsbweb.nih.gov/ij/docs/user-guide.pdf

Feuerherdt, L. (2009). Overcoming a deep rooted perennial problem - silverleaf nightshade (Solanum elaeagnifolium) in South Australia. Plant Protection Quarterly 24(3): 123-124.

Hammer, Ø., Harper, D. A. T., and Ryan, P. D. (2001). PAST: Paleontological Statistics Software Package for education and data analysis. Palaeontologia Electronica 4(1): 9.

Heiser, C. B., Jr., and Whitaker, T. W. (1948). Chromosome Number, Polyploidy, and Growth Habit in California Weeds. American Journal of Botany 35(3): 179-186.

Hosking, J. R., Sainty, G. R., and Jacobs, S. W. L. (2000). Certainty and uncertainty in plant identification. The new Mexica botanist 15: 1-8.

Johnson, A., Brooke, G., Thompson, R., Roberts, K., Hertel, K., Border, N., McNee, T., and Sullivan, P. (2006). Weed control for cropping and pastures in central west NSW. http://www.dpi.nsw.gov.au/agriculture/pests-weeds/weeds/publications/central-west

Kidston, J., Thompson, R., and Johnson, A. (2006). Silverleaf nightshade. from NSW Department of Primary Industries:

Kwong, R. (2006). Feasibility of biological control of solanaceous weeds of temperate Australia (Final report No. Weed 1.20). North Sydney.

Lopez-Martinez, N., Salva, A. P., Finch, R. P., Marshall, G., and De Prado, R. (1999). Molecular markers indicate intraspecific variation in the control of Echinochloa spp. with quinclorac. Weed Science 47(3): 310-315.

Nissen, S. J., Masters, R. A., Lee, D. J., and Rowe, M. L. (1995). DNA-based marker systems to determine genetic diversity of weedy species and their application to biocontrol. Weed Science 43(3): 504-513.

ISBN Number: 978-0-9871961-0-1 608

81 23rd Asian-Pacific Weed Science Society Conference The Sebel Cairns, 26-29 September 2011

Raman, R., Raman, H., Johnstone, K., Lisle, C., Smith, A., Matin, P., and Allen, H. (2005). Genetic and in silico comparative mapping of the polyphenol oxidase gene in bread wheat (Triticum aestivum L.). Functional & Integrative Genomics 5(4): 185-200.

Roe, K. E. (1971). Terminology of Hairs in the Genus Solanum. Taxon 20(4): 501-508.

Sambrook, J., Fritsch, E. F., and Maniatis, T. (Eds.). (1989). Molecular cloning : a laboratory manual (Second ed.). New York: Cold Spring Harbor Laboratory.

Stanton, R. A., Heap, J. W., Carter, R. J., and Wu, H. (2009). Solanum elaeagnifolium (Vol. 3). Melbourne: R. G. and F. J. Richardson.

Weese, T. L., and Bohs, L. (2007). A three-gene phylogeny of the genus Solanum (Solanaceae). Systematic Botany 32(2): 445-463.

Zhu, X. C., Wu, H. W., Raman, H., Lemerle, D., and Stanton, R. (2011). Assessment of genetic variation in silverleaf nightshade (Solanum elaeagnifolium Cav.) using cross- species SSR markers. submit to weed research.

ISBN Number: 978-0-9871961-0-1 609

82 CHAPTER 9 PUBLISHED PAPER

The previous chapters highlighted the morphological and genetic variation of SLN in Australia (Chapters 3-7). In Chapter 9 the reproductive strategies of SLN were studied. Modes of reproduction could be one of the important factors affecting the morphological and genetic variation.

Zhu, X. C., Wu, H. W., Stanton, R., Raman, H., Lemerle, D., & Burrows, G.

E. (2013). Time of emergence impacts the growth and reproduction of silverleaf nightshade (Solanum elaeagnifolium Cav.). Weed Biology and

Management. doi:10.1111/wbm.12015.

83 bs_bs_banner

Weed Biology and Management ••, ••–•• (2013)

RESEARCH PAPER Time of emergence impacts the growth and reproduction of silverleaf nightshade (Solanum elaeagnifolium Cav.)

XIAOCHENG ZHU1,2*, HANWEN WU1,3, REX STANTON1,2, GEOFFREY E. BURROWS1,2, DEIRDRE LEMERLE1,2 and HARSH RAMAN1,3 1EH Graham Centre for Agricultural Innovation, 2School of Agricultural and Wine Sciences, Charles Sturt University and 3Wagga Wagga Agricultural Institute, Wagga Wagga, New South Wales, Australia

Silverleaf nightshade is one of the worst agricultural weeds on a worldwide basis. Improved understanding of its life cycle will be useful for weed management.This research showed that the growth and reproduction of silverleaf nightshade were affected significantly by the time of emergence. Plants (both root- and seed-generated) that emerged early in the growing season, such as in September or November (spring), were significantly taller and produced more biomass and fruit than plants that emerged in January (summer) and March (autumn). Delayed emergence resulted in a shorter vegetative phase and less fruit production. As a result, the silverleaf nightshade plants that emerged in September produced a large amount of seeds, while the plants that emerged later in the season did not flower. Thus, control of early-emerging plants is important.The dynamics of fruit production indicated that silverleaf nightshade fruit formed in December and peaked in March, which suggests that a control action is required before December to control the soil seed bank.

Keywords: emergence delay, perennial weed, reproduction, root, seed.

Silverleaf nightshade (Solanum elaeagnifolium Cav.) origi- Silverleaf nightshade is a deep-rooted, summer- nated in the Americas and has been considered to be one growing perennial that reproduces sexually through a of the worst agricultural weeds around the world. It is a seed bank and vegetatively through its root bank. Under noxious weed in Australia, South Africa and Morocco Australian climates, silverleaf nightshade emerges from and also has been recorded in India, Pakistan and China September to April, ﬂowers and sets seed from Novem- (Mekki 2007). Silverleaf nightshade competes with crops ber to March, with the above-ground parts senescing in and pastures for soil water and nutrients, causing Յ77% May.The life cycle repeats with shoots regenerating from yield loss in cereal (Heap & Carter 1999) and 64% in the underground perennial root systems. Sexual repro- corn (Baye & Bouhache 2007). It is also neurotoxic to duction is critical for long-distance dispersal of this weed cattle (Boyd et al. 1984). Silverleaf nightshade recently (Richardson & McKenzie 1981; Wapshere 1988). By has been listed as one of the Weeds of National Signiﬁ- contrast, vegetative propagation mainly contributes to cance in Australia (Australian Weeds Committee 2012), localized infestations and reinfestation (Stanton et al. infesting at least 350 000 ha in Australia, with the poten- 2011). The root system of silverleaf nightshade can tial to infest 398 million ha (Feuerherdt 2009). extend up to 2.8 m in depth.All parts of the root system have a strong ability to regenerate and can survive for Communicated by M. Nashiki. Յ15 months under moist conditions (Mekki 2007; *Correspondence to: Xiaocheng Zhu, School of Agricultural andWine Stanton et al. 2011). The roots can be fragmented and Sciences, Charles Sturt University, Wagga Wagga, NSW 2678, spread by cultivation and machinery. Australia. It is estimated that each silverleaf nightshade plant can Email: [email protected] produce 40–60 fruits and each mature fruit contains Received 15 November 2012; accepted 1 June 2013 24–149 seeds (Boyd & Murray 1982a). The soil seed doi:10.1111/wbm.12015 © 2013 Weed Science Society of Japan

84 2 X. Zhu et al. bank of silverleaf nightshade can persist for at least manually from a roadside near Narrandera, NSW,Aus- 10 years (Boyd & Murray 1982a). Silverleaf nightshade tralia (-34.77°N, 146.43°E), in January 2011.The fruits seedlings usually emerge after heavy rainfall between were air-dried and crushed to collect the seeds, which spring and autumn (Stanton et al. 2009). Anecdotal evi- were washed and pregerminated as described by Stanton dence suggests that these seedlings do not flower during et al. (2012). Ten pregeminated seeds were transplanted the first year of growth because of subsequent dry con- per pot at a 1 cm depth in September, November, ditions in summer or frost in winter. January and March 2011, separately, to mimic seedling Ecological and phenological studies provide impor- emergence over time. The plants were thinned to six tant information for predicting the infestation density seedlings per pot 25 days after sowing. and fruit and seed production dynamics, which are criti- The January and March cohorts were terminated in cal for weed management (Swanton & Murphy 1996), May 2011, while the September and November cohorts especially for perennial weeds propagating both sexually were terminated in May 2012.The plant height (PH) and and asexually (Navas 1991). Silverleaf nightshade seed- fruit number (NOF) were recorded monthly and the lings from seeds that emerged later in the season pro- number of days to the first flower (TF) also was recorded. duced a significantly lower amount of plant dry weight The root (RDW) and shoot (SDW) dry weights were and fruit in the USA (Boyd & Murray 1982b). However, determined after harvest in May.The average fruit diam- information is not available on the impacts of the delayed eter (AFD), average seeds per fruit (ASF), thousand-seed emergence on the growth and development of root- weight (TSW) and average seed weight per fruit (ASWF) regenerated silverleaf nightshade. The objective of this were measured on 40 randomly selected fruits per pot study was to estimate the impact of the timing of seed- (or all fruits if <40 fruits were produced).The total seed ling (from seeds) and stem emergence (from roots) on production (TSP) and total seed weight (ToSW) were the growth and reproduction of silverleaf nightshade. calculated using the above data. The weather data were obtained from the Wagga Wagga Agricultural Institute’s meteorological station and the growing degree days MATERIALS AND METHODS (GDD) were calculated by using the method described by Liu and Bull (2001), where the base temperature, The experimental design used four emergence times, optimum temperature and ceiling temperature for silver- arranged in a randomized complete block design with leaf nightshade were assumed to be 10°C, 30°C and three replicates.All the plants were established initially in 40°C, respectively. a glasshouse (25°C for 16 h and 10°C for 8 h) for 1 month, then transferred to the outdoor horticultural area of Charles Sturt University (-35.05°N, 147.35°E), Experiment 2: Stem growth and reproduction Wagga Wagga,Australia, and maintained under ambient This experiment was conducted between March 2011 climate conditions. The average maximum and average and May 2012. The roots of silverleaf nightshade were minimum temperatures of the horticultural area were collected from Narrandera in March, September and 20.9°C and 7.9°C, respectively, during the experimental November 2011 and January 2012. The root samples period and the mean daily evaporation rate was 3.4 mm were placed in ziplock plastic bags with moist soil (Australian Bureau of Meteorology, 2012, unpublished for transport.The roots were trimmed to 10 cm lengths data).The plants were grown in 20 cm diameter pots that and buried horizontally at a 4 cm depth in trays were filled with a 4:1 mixture of sandy loam and potting (13 cm ¥ 8cm¥ 4.5 cm) in March, September and mix containing 10 g of slow-release fertilizer (16.6% November in 2011 and January in 2012, using the same nitrogen, 2% phosphorus, 6.6% potassium and 7.9% soil medium as described above. The trays were main- sulphur; Osmocote, Scotts Australia, Bella Vista, NSW, tained in a glasshouse for 1 month while stem emer- Australia) and they were watered as needed. The terms gence occurred. The emerged plants then were “seedling” and “stem” are used to describe the plants that transplanted into 20 cm diameter pots (one plant in each were generated from the seeds and root fragments, pot) in the horticultural area. The March cohort was respectively. terminated in May 2011, while the September, Novem- ber and January cohorts were terminated in May 2012. The same measurements were taken as in Experiment 1, Experiment 1: Seedling growth with an additional measurement of the number of stems and reproduction (NOSt) for the root-generated plants. The experiment was conducted between January 2011 The data were analyzed by using a one-way ANOVA and May 2012.Silverleaf nightshade fruits were harvested in Genstat 5 (Payne et al. 1995). All the data were

85 Emergence time impacts S. elaeagnifolium 3

Table 1. Vegetative production of the silverleaf nightshade cohorts that were generated from seed or root fragments (as assessed in May)

Seedling Stem

Cohort September November January March September November January March

RDW (g) 4.73a 4.74a 0.06b 0.02c 9.33a 3.62b 3.19b 0.05c SDW (g) 10.58a 4.20a 0.03b 0.02b 58.65a 15.59b 3.66c 0.14d PH (cm) 49.07a 38.40a 3.26b 2.15c 49.93a 30.13b 16.93c 1.40d NOSt NA NA NA NA 1.67a 1.33a 2.00a 1.00a

Values sharing the same letter are not significantly different between cohorts according to Fisher’s Least Significant Difference test (P < 0.05). NA, not available; NOSt, number of stems; PH, plant height; RDW,root dry weight; SDW,shoot dry weight. log-transformed before analysis to normalize the vari- (110.00 days). The corresponding GDDs were 1005.76 ances. Fisher’s Least Significant Difference test at and 871.57 for the plants that emerged in September and P < 0.05 was used to determine the statistically different November, respectively. means. The NOF for the seed-generated plants decreased with delayed emergence (Table 2), resulting in a higher number of fruit being produced in the September RESULTS cohort (21.54 fruits per plant), compared to the Vegetative growth November cohort (5.49 fruits per plant). The TSP was significantly (P < 0.01) different between the September The time of emergence significantly (P < 0.001) affected (1200.28 seeds per plant) and November (218.48 seeds the RDW, SDW and PH of both the seed- and root- per plant) cohorts. In addition, the silverleaf nightshade generated silverleaf nightshade plants (Table 1). The plants that emerged in September had a larger ASWF plants, either seed- or root- generated, that emerged in and ToSW. However, the time of emergence did not March only produced a small amount of root (0.02– affect the TSW, AFD and ASF of the seed-generated 0.05 g per plant) prior to the first frost in May 2011, plants. while a large amount of root growth was produced by The reproduction of root-generated silverleaf night- the September cohort.The plants that emerged in early shade was affected by the time of emergence (Table 2). and late spring (September and November) grew taller The stems from the September, November and January and had a larger root and shoot biomass than those that cohorts flowered and set seed, but those that emerged in emerged in January and March. For example, the root- March did not flower before the final harvest in May. generated silverleaf nightshade plants from the Sep- Similarly to the seedlings, a significantly (P < 0.001) tember cohort (49.93 cm) were 36-fold taller than the longer vegetative phase (TF) was observed in the Sep- root-generated plants in the March cohort (1.40 cm). tember and November cohorts. The first flowers were Similarly, the shoot dry weight was 58.65 g in the Sep- observed at 79.67,63.33 and 50.00 days after planting for tember cohort but only 0.14 g in the March cohort for the September, November and January cohorts, respec- the root-generated silverleaf nightshade plants. However, tively (Table 2), which corresponded to 706.64, 643.06 a delay in the emergence time did not impact the NOSt and 449.50 GDD, respectively. from the root-generated plants. The stems that emerged early in each growing season produced more (P < 0.001) fruits and seeds than those that emerged later (Table 2). A delayed emergence Reproductive growth reduced the FDW, NOF, TSP and ToSW. The NOF Delayed emergence affected the reproduction of seed- produced by each individual decreased from 121.67 to generated silverleaf nightshade plants (Table 2). The 4.00 fruits per plant for the September and January seedlings that emerged early in each season (September cohorts, respectively.A very limited number of seeds was and November) flowered and set seed, while those that produced by the January cohort (51.33 seeds per plant), emerged after January did not flower in this study.The compared to the number that emerged in September TF was significantly (P < 0.05) shorter for the Novem- (10,405.04 seeds per plant). In addition, the plants that ber cohort (83.67 days) than for the September cohort emerged early produced a larger AFD (12.72 mm

86 4 X. Zhu et al.

Table 2. Key reproductive features of silverleaf nightshade

Trait Seedling† Stem†

September November September November January

TF (days) 110.00a 83.67b 79.67a 63.33b 50.00c GDD 1005.76a 871.57b 706.64a 643.06b 449.50c FDW (g) 5.72a 1.44b 35.29a 7.73b 0.60c NOF 21.54a 5.49b 121.67a 33.33b 4.00c AFD (mm) 12.46a 11.42a 12.72a 10.28b 7.49c ASF 55.84a 39.03a 85.77a 37.93a 11.26b ASWF (g) 0.19a 0.09b 0.23a 0.10a 0.02b TSW (g) 3.50a 2.52a 2.69a 2.53a NA ToSW (g) 4.18a 0.52b 27.67a 3.21b 0.10c TSP 1200.28a 218.48b 10 405.04a 1261.33b 51.33c

† Seedlings from the January and March cohorts and stems from the March cohort did not reach reproductive growth stage.Values sharing the same letter are not significantly different between cohorts according to Fisher’s Least Significant Difference test (P < 0.05). AFD, average fruit diameter; ASF,average seed number per fruit; ASWF,average seed weight per fruit; FDW,fruit and flower dry weight per plant; GDD, growth degree days; NA, not available; NOF,number of fruits per plant;TF,time to first flowering;ToSW,total seed weight per plant;TSP,total seed production per plant; TSW,thousand-seed weight (not available for the root-generated plants from the January cohort due to limited seed production).

compared to 7.49 mm in the September and January 25 cohorts, respectively) and had more ASF (85.77 seeds per fruit compared to 11.26 seeds per fruit for those that 20 emerged in September and January, respectively). The ASWF was also signiﬁcantly smaller for the plants that emerged in January. However, delayed emergence did 15 not affect the TSW of root-generated silverleaf nightshade (Table 2). 10

Dynamics of silverleaf nightshade reproduction 5 Numberfruit ofper plant A delay in emergence resulted in a delay in fruit forma- tion and a reduction in fruit production in the seed- 0 generated plants (Fig. 1).The fruit formed in December Sep Oct Nov Dec Jan Feb Mar Apr May or January.The number of fruits sharply increased after Time (month) December, peaked in March and then ﬂuctuated slightly until May.The September cohort produced 22.29 fruits, Fig. 1. Fruit production of the seedlings that emerged in ᭡ when assessed in March, with ~3.72 fruits formed per different months. ( ), September cohort; (᭹), November month, while 5.23 fruits were produced by the Novem- cohort. ber cohort, with ~1.30 fruits formed per month. The dynamics of fruit production in the root- generated plants was very similar to that of the seedlings, November cohorts, respectively. By contrast, the January but with a much larger amount of fruit production per cohort only produced 4.00 fruits prior to the ﬁnal plant (Fig. 2).The fruit of the September and November harvest in May. cohorts formed in December and January, respectively, rapidly increased and peaked in March for both the DISCUSSION plants that emerged in September (126.00 fruits per plant) and those that emerged in November (35.00 fruits This study highlights the importance of emergence per plant). On average, 21.00 and 8.75 fruits were pro- timing on the growth and reproduction of silver- duced per month by the stems from the September and leaf nightshade. Delayed emergence reduced the fruit

87 Emergence time impacts S. elaeagnifolium 5

160 the photoperiod also might affect the life cycle of silverleaf nightshade, as found in other species such as wild 140 radish (Raphanus raphanistrum) (Norsworthy et al. 2010). 120 Further experiments under a controlled environment could be conducted in order to investigate the impact of 100 the photoperiod on ﬂowering. 80 Although the seedling and stem emergence experiments were independent and could not be compared 60 statistically,the root-generated plants obviously produced

40 many more fruits (121.67 fruits per plant) and seeds Number fruitof plant per (10,405.04 seeds per plant). The stems from the root 20 fragments developed more rapidly and required much

0 shorter GDDs for flowering, compared to the seedlings, Sep Oct Nov Dec Jan Feb Mar Apr May which was probably related to the larger amount of non-structural carbohydrates in the root fragments that Time (month) encouraged growth and development. Competition Fig. 2. Fruit production of the stems that emerged in between individuals might have occurred in this study different months. (᭝), September cohort; (᭺), November (six seedlings per pot) and impacted fruit production, but cohort; (◊), January cohort. the effect would be minor, as the fruit production of these seedlings was similar to the non-competition field experiment in the USA (Boyd & Murray 1982b). Con- production of both the seedlings and stems. A previous sidering the extensive root systems, fruit and seed pro- investigation produced similar results in seed-generated duction in the field can be even larger, which highlights silverleaf nightshade in the USA, with the maximum the importance of rootbank control. plant biomass and fruit production being detected in the plants that emerged in spring (Boyd & Murray 1982b). CONCLUSION Anecdotal evidence has suggested that silverleaf nightshade seedlings do not flower in their first year under These pot studies have showed the significant effect of Australian climatic conditions. However, this study sug- the emergence time on silverleaf nightshade growth and gested that silverleaf nightshade seedlings that emerge reproduction. However, silverleaf nightshade is a peren- before November can flower and set seed during the first nial weed, whose root system can grow up to 2.8 m in year of growth, especially when there is sufficient soil depth. Such massive root systems, as compared to the moisture. Stanton et al. (2012) demonstrated that silver- plant, starting from 10 cm root fragments as used in this leaf nightshade germination improved with reduced study, could have a potential impact on the growth and osmotic stress.This research showed that the plants that development of this weed. Nevertheless, the difference in emerged in September produced an average of 21.54 root systems might have limited impacts on the NOF fruits and 1200.28 seeds per plant prior to the final and ASF.The NOF and ASF from the root-generated harvest in May,which is similar to Boyd’s (1982b) study, plants (September and November cohorts) were with 13.3 fruits per plant. In addition, this research iden- between 33 and 122 fruits per plant and 38 and 86 seeds tified that silverleaf nightshade fruits rapidly increase per fruit, respectively,which was similar to the 45 and 74 from December to March (Figs 1,2). It is therefore nec- fruits per plant and 38 and 89 seeds per fruit, respectively, essary to control silverleaf nightshade at the flowering that were reported previously by Stanton et al. (2012) in stage before December to avoid the replenishment of the a field survey in Australia. It is important to control the soil seed bank. Silverleaf nightshade fruits and seeds early-emerging plants to stop seedset. Under Australian could be spread potentially by livestock, machinery and climatic conditions, silverleaf nightshade seedlings can fodder movements and infect new areas. It is therefore flower and set seed during the first year of growth; thus, important to control these seedlings to minimize the the early detection of seedlings is needed so that they can input into the seed bank and root bank. In addition, be controlled in order to avoid the establishment of a silverleaf nightshade required various GDDs to flower at persistent root bank in coming years. This study also different emergence times. The September cohort highlighted that silverleaf nightshade fruit form in required more GDDs than the November cohort, irre- December and sharply increase until March. Therefore, spective of whether the plants were generated from seeds management before December is important in order to or roots.These results indicate that, besides temperature, avoid the replenishment of the soil seed bank.

88 6 X. Zhu et al.

ACKNOWLEDGMENT Mekki M. 2007. Biology, distribution and impacts of silverleaf nightshade (Solanum elaeagnifolium Cav.). Bull. EPPO 37, The authors wish to acknowledge Charles Sturt Univer- 114–118. Navas M.L. 1991. Using plant population biology in weed research: a sity,Australia, for funding this research. strategy to improve weed management. Weed Res. 31, 171–179. Norsworthy J.K., Malik M.S., Riley M.B. and Bridges W. 2010. Time of emergence affects survival and development of wild radish REFERENCES (Raphanus raphanistrum) in South Carolina. Weed Sci. 58, 402–407. Australian Weeds Committee 2012. Weeds of National Signiﬁcance: Payne R.W.,Lane P.W.,Baird D.B., Ainsley A.E., Bicknell K.E., Digby Silverleaf Nightshade (Solanum elaeagnifolium) Draft Strategic Plan. P.G.N. et al. 1995. Genstat 5 Release 3.2 Reference Manual Supplement. Australian Weeds Committee, Canberra. Lawes Agricultural Trust, Harpenden, UK. Baye Y. and Bouhache M. 2007. Competition between silverleaf Richardson R.G. and McKenzie D.N. 1981. Regeneration of, and nightshade (Solanum elaeagnifolium Cav.) and spring maize (Zea mays toxicity of 2,4-D to, root fragments of silver-leaf nightshade (Solanum L.). Bull. OEPP 37, 129–131. elaeagnifolium Cav.). J.Aust. Inst. Agric. Sci. 47, 48–50. Boyd J.W. and Murray D.S. 1982a. Effects of shade on silverleaf Stanton R., Wu H. and Lemerle D. 2011. Root regenerative ability of nightshade (Solanum elaeagnifolium). Weed Sci. 30, 264–269. silverleaf nightshade (Solanum elaeagnifolium Cav.) in the glasshouse. Boyd J.W. and Murray D.S. 1982b. Growth and development of Plant Prot. Q. 26, 54–56. silverleaf nightshade (Solanum elaeagnifolium). Weed Sci. 30, 238–243. Stanton R., Wu H.W. and Lemerle D. 2012. Factors affecting Boyd J.W.,Murray D.S. and Tyrl R.J. 1984. Silverleaf nightshade, silverleaf nightshade (Solanum elaeagnifolium) germination. Weed Sci. Solanum elaeagnifolium, origin, distribution, and relation to man. Econ. 60, 42–47. Bot. 38, 210–217. Stanton R.A., Heap J.W.,Carter R.J. and Wu H. 2009. Solanum Feuerherdt L. 2009. Overcoming a deep rooted perennial problem – elaeagnifolium. In: The Biology of Australian Weeds (ed. by Panetta F.D.). silverleaf nightshade (Solanum elaeagnifolium) in South Australia. Plant R.G. and F.J. Richardson, Melbourne, 274–293. Prot. Q. 24, 123–124. Swanton C.J. and Murphy S.D. 1996. Weed science beyond the weeds: Heap J.W. and Carter R.J. 1999. The biology of Australian weeds. 35. the role of integrated weed management (IWM) in agroecosystem (Solanum elaeagnifolium Cav.). Plant Prot. Q. 14, 2–12. health. Weed Sci. 44, 437–445. Liu D.L. and Bull T.A. 2001. Simulation of biomass and sugar Wapshere A.J. 1988. Prospects for the biological control of silver-leaf accumulation in sugarcane using a process-based model. Ecol. Model. nightshade, Solanum elaeagnifolium, in Australia. Aust. J.Agric. Res. 39, 144, 181–211. 187–197.

CHAPTER 10 GENERAL DISCUSSION AND

CONCLUSIONS

This is the first large scale study of SLN diversity in Australia, covering 94 locations. To avoid any misidentification, individuals at anthesis were sampled and roughly identified according to the corolla lobe colour, prickle density, calyx shape and fruit shape. All of the SLN and S. esuriale individuals were further analysed using the species-specific marker

EM 135 (Chapter 8). This study improved the understanding of SLN on morphology and genetics, provided reliable methods for SLN identification, highlighted the possible correlation between SLN genotypes, phenotypes and abiotic factors, and partly explained the reason for its wide distribution.

The reproductive strategies of SLN have enabled it to establish and spread in Australia.

Morphological variation

High morphological variation was found in the collection of 642

SLN individuals from south-eastern Australia for characteristics such as leaf length, width, area and roundness, plant height and prickle, trichome and stomatal densities (Chapter 3). Leaf area ranged from 0.41 to 25.8 cm2.

Larger leaves could intercept more herbicide droplets than smaller leaves

(Richburg et al., 1994), thus increasing the amount of herbicide available for uptake. However, retaining more herbicide on the leaf does not necessarily increase uptake (Kraemer, et al., 2009; Leaper & Holloway,

2000), as droplets can be retained on trichomes and evaporated without contacting with the leaf surface (Xu, et al., 2011).

91 High trichome densities were found in this study (Chapter 3), with an average adaxial (upper surface) and abaxial (lower surface) trichome density at 67 and 131 trichome/mm2, respectively, which are much higher than reported for nine other Solanum species (0.84 – 7.13 trichome/mm2)

(de Rojas & Ferrarotto, 2009; Leite et al., 2003). High trichome density contributes to the protection of the leaves of SLN from high summer temperatures (Perez-Estrada, et al., 2000), solar radiation (Jordan, et al.,

2005) and herbivores (Canosantana & Oyama, 1992; Levin, 1973), thus contributing to the adaptability of SLN to the hot and dry summer climate in

Australia. In this study, larger leaves tended to have lower trichome densities. A similar trend was also reported in mustard (Sinapis arvensis)

(Roy et al., 1999).

Dense trichomes on the leaf surface can also block herbicide uptake

(Brewer, et al., 1991; Chachalis, et al., 2001; Kraemer, et al., 2009). Certain adjuvants, such as modified seed oil, can help droplets penetrate the trichome barrier and adhere to the leaf surface of peppermint-scented geranium (Xu, et al., 2011). However, trichomes of SLN are much more complex than the simple trichomes of peppermint-scented geranium; therefore the relationship between trichome density, adjuvants and herbicide uptake may be different in SLN. In addition, the effects of adjuvants on overcoming the trichome barrier and herbicide uptake are highly dependent on adjuvant concentration and weed species (Wang & Liu, 2007; Xu, et al.,

2011).

Once the trichome barrier is crossed, herbicide uptake may be promoted by the high stomatal density of SLN, as guard cells are more

permeable than other epidermal cells (Ricotta & Masiunas, 1992). There were 603 stomata/mm2 on the adaxial surface and 814 stomata/mm2 on the abaxial surface of SLN (Chapter 3). The stomatal density detected here is extremely high compared to more than 800 other species (varied from 0-864 stomata/mm2, most with less than 500 stomata/mm2) reported previously

(Camargo & Marenco, 2011; Peat & Fitter, 1994; Tay & Furukawa, 2008;

Zarinkamar, 2007). In addition, tetraploid (2n = 48) and hexaploid (2n = 72)

SLN were found in Argentina (Scaldaferro, et al., 2012). Variation of ploidy level may influence the size of guard cells and impact herbicide uptake.

Therefore the high stomatal density detected here will be a positive factor for the successful control of SLN.

The morphological study supports the hypothesis that “SLN is highly morphologically diverse in Australia”. Further research is required to determine the relationship between leaf size and trichome and stomatal densities, and adjuvants and herbicides to improve recommendations for optimal herbicide efficacy.

Genetic diversity

Genetic diversity studies were undertaken after a high level of morphological variation in SLN was identified. A total of 36 SSR markers were developed, including 13 cross-species markers from other Solanum species (Chapter 4) and 23 SLN-specific markers (Chapter 5). These markers detected an average of 0.79 Jaccard’s genetic similarity between

SLN individuals in Hopetoun, Jarklin, Keith, Loxton, Narrandera,

Serpentine, Temora, Ungarie and Wirrabara (Chapters 4 and 5). Diploid (2n

93 = 24), tetraploid, and hexaploid individuals were found in Argentina

(Scaldaferro, et al., 2012). All of the SLN individuals in this study were analysed using a highly polymorphic species-specific marker EM 135

(Chapter 8). Only single locus amplifications were detected among the 670 individuals, which may suggest only diploid individuals are present in

Australia. This is in agreement with previous SLN ploidy studies in

Australia (Randell & Symon, 1976) and California (Heiser & Whitaker,

1948). Polyploidy level in angiosperms was estimated between 30%-80%

(Hegarty & Hiscock, 2008). The success of polyploids species probably due to the larger genome size, high level of plasticity of genome structure, increased heterozygosity and genetic diversity (Hegarty & Hiscock, 2008;

Leitch & Leitch, 2008; Soltis & Soltis, 1993; Van de Peer,2009; Wendel,

2000).

Nineteen high polymorphic SSR markers were further used to detect the genetic diversity among 94 SLN populations using bulk DNA analysis as a cost-efficient method (Chapter 6). The reliability of bulk DNA analysis was checked using a subset of individuals from nine locations, which achieved similar results (average Jaccard similarity: 0.73 and 0.79 for bulk and individual analysis, respectively). This method has been proven to be highly repeatable and reliable for population genetic study in other species such as maize (Zea mays) (Eschholz et al., 2008) and wild tomato (Solanum peruvianum and S. chilense) (Arunyawat et al., 2007). Data were scored in a binary form, because some of these primers (STG0010, EM117, EM127,

EM155, ESM3, SLNZ 7 and SLNZ 22) amplified multiple bands (three or

more) in a single genotype. Therefore, it is not possible to determine allele sizes as for single locus markers.

These SSR markers revealed a mean Jaccard’s genetic similarity between populations at 0.43 (Chapter 6). The sample size varies between locations according to the population size. Increasing the sample size may reduce the within and between population genetic diversity as more shared alleles may be found (Bashalkhanov et al., 2009). In addition, high genetic diversity between individuals was found using four AFLP primer combinations with a mean Jaccard’s genetic similarity at 0.26 (Chapter 7).

High levels of genetic diversity can pose challenges for effective weed control. Genetic variation may lead to phenotypic variation of weed species, contributes to adaptations and may result in herbicide resistance (Dekker,

1997). Herbicide resistance was not found in SLN, but has been reported in many other weeds, including three Solanum species: S. americanum (Chase et al., 1998), S. ptycanthum (Milliman et al., 2003) and S. nigrum

(Stankiewicz et al., 2001).

Genetic diversity also impacts on the effectiveness of biocontrol agents. For example, the stem-mining midge is a biocontrol agent of hydrilla.

Larvae survival rate was significantly higher on the fluridone susceptible genotypes (38.3%) than on the moderate (3.3%) and high (1.7%) fluridone resistant genotypes (Schmid et al., 2010). Similarly, the genetic diversity of gorse (Ulex europaeus) influences the density of its bio-control agent, gorse thrips (Sericothrips staphylinus) (Ireson et al., 2008).

95 The high genetic diversity of SLN in Australia may be caused by multiple introductions (Cuthbertson, et al., 1976), the heterogeneous nature of the initial introduction(s), and/ or inter- and intraspecific hybridisation events. SLN is a self-incompatible species. Hardin et al. (1972) suggested that the hybridisation between SLN and another two Solanum species

(western horsenettle (Solanum dimidiatum) and Carolina Horsenettle (S. carolinense) can form berries. Solanum esuriale is an Australian native species often occupying the same habitat as SLN and could potentially hybridise with SLN. These two species are morphologically similar (Bean,

2004). Both belong to the Leptostemonum subgenus, and share the same chromosome number (n = 12) (Randell & Symon, 1976). Preliminary experiments were conducted here under glasshouse conditions to determine if hybridisation occurs between SLN and S. esuriale. The results suggested that crosses between the two species can form berries. However, the viability and identity of the suspected F1 seeds need to be further investigated. Once confirmed, this inter-specific hybridisation could potentially increase the genetic diversity in SLN.

The SSR and AFLP studies identified genetic diverse SLN in

Australia. The codominant nature of high polymorphic SSR markers developed in this study allows in depth investigation on the gene flow within and between species and explains the high genetic diversity in SLN.

SLN identification

Morphological diversity increases the difficulty on differentiation between SLN and a domestic Solanum species S. esuriale, especially before anthesis. Previous studies suggested that SLN leaves are around 5-10 cm in

length, while the S. esuriale leaf is up to 5 cm long. Prickles are occasionally absent on SLN stems but rarely occur on S. esuriale stems

(Kidston, et al., 2007). However, in this study about half of the SLN plants had leaves less than 5 cm in length and around a quarter of individuals had a prickle-less stem. These plants could be misidentified as S. esuriale.

Misidentification could lead to a delay in control, resulting in replenishment of the soil seedbank and rootbank. Three cross-species SSR markers

(Chapter 8) and nine SLN specific markers (Chapter 5) amplified different sizes of bands in SLN and S. esuriale. The difference in band size can be used to differentiate between these two Solanum species (Appendix 1).

Leaf trichomes have been considered as one of the most distinguishing characters in Solanum and different trichome forms have been used as a diagnostic feature for some species such as borrachero (S. umbellatum) (Roe, 1971). Therefore, micro-morphology features were examined on the adaxial leaf surface of SLN and S. esuriale (Chapter 8).

The diameter of silverleaf nightshade trichomes varied from 437 to 849 μm and the number of lateral rays varied from eight to 16, whilst these measurements varied from 343 to 819 μm and eight to 14 in S. esuriale. No significant difference was found in trichome density, and the size and number of lateral rays of trichomes between both species. However, SLN had a significant longer (average 143 μm) trichome intrusive base structure than S. esuriale (average 33 μm), which can be used to improve the early identification. Early detection is of paramount importance for invasive weed eradication. Solanum esuriale is easier to control than SLN (Johnson, et al.,

2006). Therefore correct identification of these two species will help to

97 select appropriate control strategies. Further study may use DNA barcoding technique to investigate the relationship among SLN and other Australian native and invasive Solanum species.

Phenotypes and genotypes distribution

SLN individuals in the higher rainfall areas were significantly (p <

0.001) taller and had larger leaves, which suggests a high possibility of correlation between rainfall and morphological variation. The average leaf area of SLN reduced from 6.88 cm2 in the high rainfall area to 4.71 cm2 in the low rainfall area. By contrast, SLN samples from different rainfall areas did not have significantly different trichome density (data not shown).

Similarly, water stress does not impact on trichome density in tomato

(Wilkens et al., 1996) and mustard (Roy, et al., 1999). By contrast, reducing irrigation frequency from daily to five day intervals increased the adaxial trichome density of rose-scented geranium (Pelargonium capitatum × P. radens) from 5.3 to 15.3 trichomes/ mm2 (Eiasu et al., 2012). Eiasu et al.

(2012) suggested that such density increase is due to the reduction of leaf size, while the total trichome number on each leaf is similar. However, in this study, total leaf trichome number was not calculated, and no attempts were made to detect any possible correlation between trichome density and leaf size.

The Bayesian model-based genetic structure analysis suggested that

SLN in Australia comes from two gene pools. Two large spatial groups of these two gene pools were distributed in northern Victoria and south- western NSW, and northern NSW (Chapter 7 Fig. 5), which correlated well with the early SLN records from Bingara, NSW and Hopetoun, Victoria

(Stanton, et al., 2009a). Admixture individuals were usually found close to groups of individuals from the two gene pools, which indicates genetic admixture of these two gene pools.

Plant morphology is regulated by genetic and environmental parameters (Roy, et al., 1999). In order to understand the impact of genetic background on the morphology of SLN, individuals assigned to the two different gene pools were statistically compared on the five quantitative morphological traits (Chapter 3). None of these traits were significantly different between the genetic clusters (Appendixes 2 and 3). This suggests that environmental conditions such as rainfall may have greater impact on these morphological traits than genetic background.

These results strongly support the correlation between plant morphology and rainfall, and the correlation between genotypes and introduction events. Experiments under controlled environments are required to further prove the relationships between water availability and leaf morphology (such as leaf size, and trichome and stomatal densities) using the same genotypes. In addition, SSR study at individual level can be conducted to understand the distribution of SLN.

Reproductive strategies

Morphological and genetic studies showed that different phenotypes and genotypes of SLN are widely spread over southern Australia covering different climatic zones. The phenological study partially explained the adaptability of SLN under this Australian climate. This study (Chapter 9) showed that SLN can regenerate from seeds and roots from early spring

99 (September) to autumn (March). Regenerative stems or seedlings that emerged early in the season (September) produced larger amount of seeds and root biomass when compared to those emerged later in the season, thus providing seed and root replenishment for the following seasons. Seeds can potentially be spread through contaminated agricultural products and hybridise with other genotypes, leading to gene flow within and between populations and contribute to the SLN genetic diversity in Australia. Plants that emerged later in the season (plants from seeds later than January and plants from root later than March) produced a very limited amount of root biomass (root dry weight < 0.1g per plant). It is still not clear whether these plants produced enough root biomass to survive over winter. Furthermore, this study highlighted the importance of managing early emerged plants to reduce seed and root production. In addition, further study may need to investigate the impacts of genetic diversity and different environmental conditionals on seed production and reproduction dynamic in Australian

SLN.

Further study may focus on the impact of SLN genetic background, and climate and farming system changes on the growth, development and distribution of the weed. As a result of climate change, this region of

Australia is expected to have more floods and droughts, but also wetter and hotter summers which should favour SLN growth and spread (Australian

Bureau of Statistics, 2013). In addition, sheep numbers have dramatically decreased from 170 million in 1990 to 77 million in 2008 (Sheep CRC,

2013). Therefore, the reduction of sheep numbers may decrease the spread of SLN in the future, as sheep are one of the most important vectors for seed

100

distribution (Heap & Honan, 1993). Genotypes that are more able to adapt to the increasingly severe droughts and floods may be more competitive.

Nowadays no-till is commonly practised in Australian cropping systems of this zone; this may retard the localised spread of root fragments. In addition, our preliminary experiment suggested that new plants can also regenerate from fresh stem fragments under moist conditions (Appendix 4). Therefore, assessment of the regenerative ability of stem fragments will be useful to guide weed management such as slashing and chipping, and requires future research.

Summary and future research

Through this PhD study, high morphological and genetic variations were found in Australian SLN populations, suggesting that there are correlations between rainfall and SLN morphology, and between genotypes and introduction events. Correct identification of SLN can be achieved with the molecular markers and using SEM techniques developed in this study.

In addition, this study also highlighted the sexual and asexual reproductive strategies of SLN as another reason for its distribution.

Further research may focus on the following aspects:

1) Morphology and distribution

• Micro-morphological observation can be conducted to detect correlation between trichome and stomatal densities and leaf size, as reported in rose-scented geranium (Eiasu, et al., 2012). Glasshouse experiments under control environments are required to study the

101 correlation between rainfall and leaf size (and other morphological traits).

This will improve the understanding on morphological flexibility of SLN and explain the adaptability and distribution.

• Experiments are also required to identify appropriate application technology such as adjuvants, spraying volume to overcome the dense trichome barrier and to improve herbicide uptake through stomata.

• Studies of morphological features such as stamen length, fruit shapes and calyx lobes between SLN, S. coactiliferum and S. esuriale will be useful for distinguishing SLN from these morphologically similar species.

2) Genetic diversity and distribution

• There are around 120 native and invasive Solanum species in Australia. It is difficult to comprehensively compare the morphological features of all these species. The DNA barcoding is a fast and reliable technique (Hebert & Gregory, 2005) and will be useful for

Solanum species identification in Australia and will provide a comprehensive understanding of the relationships between these species.

• Because of the codominant nature and the ability to detect multiple alleles, the developed SSR markers allow further study on inter-species hybridisation of SLN and S. esuriale. The species-specific SSR markers developed in this study can be used to examine the possibility of natural hybridisation between SLN and S. esuriale (or other closely related

Solanum species). This study will help determine if natural hybridisation contributes to the genetic diversity in SLN.

102

• A sub-set of several highly polymorphic primer pairs can be used to study in depth analysis of gene flow within and between SLN populations. Further study of inter- and intra- species hybridisation will also help understand the mechanisms underpinning SLN genetic diversity and distribution.

• As SLN in Australia belongs to two gene pools, how these two gene pools respond to environmental conditions and management would be another area of further research. In addition, it is not clear whether

SLN of distinct genetic backgrounds differ in regenerative ability or seed production (reproductive efforts). Such studies will improve the understanding of the impact of genetic background on SLN distribution.

3) Climate and farming system change and SLN spread

Further field studies are required to understand how SLN responds to changing farming systems under different climatic conditions, thus improving the knowledge on how reproductive systems of SLN contribute to its adaptability and population dynamics. This information could be used to develop a model for SLN distribution under climate change. These studies will help reduce the impact and spread of this weed in Australia.

103 APPENDIX

Appendix 1. Species-specific SSR markers found in this study for SLN and S. esuriale identification. Band size including 19 bp of

M13-tail.

Band size (bp) Primer pairs SLN S. esuriale EM 117 - 85 EM 135 - 222 ESM 3 - 249 SLNZ 2 - 223 SLNZ 8 - 161 SLNZ 11 231 - SLNZ 12 328 - SLNZ 14 299 - SLNZ 18 185 - SLNZ 19 216 - SLNZ 20 - 220 SLNZ 21 - 285

104

Appendix 2. Relationship between genetic clusters and morphological traits of SLN*.

Genetic Clusters

Traits Cluster 1 Admix Cluster 2

Mean SE Mean SE Mean SE

Leaf Length (cm) 4.6a 0.226 4.94a 0.174 5.29a 0.235

Leaf width (cm) 1.33a 0.069 1.46a 0.07 1.45a 0.078

Leaf area (cm2) 4.77a 0.497 5.64a 0.466 6.18a 0.711

Leaf roundness 0.28a 0.011 0.27a 0.007 0.26a 0.009

Plant height (cm) 30.9a 1.821 33.2a 1.453 36.1a 1.622

* Clusters defined through the AFLP study (Chapter 7). Comparison was made using morphological data of the corresponding individuals. SE, standard error of mean. Values sharing the same letters within each row are not significantly different according to Fisher’s LSD (P = 0.05).

105 Appendix 3. Relationship between genetic clusters and adaxial trichome density of SLN*.

* Genetic cluster related to the STRUCTURE analysis in Chapter 7, TAL: low trichome density on adaxial leaf surface; TAM: medium trichome density on adaxial leaf surface and TAH: high trichome density on adaxial leaf surface.

106

Appendix 4. SLN regenerates from a 10-cm stem fragment buried at 4-cm depth under glasshouse conditions.

107 Appendix 5. Mantel test indicating no significant correlation between Jaccard’s genetic distant matrix and geographical distant matrix of the SLN samples from 94 locations in Australia.

108

REFERENCES

Abernathy, J. R., & Keeling, J. W. (1979). Silverleaf nightshade control in cotton with glyphosate. Proceedings-Southern Weed Science Society, 32, 380. Achigan-Dako, E. G., Fuchs, J., Ahanchede, A., & Blattner, F. R. (2008). Flow cytometric analysis in Lagenaria siceraria (Cucurbitaceae) indicates correlation of genome size with usage types and growing elevation. Plant Systematics and Evolution, 276(1-2), 9-19. Akkale, C., Yildirim, Z., Yildirim, M. B., Kaya, C., Ozturk, G., & Tanyolac, B. (2010). Assessing genetic diversity of some potato (Solanum tuberosum L.) genotypes grown in Turkey using the AFLP marker technique. Turkish Journal of Field Crops, 15(1), 73-78. Al-Mouemar, A., & Azmeh, F. (2009). The control of invasive species: Solanum elaeagnifolium Cav., Eichhornia crassipes, in Syria. XIIIeme Colloque International sur la Biologie des Mauvaises Herbes, Dijon, France, pp. 457-464. Anderson, G. J., Bernardello, G., & Schlehofer, M. (1999). Continuous variation among three species of Solanum sect. Anarrhichomenum (Solanaceae): the synonymy of S. carchiense and S. tetrapetalum with S. sodiroi. Kurtziana, 27: 233-242. Arunyawat, U., Stephan, W., & Stadler, T. (2007). Using multilocus sequence data to assess population structure, natural selection, and linkage disequilibrium in wild tomatoes. Molecular Biology and Evolution, 24(10), 2310-2322. Australian Weeds Committee (2012). Weeds of National Significance: silverleaf nightshade (Solanum elaeagnifolium) draft strategic plan, Canberra. Australian Weeds Committee. Weed Identification. Retrieved 25/01/2013, from http://www.weeds.org.au/cgi-bin/weedident.cgi?tpl=plant.tpl & ibra=all&card=H42. Barrett, S. C. H. (1992). Genetics of weed invasions. In Jain, S. K. & Botsford L. W. (eds.), Applied Population Biology, 67, (pp. 91-119): Springer Netherlands. Bashalkhanov, S., Pandey, M., & Rajora, O. P. (2009). A simple method for estimating genetic diversity in large populations from finite sample sizes. BMC Genetics, 10, 84. Baye, Y., & Bouhache, M. (2007). Competition between silverleaf nightshade (Solanum elaeagnifolium Cav.) and spring maize (Zea mays L.). EPPO Bulletin, 37(1), 129-131. Bean, A. R. (2004). The taxonomy and ecology of Solanum subg. Leptostemonum (Dunal) Bitter (Solanaceae) in Queensland and far north-eastern New South Wales. Austrobaileya, 6(4), 639-816. Bothma, A. (2006). Allelopathic potential of silverleaf nightshade (Solanum elaeagnifolium Cav.). MSc (Agric) Horticulture thesis, University of Pretoria, Pretoria. Retrieved from http://upetd.up.ac.za/thesis/available/etd-10072005-121704/, pp 65.

109 Boukhris-Bouhachem, S., Hulle, M., Rouze-Jouan, J., & Glais, L. K., C. (2007). Solanum elaeagnifolium, a potential source of Potato virus Y (PVY) propagation. EPPO Bulletin, 37(1), 125-128. Boyd, J. W., & Murray, D. S. (1982a). Effects of shade on silverleaf nightshade (Solanum elaeagnifolium). Weed Science, 30(3), 264- 269. Boyd, J. W., & Murray, D. S. (1982b). Growth and development of silverleaf nightshade (Solanum elaeagnifolium). Weed Science, 30(3), 238-243. Boyd, J. W., Murray, D. S., & Tyrl, R. J. (1984). Silverleaf nightshade, Solanum elaeagnifolium, origin, distribution, and relation to man. Economic Botany, 38(2), 210-217. Brewer, C. A., Smith, W. K., & Vogelmann, T. C. (1991). Functional interaction between leaf trichomes, leaf wettability and the optical properties of water droplets. Plant Cell and Environment, 14(9), 955-962. Brunel, S. (2011). Pest risk analysis for Solanum elaeagnifolium and international management measures proposed. EPPO Bulletin, 41(2), 232-242. Camargo, M. A. B., & Marenco, R. A. (2011). Density, size and distribution of stomata in 35 rainforest tree species in Central Amazonia. Acta Amazonica, 41, 205-212. Canosantana, Z., & Oyama, K. (1992). Variation in leaf trichomes and nutrients of Wigandia urens (Hydrophyllaceae) and its implications for herbivory. Oecologia, 92(3), 405-409. Carvalho, S. J. P., Damin, V., Dias, A. C. R., Melo, M. S. C., Nicolai, M., & Christoffoleti, P. J. (2009). Weed desiccation with glyphosate mixed with urea or ammonium sulfate. Planta Daninha, 27(2), 353-361. Chachalis, D., Reddy, K. N., & Elmore, C. D. (2001). Characterization of leaf surface, wax composition, and control of redvine and trumpetcreeper with glyphosate. Weed Science, 49(2), 156-163. Chalghaf, E., Aissa, M., Mellassi, H., & Mekki, M. (2007). Maîtrise de la propagation de la morelle jaune (Solanum elaeagnifolium Cav.) dans le gouvernorat de Kairouan (Tunisie). EPPO Bulletin, 37(1), 132- 136. Chiarini, F. E., & Barboza, G. E. (2009). Fruit anatomy of species of Solanum sect. Acanthophora (Solanaceae). Flora, 204(2), 146-156. Chiarini, F., & Barboza, G. E. (2007). Anatomical study of different fruit types in Argentine species of Solanum subgen. Leptostemonum (Solanaceae) Anales del Jardín Botánico de Madrid, 64(2), 165-175. Chiarini, F. E., & Barboza, G. E. (2007). Placentation patterns and seed number in fruits of South American Solanum subgen. Leptostemonum (Solanaceae) species. Darwiniana, 45(2), 163-174. Chase, C. A., Bewick, T. A., & Shilling, D. G. (1998). Differential photosynthetic electron transport and oxidative stress in paraquat- resistant and sensitive biotypes of Solanum americanum. Pesticide Biochemistry and Physiology, 60(2), 83-90. Choudhary, M., & Bordovsky, D. G. (2006). Timing of glyphosate application on control of silverleaf nightshade and glyphosate- resistant cotton yield. Weed Technology, 20(1), 198-203.

110 Christodoulakis, N. S., Lampri, P. N., & Fasseas, C. (2009). Structural and cytochemical investigation of the leaf of silverleaf nightshade (Solanum elaeagnifolium), a drought-resistant alien weed of the Greek flora. Australian Journal of Botany, 57(5), 432-438. Clausen, A. M., & Crisci, J. V. (1989). Analisis multivariado de la variacion morfologica de poblaciones diploides y tetraploides de Solanum gourlayi (Solanaceae) y species relacionadas. Darwiniana, 29: 247-259. Cooley, A. W., & Smith, D. T. (1972). Seed aspects of two perennials - woolly leaf bursage and silverleaf nightshade. Proceedings 25th Annual Meeting Southern Weed Science Society, pp. 443. Cooley, A. W., & Smith, D. T. (1973). Silverleaf nightshade (whiteweed) establishment from seed and clipped seedlings, PR-3198, Weed and Herbicide Research in West Texas, Texas, pp. 6-9. Cosa, M. T., Bruno, G., & Dottori, N. (1998). Anatomía de los órganos vegetativos en Solanum juvenale y su comparación con S. elaeagnifolium (Solanaceae). Anales del Instituto de Biología, Serie Botánica, 69(1), 9-22. Cosa, M. T., Dottori, N., & Bruno, G., (2000). Propagación y anatomía de los órganos vegetativos en Solanum hieronymi (Solanaceae). Kurtziana 28(2): 211-220. Cuthbertson, E. G., Leys, A. R., & McMaster, G. (1976). Silverleaf nightshade - a potential threat to agriculture. Agricultural Gazette of New South Wales, 87(6), 11-13. Davis, C. H., Smith, T. J., & Hawkins, R. S. (1945). Eradication of the white horsenettle in southern Arizona (Bulletin no. 195). Tucson (US): University of Arizona, pp. 1-14. de Rojas, C. E. B., & Ferrarotto, M. (2009). Morphology of foliar epidermis in two groups of Solanum section Geminata (Solanaceae). Caldasia, 31(1), 31-40. Dekker, J. (1997). Weed diversity and weed management. Weed Science, 45(3), 357-363. Diggle, P. K. (1993). Developmental plasticity, genetic variation, and the evolution of andromonoecy in Solanum hirtum (Solanaceae). American Journal of Botany, 80(8), 967-973. Dollahite, J. W., & Allen, T. J. (1960). Silverleaf nightshade poisoning in livestock, Progress Report no. 2146, Vernon, Texas (US). Edmonds, J. M. (1978). Numerical taxonomic studies on Solanum L. section Solanum (Maurella). Botanical Journal of the Linnean Society, 76(1), 27-51. Eiasu, B. K., Steyn, J. M., & Soundy, P. (2012). Physiomorphological response of rose-scented geranium (Pelargonium spp.) to irrigation frequency. South African Journal of Botany, 78, 96-103. Ellstrand, N. C., & Schierenbeck, K. A. (2000). Hybridization as a stimulus for the evolution of invasiveness in plants? Proceedings of the National Academy of Sciences of the United States of America, 97(13), 7043-7050. Encomidou, E., & Yannitsaros, A. (1975). Research on the weed flora of Greece. III. Morphology, development and phenology of Solanum elaeagnifolium Cav. Candollea, 30, 29-41.

111 Ensbey, R., Cook, T., Scott, M., Johnson, S., & van Oosterhout, E. (Eds.). (2011). Noxious and environmental weed control handbook (5th ed.). Orange: Department of Primary Industries, pp. 88. Eschholz, T. W., Peter, R., Stamp, P., & Hund, A. (2008). Genetic diversity of Swiss maize (Zea mays L. ssp mays) assessed with individuals and bulks on agarose gels. Genetic Resources and Crop Evolution, 55(7), 971-983. Esquivel, J. F., & Esquivel, S. V. (2009). Identification of cotton fleahopper (Hemiptera: Miridae) host plants in central texas and compendium of reported hosts in the United States. Environmental Entomology, 38(3), 766-780. Fahn, A., & Cutler, D. F. (1992). Xerophytes (Vol. 13). Stuttgart: Gebruder Borntraeger. Fernandez, O. A., & Brevedan, R. E. (1972). Regeneracion de Solanum eleagnifolium Cav. a partir de fragmentos de sus raices. Darwiniana, 17, 434-442. Feuerherdt, L. (2009). Overcoming a deep rooted perennial problem - silverleaf nightshade (Solanum elaeagnifolium) in South Australia. Plant Protection Quarterly, 24(3), 123-124. Field, R. P., Kwong, R. M., & Sagliocco, J. L. (2009). Host specificity of Ditylenchus phyllobius, a potential biological control agent of silverleaf nightshade (Solanum elaeagnifolium Cav.) in Australia. Plant Protection Quarterly, 24(4), 141-145. Fukuoka, H., Yamaguchi, H., Nunome, T., Negoro, S., Miyatake, K., & Ohyama, A. (2010). Accumulation, functional annotation, and comparative analysis of expressed sequence tags in eggplant (Solanum melongena L.), the third pole of the genus Solanum species after tomato and potato. Gene, 450(1-2), 76-84. Giannattasio, R. B., & Spooner, D. M. (1994). A reexamination of species boundaries between. Solanum megistacrolobum and S. toralapanum. (Solanum sect. Petota, series Megistacroloba): Morphological data. Systematic Botany, 19(1), 89-105. Green, J. D., Murray, D. S., & Stone, J. F. (1988). Soil-water relations of silverleaf nightshade (Solanum-elaeagnifolium) with cotton (Gossypium-hirsutum). Weed Science, 36(6), 740-746. Green, J. M., Barker, J. H. A., Marshall, E. J. P., Froud-Williams, R. J., Peters, N. C. B., Arnold, G. M., . . . Karp, A. (2001). Microsatellite analysis of the inbreeding grass weed Barren Brome (Anisantha sterilis) reveals genetic diversity at the within- and between-farm scales. Molecular Ecology, 10(4), 1035-1045. Greenfield, K. M. (2003). Understanding herbicide behaviour in Solanum elaeagnifolium Cav. Honours thesis, University of Adelaide. pp 77. Hackett, N. M., & Murray, D. S. (1982). Interference of silverleaf nightshade (Solanum eleagnifolium) with Spanish peanuts (Arachis hypogaea). Proceedings 35th Annual Meeting Southern Weed Science Society, pp. 66. Hall, D. G., & Teetes, G. L. (1981). Alternate host plants of sorghum panicle-feeding bugs in southeast central Texas. Southwestern Entomologist, 6(3), 220-228.

112 Hardin, J. W., Doerkson, G., Herndon, D., Hobson, M., & Thomas., F. (1972). Pollination and floral biology of four weedy species in southern Oklahoma. Southwest Naturalist, 16, 403-412. Hawker, V., Preston, C., & Baker, J. (2006, 24-28 September 2006). Genetic variation within and among silverleaf nightshade (Solanum elaeagnifolium Cav.) populations in South Australia. 15th Australian Weeds Conferences., Adelaide Conference Centre, Adelaide, South Australia, pp. 176. Heap, J. W., & Honan, I. (1993). Weed seed excretion by sheep - temporal patterns and germinability. Proceedings I of the 10th Australian Weeds Conference and 14th Asian Pacific Weed Science Society Conference, Brisbane, Australia, pp. 431-434. Heap, J. W., Smith, E., & Honan, I. (1997). Silverleaf nightshade: a technical handbook for animal and plant control boards in South Australia. Adelaide: Primary Industries South Australia Animal and Plant Control Commission, pp. 42. Hebert, P. D. N., & Gregory, T. R. (2005). The promise of DNA barcoding for taxonomy. Systematic Biology, 54(5), 852-859. Hegarty, M. J., & Hiscock, S. J. (2008). Genomic clues to the evolutionary success of review polyploid plants. Current Biology, 18(10), R435- R444. Heiser, C. B., Jr., & Whitaker, T. W. (1948). Chromosome number, polyploidy, and growth habit in California weeds. American Journal of Botany, 35(3), 179-186. Hoffmann, J. H., Moran, V. C., & Impson, F. A. C. (1998). Promising results from the first biological control programme against a Solanaceous weed (Solanum elaeagnifolium). Agriculture Ecosystems & Environment, 70(2-3), 145-150. Hosking, J. R., Sainty, G. R., & Jacobs, S. W. L. (2000). Certainty and Uncertaintly in Plant Identification. The New Mexico Botanist (15), 1-4. Huangfu, C. H., Song, X. L., & Qiang, S. (2009). Morphological disparities in the epidermal and anatomical features of the leaf among wild Brassica juncea populations. Weed Biology and Management, 9(3), 234-242. Idol, G. B., & Slosser, J. E. (2005). Spring emergence and development of the cotton fleahopper (Heteroptera : Miridae) in host plants of the Texas rolling plains. Southwestern Entomologist, 30(2), 73-84. Ireson, J. E., Holloway, R. J., & Chatterton, W. S. (2008). The influence of host plant genotype on variation in population densities of the gorse thrips, Sericothrips staphylinus (Thysanoptera: Thripidae), and its consideration in relation to release strategies. Biocontrol Science and Technology, 18(9), 949-955. Johnson, A., Brooke, G., Thompson, R., Roberts, K., Hertel, K., Border, N., . . . Sullivan, P. (2006). Weed control for cropping and pasture in central west NSW. Retrieved 28 May 2011, from http://www.dpi.nsw.gov.au/agriculture/pests- weeds/weeds/publications/central-west Jones, C. J., Edwards, K. J., Castaglione, S., Winfield, M. O., Sala, F., vandeWiel, C., . . . Karp, A. (1997). Reproducibility testing of

113 RAPD, AFLP and SSR markers in plants by a network of European laboratories. Molecular Breeding, 3(5), 381-390. Jordan, G. J., Dillon, R. A., & Weston, P. H. (2005). Solar radiation as a factor in the evolution of scleromorphic leaf anatomy in Proteaceae. American Journal of Botany, 92(5), 789-796. Kardolus, J. P. (1999). Morphological variation within series Acaulia Juz. (Solanum sect. Petota). In: Nee, M., Symon, D. E., Lester, R. N., Jessop, J. P., eds. Solanaceae IV. Advances in Biology and Utilization. Kew: Royal Botanic Gardens, 257-274. Keeling, J. W., & Abernathy, J. R. (1985). Biological control of silverleaf nightshade with nematodes. Proceedings, Southern Weed Science Society, 38th annual meeting, pp. 362. Kelly, P., & Reeder, R. (2002). Geared up: Conservation farming equipment for central west NSW: NSW Department of Primary Industries, pp. 56. Kidston, J., Thompson, R., & Johnson, A. (2007). Silverleaf nightshade. Primefact 237 Retrieved 10 June 2010, from http://www.dpi.nsw.gov.au/__data/assets/pdf_file/0008/90890/silver leaf-nightshade.pdf. Kraemer, T., Hunsche, M., & Noga, G. (2009). Surfactant-induced deposit structures in relation to the biological efficacy of glyphosate on easy- and difficult-to-wet weed species. Pest Management Science, 65(8), 844-850. Kwong, R. (2006). Feasibility of biological control of solanaceous weeds of temperate Australia, Weed 120, MLA, North Sydney, 150. Kwong, R. M., Sagliocco, J. L., Weiss, J., Hunt, T., & Morfe, T. (2008). Feasibility of biological control of silverleaf nightshade in Australia. 16th Australian Weeds Conference Proceedings: weed management 2008 hot topics in the tropics, North Queensland, Australia, pp. 532. Leaper, C., & Holloway, P. J. (2000). Adjuvants and glyphosate activity. Pest Management Science, 56(4), 313-319. Leitch, A. R., & Leitch, I. J. (2008). Perspective - Genomic plasticity and the diversity of polyploid plants. Science, 320(5875), 481-483. Leite, G. L. D., Picanco, M., Zanuncio, J. C., & Marquini, F. (2003). Factors affecting mite herbivory on eggplants in Brazil. Experimental and Applied Acarology, 31(3-4), 243-252. Lemerle, D., & Leys, A. R. (1991). Control of silverleaf nightshade (Solanum-elaeagnifolium) increases the grain-yield of wheat. Australian Journal of Experimental Agriculture, 31(2), 233-236. Levin, D. A. (1973). The role of trichomes in plant defense. The Quarterly Review of Biology, 48(1), 3-15. Levin, R. A., Myers, N. R., & Bohs, L. (2006). Phylogenetic relationships among the "spiny solanums" (Solanum subgenus Leptostemonum, Solanaceae). American Journal of Botany, 93(1), 157-169. Marshall, R., & Moss, S. R. (2008). Characterisation and molecular basis of ALS inhibitor resistance in the grass weed Alopecurus myosuroides. Weed Research, 48(5), 439-447. Martine, C. T., Vanderpool, D., Anderson, G. J., & Les, D. H. (2006). Phylogenetic relationships of andromonoecious and dioecious Australian species of Solanum subgenus Leptostemonum section

114 Melongena: Inferences from ITS sequence data. Systematic Botany, 31(2), 410-420. Marulanda, M. L., Lopez, A. M., & Aguilar, S. B. (2007). Genetic diversity of wild and cultivated Rubus species in Colombia using AFLP and SSR markers. Crop Breeding and Applied Biotechnology, 7(3), 242- 252. McGregor, C. E., Lambert, C. A., Greyling, M. M., Louw, J. H., & Warnich, L. (2000). A comparative assessment of DNA fingerprinting techniques (RAPD, ISSR, AFLP and SSR) in tetraploid potato (Solanum tuberosum L.) germplasm. Euphytica, 113(2), 135-144. McKenzie, D. N. (1975). Silverleaf nightshade - one method of dispersal. Australian weeds research newsletter, 22, 13-15. McKenzie, D. N. (1980). Report on silver-leaf nightshade research, Pamphlet No.79, Department of Crown Lands and Survey, Victoria. McKenzie, D. N., & Douglas, G. W. (1974). Report on white horsenettle (or silver-leaf nightshade) Solanum elaeagnifolium Cav., Pamphlet No. 50, Vermin and Noxious Weeds Destruction Board, Department of Crown Lands and Survey. Victoria. Mekki, M. (2007). Biology, distribution and impacts of silverleaf nightshade (Solanum elaeagnifolium Cav.). EPPO Bulletin, 37(1), 114-118. Milliman, L. D., Riechers, D. E., Wax, L. A., & Simmons, F. W. (2003). Characterization of two biotypes of imidazolinone-resistant eastern black nightshade (Solanum ptycanthum). Weed Science, 51(2), 139- 144. Molnar, V. M., & McKenzie, D. N. (1976). Progress report on silverleaf nightshade, Pamphlet No. 61, Vermin and Noxious Weeds Destruction Board, Keith Turnbull Research Institute, Victoria. Monaghan, N., & Brownlee, H. (1979). Regeneration of Solanum karsensis and S. elaeagnifolium. 7th Asian-Pacific Weed Science Society Conference, Sydney, Australia, pp. 321-323. Mkula, N. P. (2006). Allelopathic interference of silverleaf nightshade (Solanum elaeagnifolium Cav.) with the early growth of cotton (Gossypium hirsutum L.) (PhD PhD thesis), University of Pretoria. Nee, M. (1999). Synopsis of Solanum in the New World. in: Nee, M., Symon, D, E., Lester, R. N., Jessop, J. P., (eds), Solanaceae IV. Advances in biology and Utilization. Kew: Royal Botanic Gardens, pp. 285-333. Nei, M. (1972). Genetic distance between populations. American Naturalist, 106(949), 283-292. Nissen, S. J., Masters, R. A., Lee, D. J., & Rowe, M. L. (1995). DNA-based marker systems to determine genetic diversity of weedy species and their application to biocontrol. Weed Science, 43(3), 504-513. O'Hanlon, P. C., Peakall, R., & Briese, D. T. (2000). A review of new PCR- based genetic markers and their utility to weed ecology. Weed Research, 40(3), 239-254. Okada, M., Lyle, M., & Jasieniuk, M. (2009). Inferring the introduction history of the invasive apomictic grass Cortaderia jubata using microsatellite markers. Diversity and Distributions, 15(1), 148-157.

115 Olckers, T., Hoffmann, J. H., Moran, V. C., Impson, F. A. C., & Hill, M. P. (1999). The initiation of biological control programmes against Solanum elaeagnifolium Cavanilles and S-sisymbriifolium Lamarck (Solanaceae) in South Africa. In T. Olckers & M. P. Hill (Eds.), Biological Control of Weeds in South Africa (1990-1998). African Entomology Memoir 1, pp. 55-63. Olckers, T., & Hulley, P. E. (1994). Resolving ambiguous results of host- specificity tests - the case of 2 Leptinotarsa species (Coleoptera, Chrysomelidae) for biological-control of Solanum elaeagnifolium Cavanilles (Solanaceae) in South-Africa. African Entomology, 2(2), 137-144. Olckers, T., & Zimmermann, H. G. (1991). Biological-control of silverleaf nightshade, Solanum elaeagnifolium, and bugweed, Solanum mauritianum, (Solanaceae) in South-Africa. Agriculture Ecosystems & Environment, 37(1-3), 137-155. Olckers, T., Zimmermann, H. G., & Hoffmann, J. H. (1995). Interpreting ambiguous results of host-specificity tests in biological-control of weeds - assessment of 2 Leptinotarsa species (chrysomelidae) for the control of Solanum elaeagnifolium (Solanaceae) in South-Africa. Biological Control, 5(3), 336-344. Orr, C. C. (1981). Nothanguina phyllobia, a nematode biocontrol of silverleaf nightshade. Proceedings of the 5th International Symposium on Biological Control of Weeds pp. 389-391. Parsons, W. T., & Cuthbertson, E. G. (2001). Noxious weeds of Australia (2nd ed.). Collingwood, VIC: CSIRO Publishing, pp. 609-612. Peat, H. J., & Fitter, A. H. (1994). A comparative study of the distribution and density of stomata in the British flora. Biological Journal of the Linnean Society, 52(4), 377-393. Perez-Estrada, L. B., Cano-Santana, Z., & Oyama, K. (2000). Variation in leaf trichomes of Wigandia urens: environmental factors and physiological consequences. Tree Physiology, 20(9), 629-632. Petanidou, T., Godfree, R. C., Song, D. S., Kantsa, A., Dupont, Y. L., & Waser, N. M. (2012). Self-compatibility and plant invasiveness: Comparing species in native and invasive ranges. Perspectives in Plant Ecology Evolution and Systematics, 14(1), 3-12. Prentis, P. J., Wilson, J. R., Dormontt, E. E., Richardson, D. M., & Lowe, A. J. (2008). Adaptive evolution in invasive species. Trends Plant Sci, 13(6), 288-294. Porch, N., Jordan, G. J., Price, D. M., Barnes, R. W., Macphail, M. K., & Pemberton, M. (2009). Last interglacial climates of south-eastern Australia: plant and beetle-based reconstructions from Yarra Creek, King Island, Tasmania. Quaternary Science Reviews, 28(27-28), 3197-3210. Randell, B. R., & Symon, D. E. (1976). Chromosome numbers in Australian Solanum species. Australian Journal of Botany, 24, 369-379. Richburg, J. S., Wilcut, J. W., & Wehtje, G. R. (1994). Toxicity of AC 263,222 to purple (Cyperus rotundus) and yellow nutsedge (C. esculentus). Weed Science, 42(3), 398-402. Ricotta, J. A., & Masiunas, J. B. (1992). Relationship of leaf surface characteristics to acifluorfen tolerance in tomato (Lycopersicon esculentum) and related species. Weed Science, 40(3), 402-407.

116 Rodriguez-Alvarado, G., Fernandez-Pavia, S., Creamer, R., & Liddell, C. (2002). Pepper mottle virus causing disease in chile peppers in southern New Mexico. Plant Disease, 86(6), 603-605. Rodriguez, N. M. (1972). Control of perennial weeds in the semi-arid region of the pampas. Malezas y su control, 1(4), 52-56. Roe, K. E. (1971). Terminology of hairs in the genus Solanum. Taxon, 20(4), 501-508. Roy, B. A., Stanton, M. L., & Eppley, S. M. (1999). Effects of environmental stress on leaf hair density and consequences for selection. Journal of Evolutionary Biology, 12(6), 1089-1103. Rutherford, P. A. (1978). Effect of time of immersion in running and still water on the germination of silver-leaf nightshade. Proceedings of the First Conference of the Council of Australian Weed Science Societies, National Science Centre, Parkville, Victoria, Australia, pp. 372-378. Scaldaferro, M., Chiarini, F., Santiñaque, F., Bernardello, G., & Moscone, E. (2012). Geographical pattern and ploidy levels of the weed Solanum elaeagnifolium (Solanaceae) from Argentina. Genetic Resources and Crop Evolution, 59(8), 1833-1847. Schmid, T. A., Cuda, J. P., Macdonald, G. E., & Gillmore, J. L. (2010). Performance of two established biological control agents on hydrilla genotypes susceptible and resistant to fluridone herbicide. Journal of Aquatic Plant Management, 48, 102-105. Seithe, A. (1979). Hair types as taxonomic characters in Solanum. In Hawkers, J. G., Lester, R. N., & Skelding, A. J., (Eds.), The biology and taxonomy of the Solanaceae. UK: Academic Press, pp. 307-319. Shirasawa, K., Asamizu, E., Fukuoka, H., Ohyama, A., Sato, S., Nakamura, Y., . . . Isobe, S. (2010). An interspecific linkage map of SSR and intronic polymorphism markers in tomato. Theoretical and Applied Genetics, 121(4), 731-739. Smarda, P., & Bures, P. (2006). Intraspecific DNA content variability in Festuca pallens on different geographical scales and ploidy levels. Annals of Botany, 98(3), 665-678. Smith, B. S., Pawlak, J. A., Murray, D. S., Verhalen, L. M., & Green, J. D. (1990). Interference from established sands of silverleaf nightshade (Solanum elaeagnifolium) on cotton (Gossypium hirsutum) lint yield. Weed Science, 38(2), 129-133. Soltis, D. E., & Soltis, P. S. (1993). Molecular data and the dynamic nature of polyploidy. Critical Reviews in Plant Sciences, 12(3), 243-273. Spooner, D. M., & van den berg, R. G. 2001. Quantitative assessment of corolla shape variation in Mexican Solanum section Petota. In: van den berg, R. G. , Barendse, G. W. M., van den Weerden, G. M., Mariani, C., eds. Solanaceae V. Advances in taxonomy and utilization. Nijmegen: Nijmegen University Press, 61-71. Stankiewicz, M., Gadamski, G., & Gawronski, S. W. (2001). Genetic variation and phylogenetic relationships of triazine-resistant and triazine-susceptible biotypes of Solanum nigrum - analysis using RAPD markers. Weed Research, 41(4), 287-300.

117 Stanton, R., Heap, J. W., Carter, R. J., & Wu, H. (2009a). Solanum elaeagnifolium. In F. D. Panetta (Ed.), The biology of Australian weeds (Vol. 3, pp. 274-293). Melbourne: R. G. and F. J. Richardson. Stanton, R., Piltz, J., Rodham, C., & Wu, H. (2012a). Silage for managing weed seeds. 18th Australasian Weeds Conference, Melbourne, pp. 219-221. Stanton, R., Wu, H., & Lemerle, D. (2009b). Innovative management of silverleaf nightshade (SLN) and prairie ground cherry (PGC) (Stage 1), Weeds 135, North Sydney, 82. Stanton, R., Wu, H., & Lemerle, D. (2011). Root regenerative ability of silverleaf nightshade (Solanum elaeagnifolium Cav.) in the glasshouse. Plant Protection Quarterly, 26(2), 54-56. Stanton, R., Wu, H. W., & Lemerle, D. (2012b). Factors affecting silverleaf nightshade (Solanum elaeagnifolium) germination. Weed Science, 60(1), 42-47. Steptoe, P. J., Vencill, W. K., & Grey, T. L. (2006). Influence of moisture stress on herbicidal control of an invasive weed, Benghal dayflower (Commelina benghalensis). Journal of Plant Diseases and Protection, 907-914. Symon, D. E. (1981). A revision of the genus Solanum in Australia. Journal of the Adelaide Botanic Gardens, 4, 1-367. Symon, D. E. (1986). A survey of Solanum prickles and marsupial herbivory in Australia. Annals of the Missouri Botanical Garden, 73(4), 745-754. Taleb, A., Bouhache, M., Ameur, A., & Baye, Y. (2007). Etat actuel de l’infestation par la morelle jaune (Solanum elaeagnifolium Cav. ) au Maroc. EPPO Bulletin, 37(1), 119-124. Tay, A. C., & Furukawa, A. (2008). Variations in leaf stomatal density and distribution of 53 vine species in Japan. Plant Species Biology, 23(1), 2-8. Tejada, M. L. O., & Reyes, B. E. (1986). Bioecological study, alternative hosts and mass rearing of the pepper weevil Anthonomus eugenii Cano. informe de Investigacion, Division de Ciencias Agropecuarias y Maritimas, Instituto Tecnologico y de Estudios Superiores de Monterrey, 19, 60-61. Thorne, G. (1934). Some plant-parasitic nemas, with descriptions of three new species. Journal of Agricultural Research, 49(8), 755-763. Tideman, A. F. (1960). Tomato weed in South Australia (Solanum elaeagnifolium Cav.). Journal of the Department of Agriculture of South Australia, 63(8), 329-331. Timbal, B., & Jones, D. A. (2008). Future projections of winter rainfall in southeast Australia using a statistical downscaling technique. Climatic Change, 86(1-2), 165-187. Travlos, I. (2013). Responses of invasive silverleaf nightshade (Solanum elaeagnifolium) populations to varying soil water availability. Phytoparasitica, 41(1), 41-48. Trione, S. O., & Cony, M. A. (1990). Thermoperiodism and other physiological traits of Solanum elaeagnifolium seeds in relation to germination. Seed Science and Technology, 18(3), 525-539.

118 Tscheulin, T., Petanidou, T., & Settele, J. (2009). Invasive weed facilitates incidence of Colorado potato beetle on potato crop. International Journal of Pest Management, 55(2), 165-173. University of Georgia. Discover life. Retrieved 25/01/2013, from http://pick4.pick.uga.edu/mp/20m?act=make_map&w=720&r= 0.2&e=0.00000&n=0.00000&z=0&kind=Solanum+elaeagnifolium& la=0&lo=0 Van Buskirk, J., & Willi, Y. (2006). The change in quantitative genetic variation with inbreeding. Evolution, 60(12), 2428-2434. Van de Peer, Y., Maere, S., & Meyer, A. (2009). OPINION The evolutionary significance of ancient genome duplications. Nature Reviews Genetics, 10(10), 725-732. Van den Berg, R. G., & Groendijk-Wilders, N. 1999. Numerical analysis of the taxa of series Circaefolia (Solanum sect. Petota). In: Nee M, Symon DE, Lester RN, Jessop JP, eds. Solanaceae IV. Advances in biology and Utilization. Kew: Royal Botanic Gardens: 213-226. van Heerwaarden, B., Willi, Y., Kristensen, T. N., & Hoffmann, A. A. (2008). Population bottlenecks increase additive genetic variance but do not break a selection limit in rain forest Drosophila. Genetics, 179(4), 2135-2146. Viljoen, B. D., & Wassermann, V. D. (2004). Suppression of silver-leaf bitter apple (Solanum elaeagnifolium Cav.) by cultivated pasture crops under dry-land conditions: a preliminary study. South African Journal of Plant and Soil, 21(1), 63-66. Wakhloo, J. L. (1970). Role of mineral nutrients and growth regulators in the apical dominance in Solanum sisymbrifolium. Planta, 91(3), 190- 194. Wakhloo, J. L. (1975). Studies on growth, flowering, and production of female sterile flowers as affected by different levels of foliar potassium in Solanum sisymbrlifolium Lam. III. Interaction between foliar potassium and applied daminozide, chlormequat chloride, and chlorflurecolpmethyl. Journal of Experimental Botany, 26(92), 441- 450. Wang, C. J., & Liu, Z. Q. (2007). Foliar uptake of pesticides - Present status and future challenge. Pesticide Biochemistry and Physiology, 87(1), 1-8. Wapshere, A. J. (1988). Prospects for the biological-control of silver-leaf nightshade, Solanum elaeagnifolium, in Australia. Australian Journal of Agricultural Research, 39(2), 187-197. Wassermann, V. D., Zimmermann, H. G., & Neser, S. (Eds.). (1988). The weed silverleaf bitter apple ("Satansbos") (Solanum elaeagnifolium Cav.): with special reference to its status in South Africa. Pretoria: Plant Protection Research Institute. Waycott, M., Jones, B. L., Van Dijk, J. K., Robson, H. L. A., & Calladine, A. (2011). Microsatellite markers in the Australian desert plant, Solanum centrale (Solanaceae). American Journal of Botany, 98(4), E81-E83. Wen, A., Mallik, I., Alvarado, V. Y., Pasche, J. S., Wang, X., Li, W., . . . Gudmestad, N. C. (2009). Detection, distribution, and genetic

119 variability of 'Candidatus liberibacter' species associated with zebra complex disease of potato in North America. Plant Disease, 93(11), 1102-1115. Wendel, J. F. (2000). Genome evolution in polyploids. Plant Molecular Biology, 42(1), 225-249. Whalen, M. D. (1984). Conspectus of species groups in Solanum subgenus Leptostemonum. Gentes Herbarum, 12: 179-282. Wilkens, R. T., Shea, G. O., Halbreich, S., & Stamp, N. E. (1996). Resource availability and the trichome defenses of tomato plants. Oecologia, 106(2), 181-191. Xu, L. Y., Zhu, H. P., Ozkan, H. E., Bagley, W. E., & Krause, C. R. (2011). Droplet evaporation and spread on waxy and hairy leaves associated with type and concentration of adjuvants. Pest Management Science, 67(7), 842-851. Yu, Q., Han, H. P., & Powles, S. B. (2008). Mutations of the ALS gene endowing resistance to ALS-inhibiting herbicides in Lolium rigidum populations. Pest Management Science, 64(12), 1229-1236. Zarinkamar, F. (2007). Stomatal observations in dicotyledons. Pakistan Journal of Biological Sciences, 10(2), 199-219.

120