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.
x
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
2
(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.
7
Chapter 10 General discussion and conclusion:
This chapter discusses the outcome of this PhD study and aspects of further research.
8
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).
10
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
14
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” sub- genus 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).
18
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
20
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
22
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
24
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).
26
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,
28
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 signifi- cantly 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
© 2013 European Weed Research Society 33 Morphological variation of Solanum elaeagnifolium 3
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
© 2013 European Weed Research Society 34 4 X C Zhu et al.
AB
CD
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 sur- face; and (D): stomata on shaved abaxial leaf surface.
Fig. 3 Variation of Solanum elaeagnifoli- um leaf size and shape. All leaves were the seventh leaf from the shoot apex of flowering plants, showing the adaxial sur- face 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).
© 2013 European Weed Research Society 35 Morphological variation of Solanum elaeagnifolium 5
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).
© 2013 European Weed Research Society 36 6 X C Zhu et al.
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 coef- ficient 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.
© 2013 European Weed Research Society 37 Morphological variation of Solanum elaeagnifolium 7
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 tri- chome 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 differ- ence may be associated with genetic diversity of this 0 weed (Zhu et al., 2012) or edaphic and climatic differ- ence between locations. Individuals collected from –2 higher rainfall areas are often taller and have larger leaves, indicating a possible correlation between pheno- typic 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 establish- ment 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 varia- tion 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.
© 2013 European Weed Research Society 39 Morphological variation of Solanum elaeagnifolium 9
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.
© 2013 European Weed Research Society 40 10 X C Zhu et al.
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
© 2013 European Weed Research Society 41 Morphological variation of Solanum elaeagnifolium 11
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
© 2013 European Weed Research Society 42 CHAPTER 4 PUBLISHED PAPER
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]