variation of the germinable soil seed banks in a pasture infested by parthenium hysterophorus l.at kilcoy, south-eastern queensland

The disproportional low percentage of highly desirable pasture species including legume, palatable perennial grass and other native species coupled with the presence of parthenium weed i

VARIATION OF THE GERMINABLE SOIL SEED BANKS

IN A PASTURE INFESTED BY

PARTHENIUM HYSTEROPHORUS L

AT KILCOY, SOUTH-EASTERN QUEENSLAND

An Thi Kieu Do

A report submitted as a requirement for the degree of Master of Agricultural

Studies in The University of Queensland

School of Land, Crop and Food Sciences Faculty of Natural Resources, Agriculture and Veterinary Science

The University of Queensland, Gatton

DECLARATION

The work reported in this research project report has been carried out by the undersigned Where reference has been made to the results of other workers, appropriate acknowledgment of the source of information has been made

Author

AN DO

This is to certify that the academic style and manner of presentation of the research project report are appropriate to the discipline, that all requirements of the University in relation to the deposition of records of research have been met and that the information contained in the report is a true representation of the data collected

Project Supervisors

Prof STEVE ADKINS Dr CHRIS O’DONNELL Dr DOUG GEORGE

ACKOWLEDGEMENTS

I would sincerely like to thank Professor Steve Adkins (principal supervisor), Dr Chris O’Donnell (associate supervisor) and Dr Doug George (associate supervisor) - who offered endless hours of patience, support and guidance - for their valuable roles in discussing ideas and problems associated with the project and for reading many drafts Special thanks

to Dr Sheldon Navie who played a great role in identification botanical species and to Mr Allan Lisle who assisted me with data analysis Many thanks also to all of the staff of the School of Land, Crop and Food Sciences - The University of Queensland - who supported

me with equipment and materials The land owner, Mr Syd Smith, was kind enough to provide me with the property-related information and to let me set up sites on his property Thanks to Mr Shane Noon, a Kilcoy officer, for helping locating the study sites

Special thanks to the UQ’s fellows, who accompanied with me on field trips and who made my life in Australia pleasurable

I greatly appreciate the financial support of the Australian Development Scholarship for my study at the University of Queensland

Finally, I gratefully acknowledge the support of my family members, who cared and encouraged me during the project

ABSTRACT

Seed germination method was used to determine the germinable soil seed bank at two contrasting sites (one in a gully and another at the top of a ridge) within a sub-tropical pasture at Kilcoy, south-eastern Queensland, in January 2008 (mid-summer) and July 2008 (mid-winter) These two sites demonstrated differences in topography, soil type and moisture content and standing vegetation composition During the period of study, the germinable soil seed bank at the two sites varied between 11,526 to 23,252 seeds/m2 in the gully and between 10,136 to 12,799 seeds/m2 at the top of the ridge Parthenium weed

(Parthenium hysterophorous L.) exhibited a moderately abundant and persistent seed bank

in the gully (varying from 358 to 821 seeds/m2) and at the top of the ridge (from 863 to 1,189 seeds/m2) The weed was not the dominant species at either site, representing 1.5 to 7.1% of the total germinable seed bank in the gully and 8.5 to 9.3% of the total seed banks

at the top of the ridge The species richness and the species diversity of the germinable seed bank were high at both sites with no significant difference showing up between them or between summer and winter This management plan, designed around chemical control, had not yet been in operation long enough to have the the desired effect of reducing the seed bank to below a desirable maximum level The disproportional low percentage of highly desirable pasture species (including legume, palatable perennial grass and other native species) coupled with the presence of parthenium weed in the germinable soil seed bank indicates that the pasture is still in a degraded state

TABLE OF CONTENTS

DECLARATION i

ACKOWLEDGEMENTS ii

ABSTRACT iii

TABLE OF CONTENTS iii

LIST OF TABLES vii

LIST OF FIGURES viii

LIST OF PLATES ix

CHAPTER 1 1

INTRODUCTION 1

CHAPTER 2 3

LITERATURE REVIEW 3

2.1 Scope of the review 3

2.2 Parthenium weed (Parthenium hysterophorus L.) 3

2.2.1 Taxonomy and description of parthenium weed 3

2.2.2 The history of spread and present distribution 5

2.2.3 Habitat 7

2.2.4 Growth and development 8

2.2.5 Biology of parthenium weed seed 9

2.2.6 Parthenium weed and its significance 12

2.2.7 Parthenium weed management 16

2.3 Soil seed bank 23

2.3.1 Soil seed bank concept 23

2.3.2 Soil seed bank study approaches 27

2.3.3 Composition of the soil seed bank in grasslands 31

2.3.5 Soil seed bank of parthenium weed 35

2.4 Kilcoy climate/land use of the study site 36

2.4.1 Kilcoy location 36

2.4.2 Kilcoy climate 36

2.4.3 Land use of the study site 38

2.5 Conclusion 38

CHAPTER 3 40

MATERIALS AND METHODS 40

3.1 Study sites 40

3.2 Sampling method 40

3.3 Analytical method 41

3.4 Statistical analysis 43

CHAPTER 4 44

RESULTS 44

4.1 Temporal variation in seed density and composition 44

4.2 Species richness and diversity 53

4.3 Life history 54

4.3.1 Annuals/biennials versus perennials 54

4.3.2 Life form 56

4.3.3 Introduced species versus native species 57

4.4 Germination rate 57

CHAPTER 5 61

DISCUSSION 61

5.1 Temporal variation in seed density and composition 61

5.2 Species richness and diversity 66

5.3 Life history 67

5.4 Germination rate 69

CHAPTER 6 71

CONCLUSION 71

REFERENCES 72 APPENDICES 82 Appendix 1 Temperatures (0C) and humidity (%) at Kirkleagh (latitude 27.2’S, 82

Appendix 2 Mean (■), max(▲) and min (♦) temperature of the first 20 days of

the germination trials in the glasshouse in January (A) and in July (B) 83

Appendix 3 Above ground vegetation species versus seed bank species at the

gully 84

Appendix 4 Above ground vegetation species versus seed bank species at the top

ridge 88

LIST OF TABLES

Table 2.1 Recommendations on seed bank sample sizes from the literature 27 Table 2.2 The main advantages and disadvantages of the germination and the seed

extraction methods for analysing soil seed banks 30

Table 2.3 Temperatures (0C) and humidity (%) at Kirkleagh (latitude 27o2’S,

longtitude 152o34’E) for the period 1978 – 1993 37

Table 4.1 Temporal variation in the germinable soil seed bank at the gully of the Kilcoy

pasture 45

Table 4.2 Temporal variation in the germinable soil seed bank at the top ridge of the

Kilcoy pasture 49

Table 4.3 Species richness and diversity (Shannon-Weiner Index) of the seed banks at

the gully and the top ridge of the Kilcoy pasture 54

Table 4.4 The spatial variation in the seed densities of annuals and perennials 55 Table 4.5 Spatial variation in the seed densities of native and introduced species 57 Table 5.1 The contribution of difference plant groups to the viable seed banks in

difference grassland types 63

LIST OF FIGURES

Figure 2.1 A map of parthenium weed distribution in Australia based on herbarium

records, where red dots represent the presence of parthenium weed 6

Figure 2.2 Diagrammatic representation of the four seed bank types 24 Figure 2.3 Long-term average rainfall at Lindfield (latitude: 26o84’S, longitude:

152o58’E, 14km away from the study sites) 37

Figure 3.1 Monthly rainfall (blue bars) in the year 2008, compared with long-term

rainfall (black bars) averages of the period from 1988 to 2008 at Lindfield

(14 km away from the study sites) 41

Figure 4.1 The germinable seed proportion of parthenium weed in comparison with

that of other plant groups in the germinable soil seed banks at the gully in the Kilcoy pasture 48

Figure 4.2 The proportion of parthenium weed in comparison with other groups of

species in the germinable soil seed banks at the top of a ridge in the Kilcoy pasture 52

Figure 4.3 Life forms of species detected in the germinable soil seed banks from

samples collected at the gully and the top ridge of the Kilcoy pasture during the entire study period 56

Figure 4.4 Seedling emergence of parthenium weed (○) and all other species (■) during

the first 20 days after wetting the soil samples collected in January from a gully (A) and a top ridge (B) of the Kilcoy pasture 58

Figure 4.5 Cumulative seedling emergence of parthenium weed (○) and all other

species (■) during the first 20 days after wetting the soil samples collected in January from a gully (A) and a top ridge (B) of a Kilcoy pasture 58

Figure 4.6 Seedling emergence of parthenium weed (○) and all other species (■) during

the first 20 days after wetting the soil samples collected in July from a gully

Figure 4.7 Cumulative seedling emergence of parthenium weed (○) and all other

species (■) during the first 20 days after wetting the soil samples collected in July from a gully (A) and a top ridge (B) of a Kilcoy pasture 60

LIST OF PLATES

Plate 2.1 Parthenium weed (Parthenium hysterophorus L.) (a) Early seedling growth

with hairy true leaves and alternate leaf arrangement; (b) Late seedling growth with rosette growth habit and deeply lobed leaves 4

Plate 2.2 Mature plant Note the small white flower heads and much-branched terminal

panicles 5

CHAPTER 1

INTRODUCTION

Parthenium weed (Parthenium hysterophorus L.), an annual herbaceous weed belonging

to the Asteraceae family, is a noxious and invasive weed, which is native to the Americas including the United States of America, Mexico, Cuba, and the West Indies (Evans 1997; Mahadevappa 1999) Often accidentally introduced into new locations as seed, it has now spread over many regions of East and South Africa, parts of South East Asia, to certain Pacific Islands, to India, Pakistan, and Australia It has spread at an alarming rate achieving the status of a major weed within a relatively short period of time (Adkins &

Navie 2006; Kohli et al 2006; Mahadevappa 1999; Navie et al 1998b; Picman & Picman 1984; Reddy et al 2007; Singh et al 2004) In Australia, parthenium weed was declared

one of the 20 “Weeds of National Significance” (WONS), mainly due to its potential to spread rapidly and its adverse impacts on the economy, the environment and society (Thorp & Lynch 2000)

Parthenium weed is a serious problem in many agricultural production systems as well as in non-agricultural areas, especially those that are sparsely covered or disturbed

(Brooks et al 2004; Navie et al 1998b; O'Donnell & Adkins 2005; Reddy et al 2007)

Parthenium weed infestations result in the reduction of productivity, increasing production costs and damage to the environment Its aggressive growth habit reduces the diversity and size ofother plant populations (Sridhara 2005 cited in Dhileepan 2007) as well as their seed

banks (Navie et al 2004), and it reduces the productivity of many kinds of pastures

(Adkins & Navie 2006; Chippendale & Panetta 1994; Haseler 1976) Parthenium weed has also been reported to aggressively colonize several kinds of arable systems leading to crop

yield loss of as much as 70% (Channappagoudar et al 1990; Tamado & Milberg 2004)

Additionally, the weed acts as an alternative host for a wide range of crop pests, which include certain kinds of nematodes, mycoplasma and black scarab (Pseudoheteronyx sp)

(Basappa 2005; Navie et al 1998b). In addition, severe parthenium weed infestations are to

Haseler 1976) Nonetheless, it is hardly ever a problem in areas that are ecologically healthy

and well covered with desirable plants (Brooks et al 2004; O'Donnell & Adkins 2005)

Parthenium weed poses various detrimental effects upon human and animal health Upon contacting parthenium weed plant parts or its pollen, people may suffer severe allergic reactions such as itching, alopecia, dermatitis, hay fever or asthma (McFadyen

1995; Navie et al 1998b) Likewise, parthenium weed is poisonous to grazing animals and the meat and milk production from livestock eating the weed can be tainted (Kadhane et al 1992; Navie et al 1998b)

The wide distribution of parthenium weed and its rapid increase in population size may be due to a number of factors including its prolific seed production, allelopathic

properties, or lack of effective management practices (Haseler 1976; Navie et al 1998b; Reddy et al 2007)

Understanding the size and composition of soil seed banks is of great importance in monitoring plant community structure and function The buried seed population has a role

in re-establishing a population or altering its composition (Viragh and Gerencser 1988;

Coffin and Lauenroth 1989; Rice 1989; Navie et al 2004) Knowledge of the germinable

seed bank is an integral part of any study on the ecology of communities and the

recruitment of species into those communities (Coffin & Lauenroth 1989; Navie et al

2004; Rice 1989) In turn, the composition and density of the soil seed bank is influenced

by the seed-production capability of above-ground vegetation (Coffin & Lauenroth 1989)

The main objective of this study is to measure the size and temporal and spatial variation in the germinable seed bank of parthenium weed in a grazed pasture at Kilcoy, south-eastern Queensland The composition of the entire germinable seed bank was also determined in order to characterize the structure and dynamics of the seed banks of communities infested by this invasive noxious plant species This will give an estimate of future plant population abundance and biodiversity Further work outside of the scope of this present project will monitor change as further weed management practices are applied

CHAPTER 2

LITERATURE REVIEW

2.1 Scope of the review

This review focuses on the biology of parthenium weed (Parthenium hysterophorus L.) and

its management and the soil seed banks The first part of the review is devoted to reviewing the basic biology of the weed and its management This includes a description of parthenium weed, its growth, spread and distribution, and its significance to agricultural production, pastoral industries, human health, and the environment The various approaches that are used to manage the weed are also mentioned The second part of the review stresses the importance of studying the soil seed bank in relation to pasture management and briefly reviews factors affecting the composition and size of soil seed banks An overview of parthenium weed soil seed banks is included as well

2.2 Parthenium weed (Parthenium hysterophorus L.)

2.2.1 Taxonomy and description of parthenium weed

The following description of parthenium weed is based on Everist (1974); Haseler (1976);

Navie (2002); Navie et al (1998b, 2004); and Wilson et al (1995)

Parthenium weed is a plant species belonging to the tribe Heliantheae of the Asteraceae (Daisy family), an extremely diverse family distributed worldwide The weed is

commonly named parthenium weed in Australia but various alternative names are used abroad For example, in India, it is known as bitter weed, carrot weed, broom-brush and congress weed; in the Caribbean, it is called whitetop, escobar amarga and feverfew; and known as false ragweed or ragweed parthenium in the United States of America

Parthenium weed is a profusely branched, erect and fast-maturing annual herb, which can flower within 4 weeks The weed may emerge, grow and flower vigorously all year round if conditions are favourable Its life cycle is divided into two main stages namely juvenile (rosette) stage and mature (adult) stage

Juvenile stage: Parthenium weed seed leaves are hairless, broadly elliptic with short

stalks (about 1.5 mm) while the first two true leaves are hairy, egg-shaped and apex rounded (Plate 2.1a) In its early growth stage, parthenium weed forms a rosette with large, hairy, dark green and radial leaves which are deeply divided into narrow pointed lobes The leaves are up to 20 cm long and 12 cm wide The large lower leaves are spread

on the ground like a carpet (Plate 2.1b)

Plate 2.1 Parthenium weed (Parthenium hysterophorus L.) (a) Early seedling growth

with hairy true leaves and alternate leaf arrangement; (b) Late seedling growth

with rosette growth habit and deeply lobed leaves (Wilson et al 1995)

Mature stage: Mature plants are erect, reaching up to 2.5 m in height although most plants

do not exceed 1.5 m Upon stem elongation, smaller, narrower and less divided leaves, in comparison to the radial leaves, are formed along the upper stem The stem is hairy, rigid and longitudinally grooved Both leaves and stem of parthenium weed are coated with fine soft hairs The whole plant is bluish or grayish-green in colour Flower heads, which produce five black seeds, are creamy-white, about 3 to 5 mm across and are clustered on large branched stalks arising from the leaf axils (Plate 2.2) The achene complex, i.e seed,

is shed gradually or retained on the inflorescence until the stem senescences Seeds are black and 1 to 2 mm in length

It is easy to confuse parthenium weed with species

such as annual ragweed (Ambrosia artemisiifolia L.), perennial ragweed (A psilostachya DC.), burr ragweed (A confertiflora DC.) or lacy ragweed (A tenuifolia Sprengel), especially when only the

vegetative growth stage is seen Parthenium weed can be distinguished from the above species by the following distinct characteristics: (i) in its early vegetative growth stage, parthenium weed does not possess opposite leaves; (ii) parthenium weed has a distinguishing longitudinally grooved stem; (iii) flower heads of parthenium weed are small, white and borne on much-branched terminal panicles

while those of Ambrosia species are monoecious,

unobtrusive and predominantly green in colour

2.2.2 The history of spread and present distribution

Parthenium weed, naturally growing in the tropical and subtropical Americas, from the southern USA to the southern Brazil and northern Argentina (Dale 1981), has been accidentally introduced into many regions around the world and achieved major weed status

within a relatively short period (Evans 1997; Navie et al 1998b) Until 1977, it did not have any place in the list of the world’s worst weeds (Holm et al 1977) Within the last few

decades, it has become one of the most dreaded weeds of the world (Mahadevappa 1999)

The introduction of parthenium weed into Australian has occurred on at least two

Plate 2.2. Mature plant Note the small

white flower heads and

much-branched terminal panicles

(Wilson et al 1995).

recorded in the 1950s) and the other in central Queensland, near Clermont, in 1958 (Haseler

1976; Navie et al 1998b) The most serious introduction was the second one, when

parthenium weed was imported in a contaminated pasture seed lot from Texas (USA) (Haseler 1976) This second introduction was not noticed until 1973, when a series of mild

winters and high rainfall summers favoured the rapid spread of the plant (Navie et al

1998b) The weed then rapidly spread through central Queensland and infested 170,000

km2 of prime grazing lands or 10% of the entire state (Chippendale & Panetta 1994) The

weed gained the status as a serious pest in Queensland in 1974 (Navie et al 1998b) and as a

weed of national significance in 2000 (Thorp & Lynch 2000) Although parthenium weed

is not yet a major weed in Australian

cropping systems and lands, it is a major

pest of grazing areas and is often

dominant along roadsides (Navie et al

1998b; Williams & Groves 1980)

Presently, parthenium weed occurs in

Queensland, New South Wales, Victoria,

Western Australia and the Northern

Territory (Figure 2.1) but has been noted

to have significant potential to spread

throughout all of the warm and

temperate, humid and sub-humid areas of

Australia (Navie 2002) However, it seems

to be confined to the areas not experiencing

extreme temperatures (<50C or >400C)

(Dale 1981) and heavy shading (>80%

shade) (Williams & Groves 1980)

Parthenium weed has been introduced into India between the 1950s and 1960s through importation of food grains from the USA It was first noticed in 1955 in Pune

(Maharashtra) (Kohli et al 2006) and has since spread throughout the country, infesting

about 5 million ha of land (Adkins & Navie 2006; Mahadevappa 1999) Parthenium weed

Figure 2.1 A map of parthenium weed distribution in Australia based on herbarium records, where red dots represent the presence of parthenium weed Map sourced from Australia’s Virtual Herbarium, 23rd December,

2008, <http://www.anbg.gov.au/avh/>

has spread gradually from one place to another becoming common along the highways, petrol bunks, railway tracks, bus stops on road sides and other waste lands (Mahadevappa 1999)

Parthenium weed has also spread into Trinidad, Guyana, Jamaica, Nepal, Israel, South Korea, Taiwan, North Vietnam, Bangladesh, Sri Lanka, southern China, certain Pacific islands (Vanuatu, New Caledonia, Tahiti and Hawaii) and Indian islands (the Mascarenes, the Seychelles, Rodriguez, Bourbon and Mauritius) (Mahadevappa 1999; Navie 2002) It has reached Africa, being recorded in Kenya, Madagascar, Mozambique, South Africa, Zimbabwe, Ethiopia, and may become more prominent in other parts of the

world in the near future (Adkins & Navie 2006; Evans 1997; Kohli et al 2006;

McFadyen 1992)

2.2.3 Habitat

Parthenium weed is a weed of a wide range of habitats throughout the world, including wastelands, pastures, agricultural areas, forest nurseries, open urban areas, roadsides, public lawns, railway track sides, petrol bunks, bus stops, along the edge of canals, new construction sites and along streams and rivers, and even on rooftops (Mahadevappa 1999;

Navie 2000) However, its preferred habitat varies depending on different geographical

locations For instance, in the Americas and India, parthenium weed commonly occurs in cultivated areas, being recorded in a wide range of crops; whereas in Australia, it is a serious problem in pastures, with very little adverse impacts in cultivated areasthough having been detected in some crops (Navie 2000)

Parthenium weed grows vigorously and prolifically in areas that are disturbed, both naturally and deliberately (i.e those disturbed by the traffic of vehicles and/or livestock) (Haseler 1976; Holman 1981) It is particularly troublesome along floodplains where soil

deposition and erosion occur (Navie et al 1999) The weed exists in a wide range of soil

types but favours soils of high fertility such as black, alkaline and cracking clay soils (Dale 1981) In other soil types, more severe soil disturbances are required for the establishment of

(Dale 1981) Parthenium weed can adapt to the areas with a wide range of annual rainfall, varying from 200 mm to 1,150 mm (Dale 1981) It has been reported that the major parthenium weed infestation areas in Australia are in the sub-coastal regions of central Queensland, where annual rainfall ranges between 500 and 700 mm with a dominant summer incidence (Haseler 1976) Parthenium weed particularly favours areas where brigalow

(Acacia harpophylla F.Muell ex Benth.) and gidyea (Acacia cambagei R.T.Baker) low open forests have been cleared (Dale et al 1978) Recently cleared lands used for livestock

are the common areas with heavy parthenium weed infestations (Holman 1981) This is the case in Australia where the main parthenium weed-infested areas are beef feeding pastures

(Anon 1985a cited in Navie et al 1998)

2.2.4 Growth and development

Although parthenium weed reproduces only by seed, it is able to successfully compete with the growth of other species owing to its substantial abilities to establish and grow vigorously Parthenium weed can germinate, grow and flower over a wide range of photoperiods and temperatures if moisture is favourable (Haseler 1976) It germinates more rapidly than other

plant species (Navie et al 2004) In its early growth stage, parthenium weed has a rosette

growth habit and spreads radially close to the ground, suppressing the growth of any

vegetation underneath (Mahadevappa 1999) Parthenium weed can flower within 4 weeks

after germination (Haseler 1976), seed prolifically and form an enormous seed bank in the

soil (Navie et al 1998b, 2004) The mature plant is erect (up to 2.5 m in height), branched and may form axillary branches down the stem when it gets older (Navie et al

much-1998b) Furthermore, a long tap root allows the weed to extract deep water within the soil profile for its vigorous growth and to reserve energy for fast regrowth if the weed is slashed

or grazed (Navie et al 1998b) Besides, parthenium weed seeds, which are small and light,

are easily dispersed for quick colonization in new areas

The aggressive invasiveness of parthenium weed in various parts of the world has been attributed to its allelopathic properties, which enable the weed to successfully suppress the growth of other crop, pasture and tree species Parthenium weed affects the

emergence, early growth and development of other species by releasing phenolic acids and sesquiterpane lactones from fresh plant parts (Adamson 1996; Mersie & Singh 1987;

Picman & Picman 1984; Tefera 2002; Wakjira et al 2005) The autotoxic effects of

parthenium weed were also reported as inhibition of pollination and fruit set of

neighbouring species (Evans 1997; Navie et al 1998b)

From Haseler’s observation (1976), soil moisture seems to be the major factor limiting the emergence and growth of parthenium weed, thus its main growth season in Australia is summer, coinciding with the abundance of rainfall Soil moisture is also a contributing

factor affecting the duration of flowering (Navie et al 1998b)

Although temperature is not the limiting factor controlling the growth and development of parthenium weed in Australia, it affects the duration of the vegetative phase prior to flowering (Williams & Groves 1980) Parthenium weed flowers earlier under the day/night temperature regime of 27/220C in comparison to that of 21/260C and 33/280C However, flowering does not require any specific day length (William & Groves 1980)

2.2.5 Biology of parthenium weed seed

Seed production: Seed is the only means of parthenium weed reproduction It is a prolific

seed producer, producing up to 25,000 seeds per plant or 300,000 seeds per m2 given that

sufficient moisture is available to produce a high density of vigorous plants (Navie et al

1998b) In abandoned fields in India, it was estimated that about 200,000 seeds per m2 were present in the seed bank (Joshi 1991) In good seasons, there may be two cohorts of parthenium weed, making the number of parthenium weed seeds produced even greater

(Navie et al 1998b) In Queensland’s pastures, the number of germinable parthenium weed

seeds buried in the soil was recorded to be varying between 1,500 and 34,000 seeds/m2

(Navie et al 1998b)

The aggressive invasiveness of parthenium weed in various parts of the world has

suppress the growth of other crop, pasture and tree species Parthenium weed affects the emergence, early growth and development of other species by releasing phenolic acids and sesquiterpane lactones from fresh plant parts (Adkins & Sowerby 1996; Mersie & Singh 1987; Picman & Picman 1984; Tefera 2002; Wakjira et al 2005) The autotoxic effects of parthenium weed were also reported as inhibition of pollination and fruit set of

neighbouring species (Evans 1997; Navie et al 1998b)

Seed dispersal: Parthenium weed seed is black, light (0.045-0.049 mg) and flanked by a

pair of floats which aids dissemination by wind and water (Joshi 1991) Wind does not transport parthenium weed seeds far from the plant (only just a few metres away) but whirlwinds can disperse numerous seeds to considerable distances (Haseler 1976) Water is thought to be an important means of parthenium weed dispersal, assisting the spread of

enormous numbers of parthenium seeds along the waterways in central Queensland (Navie et

al 1998b) Moreover, parthenium weed seeds are dispersed by livestock, native and feral animals (Navie et al 1998b) They are likely to be transported long distances in mud and

debris (Haseler 1976) In most cases of long distance dispersal, parthenium weed seeds are transported on vehicles, machinery, or livestock, or with crop and pasture seeds or in fodder

(Navie et al 1998b) Consequently, parthenium weed may spread to new areas that are

thousands of kilometres away from the nearest plants

Seed dormancy: It has been noticed that parthenium weed seed exhibits more than one

dormancy mechanism (Navie et al 1998a) McFadyen (1994) assumed that the germination

of parthenium weed seeds was not controlled by primary dormancy mechanisms In contrast, Picman and Picman (1984) suggested that water soluble germination inhibitors (sesquiterpene lactones, parthenin and coronopilin) present in parthenium weed seeds, especially freshly shed seeds, need to be leached before maximum germination is attained This explained why the germination rate of newly shed seeds was lower than that of seeds buried for several months and would, therefore, result in the formation of a more persistent

seed bank in the soil (Navie et al 1996b; Navie et al 1998a) It was also noticed that the

germination rates of parthenium weed seeds increased when the distance between them (i.e

seed density in the germination dish) and the washing period preceding the germination increased (Picman & Picman 1984)

Parthenium weed seed may be induced into a state of conditional physiological dormancy by ambient environmental conditions, as is the case with many other plant

species that possess light requirements when buried (Baskin & Baskin 1989 cited in Navie

et al 1998b) It is likely that when buried in the soil, parthenium weed seeds exhibit a form

of conditional dormancy which would result in the formation of more persistent seed banks (McFadyen 1994)

Seed longevity: There are varying reports on parthenium weed longevity from different

researchers While Butler (1984)) reported that the germination rate decreased from 66% after burial for 1 week to 29% after 2 years of burial; the viability of parthenium weed

seeds was greater in the work of Navie et al (1998a) (74% after 2 years of burial) and Tamado et al (2002) (more than 50% after burial for 26 months) Surface-lying seeds,

however, remained viable for periods no longer than 6 months (Navie 2002) These findings suggested that seed incorporation into the soil - which is enhanced by some forms of soil disturbance, such us flooding, cultivation or animal activities - is important for the long-term persistence of parthenium weed (Navie 2002) Some field observations showed that parthenium weed seed buried in the soil can remain viable for at least 6 years and will

germinate when brought to the soil surface (White 1994 cited in Navie et al 1998a)

Seed germination: Parthenium weed seeds are able to germinate over a wide range of

temperatures (from 9 to 30oC) but optimum germination occurs between 22 and 25oC

(Navie et al 1999) Very warm (above 30oC) or very cool (below 5oC) temperatures may limit the germination, particularly its rate (Williams & Groves 1980) The total germination was not affected by light but the germination rate showed some effects The germination rate declined when the differential of day/night temperature increased from 5 to 11oC at low mean temperatures (Williams & Groves 1980)

The germination limiting factors in the field appear to be soil moisture availability (Williams & Groves 1980) Germination plummeted from 91% to 50% and 12% when the soil moisture decreased from the field capacity to -0.07 MPa water potential and -0.24 MPa, respectively No germination was recorded in the soil kept at -0.9 MPa (Williams & Groves 1980)

The germination rate of parthenium weed seed was affected by the length of time

between shedding and the start of germination (Navie et al 1998a) During the first month

after shedding, no seedling emergence was recorded though considerable rainfall was available (31 mm) but 30% of the surface sown-seeds germinated in the succeeding 3 months after the first notable rainfall event (> 10 mm) By the end of the fourth month, a total of 51.4% of seeds had emerged and no more seedling emergence was detected after

the fifth month However, (Tamado et al 2002) observed no germination from the depth

below 5 cm in the soil Seeds of parthenium weed reached maximum germinability in the field, 1 to 6 months after shedding

2.2.6 Parthenium weed and its significance

Parthenium weed is regarded as one of the world’s seven most dreaded weeds (Mahadevappa 1999) It has appeared as a serious problem in many regions around the world imposing deleterious effects on crop production, animal husbandry, human health and biodiversity

Crop production: With allelopathic properties, parthenium weed has great potential to

aggressively colonise a wide variety of croplands, including those used for pasture grasses, cereals, vegetables, oil crops, nut crops, sugar cane, cotton, other weeds and even tree species, by inhibiting their germination and growth (Adkins & Sowerby 1996; Evans 1997;

McFadyen 1992; Mersie & Singh 1987; Navie 2002; Navie et al 1998b) The germination

of cowpea (Vigna unguiculata (L.) Walp.) and bean (Phaseolus vulgaris L.) was inhibited

to the extent of 50 and 60%, respectively (Mahadevappa 1999) Water soluble phenolics (caffeic acid, ferulic acid, vanicillic acid, anisic acid and fumaric acid) and sesquiterpene

lactones (parthenin and coronopilin) exuded from the weed parts also exhibit an inhibitory effect on the growth and nodulation of legumes, thus inhibiting nitrogen fixation and the growth of nitrifying bacteria (Kanchan & Jayachandra 1981)

Pollen grains of parthenium weed were claimed to impose negative effects upon the chlorophyll content of leaves with which they come into contact and can interfere with the

pollination and fruit set of surrounding species (Evans 1997; Navie et al 1998b) It was

evidenced that the presence of parthenium weed pollen on the stigmatic surface of maize

(Zea mays L.) resulted in a decrease in the grain-filling by up to 40% Poor fruiting of leguminous crops smooth crotalaria (Crotalaria pellida L.) and Asian ticktrefoil (Desmodium heterocarpon (L.) DC var van Meeuwen) was also recorded in plants covered

by parthenium weed pollen grains The presence of parthenium weed in agricultural crops not only suppresses the yields but also contaminates the harvested products thus reducing their marketability (Navie 2002)

Another adverse impact parthenium weed has on crop systems is its role as an alternative host for a wide variety of crop pests, acting as an inter-season reserve or sources

of inoculum Examples of this include the scarab beetle Pseudoheteronyx sp on sunflower (Helianthus annuus L.) in central Queensland; certain kinds of plant parasitic nematodes in the USA; the American serpentine leafminer (Liriomyza trifolii (Burgess)) on bell pepper (Capsicum annuum L.) in Texas; the polyphagous lepidopteran Diacrisia obliqua Walker

(Bihar hairy moth), an important pest of agriculture and forestry; the black bean aphid

(Aphis fabae Scopoli) on black bean (Castanospermum australe A.Cunn & C.Fraser ex

Hook.) in India, etc (Basappa 2005; Evans 1997; Navie et al 1998b) Furthermore, parthenium weed may function as a secondary host of some plant pathogens, including

Xanthomonas campestris cv phaseoli, Pseudomonas solanacearum, potato virus X and Y, tomato leaf curl virus and crop virus transmitted by the cotton whitefly Bemisia tabaci

Gennadius (Evans 1997)

The yield losses caused by uncontrolled parthenium weed populations in various

the extent of 40% and 50% in tomato (Solanum lycopersicum L.) and ragi crops (Eleusine coracana L.), respectively In irrigated sorghum (Sorghum bicolor (L.) Moench) in India, a

30% decline in sorghum grain weight causing a 35% reduction in grain yields was recorded

(Channappagoudar et al 1990) Tamado and Milberg (2004) warned of a severe yield loss

of up to 75% of grain sorghum in an unweeded control

Animal husbandry: Parthenium weed poses detrimental effects on livestock production,

animal health, and quality of meat and milk The presence of parthenium weed in pastoral regions may lead to a monoculture of parthenium weed pastures which are non-nutritious and unable to sustain grazing animals (Chippendale & Panetta 1994) This results in significant reductions in pasture carrying capacity by up to 40% in Australia (McFadyen 1992) and even as much as 90% in India (Evans 1997) Cattle grazing in parthenium weed-infested grasslands gain less weight in comparison to those feeding in parthenium weed-free areas This coupled with the additional cost for parthenium weed management (including herbicide application, labour, and machinery) cause further losses to producers (Chippendale & Panetta 1994) Annually, Queensland’s pastoral industries have lost AUD$ 16 million due to parthenium weed infestations (Chippendale & Panetta 1994) Losses caused by parthenium weed are estimated to range from AUD$ 65 to AUD$ 181 million per year to the Australian beef industries by 2070 (Adamson 1996) Farmers owning parthenium weed-infested pastures face a further negative impact on supplying pasture seeds and forage because of quarantine legislation on the movement of these products out of parthenium weed-contaminated properties (Evans 1997) An average of 600 parthenium weed seeds were detected per kg of contaminated pasture seeds when sampling

in parthenium weed-infested pastures, (Navie et al 1998b)

Although parthenium weed is not actively selected by the cattle as a feed because of its irritating odour, bitter taste and numerous trichomes on leaves, in many situations, animals will graze significant amounts of the weed resulting in serious consequences (Navie 2002) Parthenium weed-consuming animals, especially horses, experience a wide range of severe toxic symptoms such as pruritis, alopecia, loss of skin pigmentation, facial and body dermatitis, erythematous eruption, anorexia, diarrhoea resulting from

gastrointestinal irritation, changes in blood chemistry and inhibition of liver

dehydrogenizes and degenerative changes in their livers and kidneys (Evans 1997; Kadhane

et al 1992; Navie et al 1998b) Consuming significant amounts (15 to 50%) of parthenium weed in their diet, animals may die within 8 to 30 days (More et al 1982) The meat and

milk production from cattle, buffalo and sheep eating the weed may also be tainted (Evans

1997; Kadhane et al 1992; Navie et al 1998b)

Human health: Not only does parthenium weed induce deleterious effects on crops and livestock, it also causes serious health hazards to humans Those working in parthenium contaminated lands may suffer allergenic eczematous contact dermatitis (AECD) as a result

of repeated skin contact with parthenium and/or allergenic rhinitis if breathing in parthenium weed pollen Allergenic rhinitis may develop into bronchitis or asthma if the pollen gets in the respiratory tract during breathing (Evans 1997) Severe dermatitis leads

patients to fatigue and weight loss (Navie et al 1998b) Even hundreds of deaths, which were attributed to allergic reactions, were recorded (Adkins et al 2005) On average,

allergenic dermatitis and asthma cost approximately AUD$ 8.0 million annually in medical treatment in Australia (Page & Lacey 2006)

Plant communities: Owing to its aggressive growth and adaptability to varying soils and microenvironments, parthenium weed potentially disrupts natural ecosystems adversely affecting the structural composition and dynamics of the diversity of the native flora The presence of parthenium weed tends to replace the dominant flora in a wide range of habitats

including native grasslands, open woodlands, river banks, floodplains, national wildlife parks and wastelands (Asad & Rukhsana 2006; Chippendale & Panetta 1994; Evans 1997; Kohli et

al 2006; Navie et al 1998b) Whenever invading natural communities, it forms a territory of

its own by replacing the indigenous natural flora, resulting in very little or no other vegetation

species being seen in parthenium weed dominated areas (Asad & Rukhsana 2006; Chippendale & Panetta 1994; Navie et al 1998b) Parthenium weed may aggressively

compete against neighbouring species in pastures creating parthenium weed monoculture

communities (Chippendale & Panetta 1994) Navie et al (2004) noted that the presence of

parthenium weed in grassland communities lowered the diversity of these seed banks, thus

reducing future regenerative ability of some of the native plant species

2.2.7 Parthenium weed management

Prevention: Reducing the spreading of parthenium weed via the movement of water,

vehicles, machinery, livestock, grain, seed and other products is highly desirable Thus, preventing its spread is considered as the most cost-effective management tactic

Parthenium weed is declared noxious in all states of Australia (Queensland Department of Natural Resources Mines and Energy 2004) and is categorised Class 2 pest

under Land Protection (Pest and Stock Route Management) Act 2002, which requires

landowners to control the weed in their land and waters under their control (Queensland Department of Primary Industries and Fisheries 2007) Legislation prohibiting the movement of parthenium weed-contaminated products out of their infested areas has been implemented

Hygiene is of significant importance in preventing the spread of the weed down facilities have been installed in parthenium weed-infested areas in Australia for contractors to properly clean their harvesters and other equipment before leaving the contaminated areas and the state (CRC for Australian Weed Management & Commonwealth Department of the Environment and Heritage 2003) It is highly recommended not moving stock in wet weather as they readily transport seeds in muddy soil Particular care should be taken on newly-arriving stock and on areas where purchased hay has been spread out (Queensland Department of Primary Industries and Fisheries 2007) Parthenium weed infestations on roadsides and public areas should be properly controlled to reduce its spread

Wash-It is recommended to avoid keeping land fallow to prevent parthenium weed invasion and further spread (Angiras & Kumar 2005)

Maintaining pastures in good condition with a high level of vegetation cover and with

a desirable pasture composition is the best way of controlling large-scale parthenium

weed-infested areas and preventing new infestations in parthenium weed-free pastures (Adkins et

al 2005; Navie 1998a) In order to avoid bare patches due to heavy grazing and livestock

congregation at watering points, it is recommended to break up large paddocks by fencing them into single units and to rotate stock between them (CRC for Australian Weed Management & Commonwealth Department of the Environment and Heritage 2003) Severe parthenium weed-infested areas can be fenced off to prevent seed spreading to parthenium weed-free areas (Holman 1981) In serious parthenium weed-infested pastures, no grazing paddock for at least one full summer followed by very light grazing in the following winter

is required for the pasture rehabilitation (Holman 1981)

Physical methods: Manually removing the weed is considered to be effective in small areas

and isolated lots such as flower beds, lawns, kitchen gardens and in intensively cultivated agricultural fields if it is uprooted before flowering, then burnt or composted (Mahadevappa 1999) Hand weeding, however, appears to be neither economical nor

practical in vast areas with heavy infestations and is not recommended in Australia (Navie

et al 1998b) This method needs to be repeated when the weed reappears

Mechanical management such as grading, slashing and ploughing is not appropriate

as this approach may support the further spread of parthenium weed seed (Haseler 1976) Mowing or slashing also leads to the rapid regeneration of the weed from lateral shoots close to the ground, which is quickly followed by flowering with abundant seed production

(Muniyappa et al 1980; Navie et al 1998b; Singh et al 2004)

Fire was noted to produce positive results only after the first heavy rains when the majority of parthenium weed seed had germinated (Holman 1981) Burning, however, usually gives only short term control as new plants sprout from remaining roots and stumps (Dhawan & Poonam 1995) and will probably be only effective when combined with the sowing of pasture seeds and other management practices (Haseler 1976) Recent work

of Vogler et al (2006) demonstrated that burning was not an effective practice for

controlling parthenium weed It did not significantly affect the germinable soil seed bank of

parthenium weed but resulted in a one-off increase in the weed population, which rapidly declined after subsequent fires

Chemical control: The availability of a wide range of effective herbicides allows

parthenium weed to be controlled successfully in almost every situation Herbicides registered for controlling parthenium weed in various situations in Australia are clearly presented by the Queensland Department of Primary Industries and Fisheries (2007) while those applied in other countries are shown in Mahadevappa (1999)

Non-crop land: Atrazine at 1.25 to 1.4 kg a.i./ha has been noted to be effective in preventing the weed emergence up to 150 days after spraying (Muniyappa et al 1980)

whereas applying up to 2.0 kg a.i./ha in late July or early August, which coincides with peak growth due to monsoon rains in India failed to provide satisfactory control of adult

plants (Singh et al 2004) Although tebthiuron (1.0 kg a.i./ha) exhibited great pre-emergent

control of parthenium weed, its use on a large scale may be limited because it restricted

annual grass cover (Brooks et al 2004) Tebuthiuron and imazapic (240 g a.i./ha) were noted to be best suited to the areas that were covered by existing perennial grass (Brooks et

al 2004)

Control efficacy varies widely with herbicides, applied rates and parthenium weed

growth stages at the time of spraying (Reddy et al 2007; Singh et al 2004) Pre-emergence

herbicides that inhibit the pigment and photosynthesis of parthenium weed (norflurazon 2,240 g a.i./ha; clomazone 1,120 g a.i./ha; fluometuron 2,240 g a.i./ha; and metribuzin 700

g a.i./ha) provide effective control of the weed 4 and 6 weeks after treatment (Reddy et al

2007) Post-emergence herbicides usually provide better control on the rosette stage than on

other stages (Reddy et al 2007) At the flowering stage in fallow land, parthenium weed is

killed by bromacil application at 2 kg a.i./ha

Crops: Application of glyphosate (0.75-1.00%), metribuzin (0.3-0.5%) and 2,4-D (0.50%) has provided satisfactory control of parthenium weed under the diverse agro-climatic

conditions of India (Yaduraju et al 2005) Kandasamy (2005) noted the efficacy of diquat 0.5

kg/ha in 500 L of water in controlling parthenium weed at all growth stages while applying 2,4-D sodium salt 2 kg/ha or MCPA 2 L/ha in 400 L of water effectively control the growth

of parthenium weed seedlings The author also noted the importance of exhausting the existing soil seed banks to prevent further germination and seed bank accumulation by applying glyphosate continuously for between 3 and 4 years

Pastures: Haseler (1976) noted that parthenium weed is susceptible to a number of

herbicides when being applied at high volume (2000 L/ha) Applying 2,4-D (4 kg a.i./ha), picloram (0.8 kg a.i./ha), dicamba (1 kg a.i./ha), diuron (2 kg a.i./ha), bromacil (2 kg a.i./ha), karbutilate (1 kg a.i./ha) or atrazine (3 kg a.i./ha) effectively controlled the weed (Haseler 1976) In general, spraying a mixture of atrazine and 2,4-D gives better control efficiency as 2,4-D kills existing plants while atrazine provides long-term residual activity but has little knockdown effect by itself (Parson & Cuthbertson 2001)

The treatment of metribuzin (0.2% solution) effectively controlled parthenium weed with no parthenium weed-regrowth observed up to 3 months after application and no effect

on the growth of native grass species (Rajkhowa et al 2005) The safety and efficacy of

metribuzin in controlling parthenium weed (even at its advance stage) in grassland was also noticed by Angiras and Kumar (2005) Metsulfuron methyl at 0.005 and 0.01% applied at the 2 to 3 leaf stage in November was found to effectively control the weed up to February and reduced its population considerably by lessening the emergence of its new flush in India (Angiras & Kumar 2005) Atrazine 0.3% and 2,4-D applied at the 2 to 3 leaf stage is less effective than metsulfuron in controlling parthenium weed (Angiras & Kumar 2005)

Chemical control over large areas of open grazing land could hasten rehabilitation but may not give an economic return (Sir Alan Fletcher Research Station & Queensland Department of Primary Industries in Central Queensland 1977) Spraying should be applied

in conjunction with burning and no grazing to allow penetration of the chemicals and a further spell to allow perennial grasses to re-establish (Sir Alan Fletcher Research Station

& Queensland Department of Primary Industries in Central Queensland 1977)

To achieve long-term control of parthenium weed, spraying the weed before it sets seed and keeping a close watch on treated areas for at least 2 years are highly recommended

(Navie et al 1998b) Unfortunately, the approach is only feasible in cultivation or in small

areas of pasture and appears to be not economically viable in very large parthenium cover areas (Jayasuriya 2005; Parson & Cuthbertson 2001)

weed-Biological method: This approach is believed to be a viable technology for large scale

applications, relatively less expensive over the long term and have fewer side effects on the

environment (Haseler 1976; Yaduraju et al 2005) The biological program in Australia has

provided a positive return bringing a AUD$ 2 benefit for every AUD$ 1 spent on this campaign (Page & Lacey 2006)

Beneficial insects: Of the nine insects that have been released in Australia since 1980, five have established successfully, one has established at a few sites and it is too early for two

of them to determine their permanent establishment status (Navie 2002) Of the successfully established species, only two have appeared to be effective in having some degree of

management of the weed; these are the leaf-feeding beetle Zygogramma bicolorata Pallister and the stem-galling moth Epiblema strenuana Walker (Dhileepan 2007)

When present in large numbers and for a long duration, Z bicolorata substantially

defoliated the weed causing reduction in shoot and root biomass, plant height, flower production, and soil seed bank The decline of the existing parthenium weed soil seed bank

in areas in which the leaf-feeding beetle has been released suggested that the density of the weed may reduce significantly within 6 to 7 years (Adkins & Navie 2006) Likewise,

management of parthenium weed in pastures has been noted to be easier after the release of the stem-galling moth Epiblema strenuana (Dhileepan 2007) It has been reported that several larvae are sufficient to stunt the plant and reduce its competitiveness as well as seed

production (McFadyen 1992) One limitation of E strenuana application is that its impacts

seem to be confined to vigorous plants (Adkins & Navie 2006) Less vigorous parthenium weed plants can avoid the gall-related damage as they were unable to sustain the gall development

Pathogens: Two rusts have been released in Australia, namely the summer rust Puccinia melampodii Dietel and Holway and the winter rust Puccinia abrupta var partheniicola (Jackson) Parmelee (Dhileepan 2007) The winter rust hastens leaf senescence, noticeably decreased the life span, dry weight and flower production of parthenium weed In heavy winter rust infestations, the flower production of the weed is reduced by up to 90% (Evans 1997) The summer rust is suitable in warmer weather

conditions (Dhileepan et al 2006) It weakens the weed by damaging the leaves over the

summer growing season Prevalence of rust infection increased on parthenium weed as the

plants matured (Dhileepan et al 2006) However, drought and consistently high night-time

temperatures have limited the dispersal of the rust from the released areas resulting in a negligible impact (Adkins & Navie 2006; Evans 1997)

The effectiveness and abundance of biocontrol agents depend largely upon seasonal conditions and green plant resources for survival of insects These are particularly highlighted during long dry periods when insect populations are reduced and need time to recover (Dhileepan 2007) The effect of parthenium weed biological control on grass recovery was also evaluated Significant increase in grass production due to biological control of beneficial insects and pathogens in Australia was observed but not consistently, only in 1 to 2 years of the 4 year trial An increase in grass biomass by 40%, 52% and 45%

as a result of defoliation by Z bicolorata, galling by E strenuana and combined effects of

E strenuana and the summer rust P melampodii was recorded (Dhileepan 2007)

Antagonistic plants: Many plant species have been screened to exploit their allelopathic properties as biocontrol agents regulating the weed population (Gautam et al 2005; Mahadevappa 1999; Rajkhowa et al 2005)

Of those exhibiting allelopathic effects on the growth of parthenium weed, Cassia sericea Sw (synonym Senna uniflora (Mill.) Irwin & Barneby) was found to be the most effective agent (Mahadevappa 1999) The growth of C sericea in pastures is not promoted

(Gautam et al 2005) Besides, it also acts as a major host of Bemisia whiteflies and as a

reservoir of the tomato leaf curl virus in India, which is easily transmitted to nearby crops (Evans 1997)

A glasshouse study in Australia identified three species which have forage value to pastoral industries and exhibit superiority over the growth of parthenium weed The three

species were butterfly pea (Clitorea terneata cv Milgarra), bisset bluegrass (Bothriochloa insculpta cv Bisset) and floren bluegrass (Dicanthium aristatum (Poir.) C.E Hubb (O'Donnell

& Adkins 2005) Further studies need to be conducted in order to determine their full value

An ecological survey in Pakistan revealed that the frequency and density of parthenium weed were considerably lower in cogongrass-dominated areas than in nearby

areas without the presence of congongrass (Imperata cylindrica L P.Beauv) (Anjum et al 2005) Further, the aqueous extracts of I cylindrica show inhibitory effects on germination and seedling growth of parthenium weed (Anjum et al 2005)

Integrated weed management: Several methods have been recommended for suppressing

the growth of parthenium weed but none of them alone appeared to be satisfactory because

of a series of limitations such as inefficiency, high cost, impracticability, environmental pollution, and slow results The integrated approach combining a number of available methods such as prevention, physical, cultural, biological and chemical tactics is likely to exhibit the best results

Preventing the spread of parthenium weed is a cost effective management approach but even strict quarantine measures cannot reduce parthenium weed movement to zero Hand weeding is effective in small areas, isolated pockets and in intensively cultivated agricultural fields but it is neither economical nor practical in vast areas with heavy parthenium weed infestation Several herbicides appear to give effective control of the weed but the need to reapply chemicals makes this approach expensive and impractical, particularly in large grasslands, and may harm the environment Biocontrol is considered to provide the best long term solution for parthenium weed management without posing any

harmful effect on the environment So far, the program has yielded considerable positive returns in Australia However, because of climatic variations, not all agents were equally effective in all places and all year round For example, from 1990 to 2000, severe

defoliations (60%) were observed only in 3 years in central Queensland (Dhileepan et al

2000) whereas galling, though evident in all years and in all areas with parthenium weed, at more than four galls per plant was observed only in 2 years The summer rust incidence exceeded 60% in 2000 but remained very low since then because of dry conditions

(Dhileepan et al 2006) These observations highlight the necessity of combining all

available management measures to ensure effective control of parthenium weed to reduce its adverse effects on communities

2.3 Soil seed bank

2.3.1 Soil seed bank concept

Definition: Seed bank is the term used to denote the reservoirs of viable seeds and fruits

present in the soil and on its surface (Robert 1981) Seeds in the litter/humus layer are often included “Viable seeds” are those being able to germinate, given appropriate germination

conditions and the lack of dormancy status (Grundy & Jones 2002)

Classification: There have been at least 10 published soil seed bank classification systems

(Csontos & Tamás 2003) Seed longevity is the main factor used for distinguishing categories, but dormancy and germination types also have an important role Systems considering relatively few seed bank categories have been the most commonly proposed in contemporary plant ecology In contrast, systems involving high numbers of categories have received limited interest because the detailed ecological knowledge of individual species required for their successful categorization is usually missing (Csontos & Tamás 2003)

The simplest classification system divides seed bank into two primary groups,

transient and persistent seed banks (Parker et al 1989) Transient (or temporary) seed

banks are made up of seeds of short life species, which do not present any dormancy

these species decreased rapidly at a rate of around 80% Persistent seed banks are composed of seeds of species that are more than one year of age and reserves of seeds remain in the soil year after year, generally buried in the soil

Figure 2.2 Diagrammatic representation of the four seed bank types

(Thompson & Grim 1979)

Shaded areas show seeds capable of germinating immediately if conditions are favorable Unshaded areas show seeds which are viable but not capable of immediate germination

In grassland communities, seed banks are sub-classified further into four main types (Figure 2.2)

Seed bank type I is composed of species with transient seed banks present only

during the late spring and summer and which germinate nearly synchronously in the cooler moist conditions of the autumn The distinct features of seeds associated with this seed bank type are: large size and/or elongated structure; lack of pronounced after-ripening or dormancy mechanisms; ability to germinate over a wide range of temperatures; and ability

to germinate in the light or in continuous darkness This type of seed bank provides a regenerative mechanism for populations of established plants

Seed bank type II consists of species with transient seed banks present during the

winter and which colonize vegetation gaps in the early spring This seed bank type is

strongly associated with species of temperate zone and represents a specific adaptation delaying germination until the beginning of the growing season

Seed bank type III comprises species whose seeds germinate soon after they are

released in autumn but a small proportion of seeds fails to germinate directly and some of these become incorporated into a persistent seed bank

Seed bank type IV: The seeds of the species belonging to this group cannot germinate immediately after dispersal as germination may be restricted by temperature and light conditions; thus a large persistent seed bank is maintained The size of this seed bank type changes little with season but is large in relation to the annual production of seeds The majority of seeds belonging to this type retain their viability in the soil for a long time

The significance of seed bank study: Seed banks in the soil are of fundamental importance

in the study of life-cycles of plant individuals, regeneration strategies, demography of plant populations and vegetation dynamical processes in communities (Coffin & Lauenroth 1989; Thompson & Grim 1979; Viragh & Gerencser 1988) The soil seed bank content is regarded as the potential or initial population which can also determine the direction of further vegetation changes (Coffin & Lauenroth 1989; Viragh & Gerencser 1988)

Some of the most important characteristics of populations reproducing primarily by seeds are the capacity of individual plants in determining the seed production and the adaptive capability of the species which can define the period when the seeds retain their viability in the soil The latter trait can mean a survival mechanism for a given population

at the same time Therefore, seed banks are significant in the maintenance of populations and can also act as a genetic stabiliser when the number of individuals fluctuates (Viragh & Gerencser 1988)

The role of soil seed banks is important during population regeneration following either natural or anthropogenic disturbances (Butler & Chazdon 1998; McIvor & Gardener

vegetation diversity and richness, which is an important implication for pasture management The regeneration role of seed banks is less important in communities that lack disturbance since germination of buried seeds is often stimulated by soil disturbance (Viragh & Gerencser 1988) The results of soil seed bank studies are one of the useful indicators in predicting successional changes and interpreting historical plant communities which occupied an area (if the viability of seeds is known), predicting and interpreting the results of various intensities of disturbance, and interpreting the pathways of natural regeneration and the strategies of establishment of individual species (Buhler 1997; Hopkins & Graham 1983; Viragh & Gerencser 1988)

Seed bank studies are also practically significant, especially in arable cultivation areas The presence of large numbers of seeds in agronomic crop production systems means that there is a continuing need for weed control to maintain the seed bank at the lowest feasible level in order to minimise interference with crop production (Buhler 1997; Robert 1981) The results of the species composition of arable seed banks can be used as a guide

on the choice of cultivated plants and of herbicides to be applied for the next year (Viragh

& Gerencser 1988)

The knowledge of weed seed banks in grasslands is particularly valuable when the swards are renewed This is because the success of a seed bank is noted to depend on the seed density ready to germinate when replacement of a plant is necessary and when the environmental conditions for establishment are favourable (Christoffoleti & Caetano 1998)

In addition, in grassland communities, the persistent seed banks of the less desirable grasses contribute to sward deterioration while the reserves of legume species seeds are valuable in the maintenance of pastures, especially in the areas subjects to drought (Buhler 1997)

Improving and applying the knowledge of weed seed bank dynamics is an important aspect of developing proper management strategies for sustainable use in arable cultivation areas (Buhler 1997) as well as grassland areas (McIvor & Gardener 1994)

2.3.2 Soil seed bank study approaches

Sampling methods

Timing of sample collection: The time of year at which samples are collected has a

significant effect on the results of a seed bank study, e.g samples taken after the seeds have been exposed to natural chilling during winter may yield remarkably more seedlings than those taken in autumn (Robert 1981) A spring sampling date appeared to be more reliable than an autumn date, probably because many viable seeds die during winter (Forcella 1992)

Sample volume: The amount of soil collected has a considerable effect on

achieving representative numbers of seed taxa and individuals of the studied sites (Csontos

2007; Page et al 2006) The recommendations for adequate seed bank sample size vary

among authors and community systems and are summarised in Table 2.1 by Page et al

(2006) With respect to numbers of seed individuals, which showed high degrees of spatial aggregation of seed in soil, it has been concluded that the use of many small samples was superior to using few large samples (Champness 1949) However, the size and number of

samples necessary for reliable seed-density analysis remains uncertain Page et al (2006)

and Bigwood and Inouye (1988) suggested that sampling effort may be optimized by taking fewer samples per transect in any one site, but selecting more transects and sites is likely to optimise the accuracy of seed bank studies

Sample depth: In general, as the depth of soil layer increases, the number of seeds

present, as would be expected, decreases drastically When testing the proportional

distribution of seeds with depth, Hodgkinson et al (1980) concluded that 78% of the seeds

were located in the top 2.5 cm and the seed content of the layers below 10 cm was very low Leck and Simpson (1987 cited in Csontos 2007) found that in the soil of flood-plain meadows, the layer between 4 and 6 cm contained only about 7-18 % of the seeds found in the top 2 cm and the percentage of seeds detected in the layer between 8 and 10 cm was

Table 2.1 Recommendations on seed bank sample sizes from the literature

(adapted from Page et al 2006)

Cultivated 60 – 87 cores

(diameter 3.3.-5 cm, depth 25 – 30 cm)

For species not abundant and m=5

Based on Poisson or Negative Binomial distribution

Gain in precision does not compensate for increase effort

Benoit et al 1989

20 – 25 cores (1500 – 2000 cm2)

For satisfactory estimates of legume seed populations

Carter 1985

70 cores (1.3 cm in diameter)

Optimum for weed seeds in grasslands to reduce SE to 10%

Champness 1949

100 cm2 Species diversity of weeds in

ryegrass pasture do not increase significantly beyond this

Forcella 1984

100 – 200 samples (core size not stated)

To estimate mean seed density (if expected mean 1-5)

Goyeau & Fablet

1982

50 cores (7.3 cm diameter)

Recommended for pasture legumes Jones & Bunch

1988

Non-cultivated 4000 – 6000 cm3 Volume for climax forests based on

species area curves

Hyashi & Numata

1968

500 – 600 cm3 Volume for Japanese grasslands

based on species area curves

Hyashi & Numata

1971

400 cm3 Volume for early successional stage,

from species area curves

Numata et al 1964

50 – 100 samples (7x7 cm)

Recommended number of samples to measure seed density

Thompson 1986

*: cited in Page et al (2006)

germinate in the field; therefore, to obtain an accurate scenario of the taxa and number of individuals potentially germinating in the field, there is no need to sample deeper (Silcock

1993 cited in Page 1997) Available literature on seed bank studies, however, showed that samples as deep as 30cm have been collected in the past (Howe & Chancellor 1983)

Seed bank analyzing techniques: There are two basic techniques commonly used to analyse

seed bank density and composition (Adkins & Navie 2004; Forcella 1992; Leck et al 1989;

Robert 1981) In both approaches, the preliminary steps are to take the representative soil samples from the chosen study site and transfer them to the laboratories for drying Then, the representative samples are further processed using either the seedling emergence or the seed

extraction method (Forcella 1992; Leck et al 1989; Robert 1981)

Seedling emergence method (germination method): This method determines the

number of seeds in a soil sample by directly placing the soil sample thinly on a suitable medium, keeping the soil sample moist, then counting and identifying the seedlings that emerge After counting, the seedlings are removed to prevent competition for light with any new seedlings that may emerge (Robert 1981; Thompson & Grim 1979) Seed germination may be promoted by soil stirring, stratification (chilling), or the application of germination stimulants (Adkins & Navie 2004; Brenchley & Warington 1930) This is the longest used approach but is a time-consuming technique (Forcella 1992) To obtain a reasonable measurement of all viable seeds present, the samples must be kept for a considerable time, may be up to as long as 3 years (Brenchley & Warington 1930) Temperature regimes during the observation period may have certain effects on the viable seed content Representative samples analysed under the temperature regime of 30/250C were dominated by monocotyledon species while dicotyledons were numerous in those at 20/150C (Mott 1972)

Seed extraction method: This method involves physically separating seeds from

the soil based on size or density differences The physical separation technique is not entirely effective and hand-sorting is also required Seed isolation may involve washing,

the seed The seed is then identified and tested for viability (Adkins & Navie 2004; Robert 1981) The sequence of these steps varies among studies and is clearly illustrated

by Robert (1981)

Table 2.2 The main advantages and disadvantages of the germination and the seed

extraction methods for analysing soil seed banks (Robert 1981)

Advantages - The required effort is spread

over a period of time;

- Each seedling represents a viable seed;

- Seedlings are easier to identify than seeds

- Results available as soon

as processing complete;

Disadvantages - Delay between sampling and

obtaining the results

- Underestimate of some species because of dormancy

- Seeds are difficult to identify;

- Over estimation of the potential flora;

- Time-consuming, 1.0 to 1.5 days for each 100g of soil

Seed identification appears to be a major problem when employing the technique of seed extraction since it is not easy to distinguish among similar species using simple morphological characters of seeds (Adkins & Navie 2004) This difficulty is exacerbated because seeds presenting in the seed bank may have lost some or all of the diagnostic characteristics after being buried in the soil

It was demonstrated that more species are detected with seed extraction methods than with seedling emergence techniques (Gross 1990) However, it is likely that the frequency and density of a species within the studied seed bank are overestimated if

inviable seeds are present in many samples While Barberi et al (1998) noted that there