Biology and management of Polygala myrtifolia

Robin Adair Australis Biological Latrobe University, Botany Department

www.australisbiological.com.au Background

. Australis Biological . Weed Sciences – Research Leader . Polygala – 1996 . IX International Sym on Biocontrol of Weeds . 1998 - Chrysanthemoides . 1999-2004 – Acacia, Victorian Coastal Council - $5000 . 2009-2012 – City Greater Geelong Building an understanding of Polygala

Essential for effective management Strengths and weaknesses Focus on weaknesses - control Understand the strengths – resilience Build effective control options Biology, ecology (native and non-native), control options, integrated control Taxonomy Polygala myrtifolia

• Polygala is Greek meaning ‘much’ ‘milk’ - reputedly promotes secretion of milk by stock • myrtifolia similarity to myrtle leaves

• Polygalaceae - 13 genera , 830 species • (related to Tremandraceae, not Fabaceae)

Bredemeyera (6) Salomonia (1) Comesperma volubile (24) Polygalaceae

. Polygala is the largest genus – 300-725 species . cosmopolitan (except NZ, Polynesia, Antarctica) . centres of diversity – South Africa, Americas . South Africa/Madagascar – 230 spp . Australia – 44 native to northern Australia . annuals . 5 exotic species . P. myrtifolia, P. virgata, P. vulgaris (Victoria) . P. japonica (native) P. chamaebuxus grandiflora P. senega (Americas)

P. vulgaris - Europe

P. calcarea - Britain

P. cruciata - USA

P. sanguinea USA

P. pauciflora - USA P. alba - USA Polygala japonica

P. isingii lateral petals wings (lateral sepals 2 )

crest keel (petal)

P. obscura – New Mexico

crest (a)

(b) P. myrtifolia var. myrtifolia

Polygala virgata P. myrtifolia var. grandiflora

P. myrtifolia – myrtle-leaf milkwort, Bellarine pea, boukappie, septemberbossie • South African – uncommon in wild, common in horticulture • high variability in South Africa • not weedy in South Africa • very weedy in Australia • coastal vegetation at risk

potential current P. myrtifolia – Betty’s Bay Ocean Grove 2010 Invasion history

• Earliest record - nursery catalogues in Adelaide, 1845 • First collection – NSW , Hasting River ,1856 • South Australia the first collection, Adelaide, 1920 • Western Australia, Katanning, 1901 • Tasmania, Taroona, 1946

• In Victoria, was introduced to Mt Eliza, 1851 • First herbarium collection - Brighton and Mentone, 1886 • Bellarine Peninsula, the earliest record is Point Lonsdale, 1912 • Mornington Peninsula, naturalization well established by 1943 • Not established until 1950s–1960s • South-western Victoria, well-established prior to 1976 • Invaded Discovery Bay Coastal Park from a hedge on private land • Earth moving equipment may have exacerbated spread, particularly along boundary firebreaks • Spread rapid 1981 and 1990 • A major threat to biodiversity values Impacts

• loss of biodiversity - competition and allelopathy (no evidence)

• increased biomass and fire-fuel levels (no evidence) • reduced accessibility • no threat to agriculture • host to Cucumber Mosaic Virus – threat to vegetable and ornamentals State Vegetation communities invaded Species and vegetation communities at risk Victoria Volcanic Plains Bioregion: Heathy Dry Forest, Heathy Acacia uncifolia, Adriana quadripartitia, Caladenia calcicola, Woodland, Coastal Dune Scrub, Coastal Headland Scrub, Caladenia insularis, Caladenia colorata, Corybas despectans, Impacts Coastal Tussock Grassland, Coastal Alkaline Scrub, Eucalyptus leucoxylon ssp. bellariensis, Poa poiformis var. Lowland Forest ramifer, Prasophyllum litorale, Prasophyllum spicatum, Otway Plain Bioregion: Coastal Saltmarsh, Grassy Pterostylis cucullata, Taraxacum cygnorum Woodlands, Coastal Tussock Grassland, Coastal Dune Scrub and Grassland Mosaic, Bellarine Yellow-Gum Communities: Coastal Moonah woodland, Bellarine Yellow- Woodland , Limestone Grassy Woodland Gum Woodland, Limestone Grassy Woodland, Wimmera Low Warrnambool Plain Bioregion: Herb-rich Foothill Forest, Rises Woodland Coast Banksia Woodland Gippsland Plain Bioregion: Moonah Woodland Glenelg Plain Bioregion: Heathy Woodland, Damp Sands Herb-rich Woodland Heathy Herb-rich Woodland Highlands - Southern Fall Bioregion: Riparian Forest, Dry Forest Wimmera: Low Rises Woodland

South Allocasuarina verticillata Woodland, Eucalyptus Acacia enterocarpa, Acacia pinguifolia, Acacia whibleyana, Australia camaldulensis var. camaldulensis Woodland, Eucalyptus Acacia rhetinocarpa, Caladenia tensa, Caladenia argocalla, diversifolia ssp. diversifolia Mallee Forest, Eucalyptus Caladenia behrii, Caladenia colorata, Caladenia diversifolia ssp. diversifolia Mallee Woodland, Eucalyptus richardsiorum, Caladenia rigida, Caladenia formosa, Correa fasciculosa (mixed) Woodland, Eucalyptus fasciculosa calycina var. calycina, Dodonaea procumbens, Glycine Woodland, Eucalyptus goniocalyx ssp. goniocalyx (mixed) latrobeana, Haloragis eyreana, Ixodia achillaeoides ssp. Woodland, Eucalyptus obliqua (mixed) Woodland, arenicola, Olearia pannosa ssp. pannosa, Pleuropappus Eucalyptus obliqua Woodland, Leucopogon parviflorus phyllocalymmeus, Pomaderris halmaturina ssp. halmaturina, (mixed) Shrubland >1m, Melaleuca brevifolia Shrubland Prasophyllum frenchii, >1m, Melaleuca halmaturorum Shrubland >1m, Prasophyllum pallidum, Prasophyllum pruinosum, Melaleuca lanceolata Shrubland >1m, Melaleuca Prostanthera calycina, Pterostylis cucullata, Pterostylis lanceolata Woodland, Olearia axillaris Shrubland >1m arenicola, Pterostylis tenuissima, Ptilotus beckerianus, Senecio macrocarpus, Stackhousia annua, Taraxacum cygnorum, Thelymitra epipactoides

New South Coastal Headland Scrub, Dry Dune Shrub Forest, Acacia georgensis, Chamaesyce psamogeton, Correa Wales Estuarine Wetland Scrub, Coastal Foothills Dry Shrub baeuerlenii, Dracophyllum oceanicum, Pultenaea pedunculata, Forest Prostanthera densa, Senecio spathulata var. attenuatus

Communities: Bangalay Sand Forest, Eastern Suburbs Banksia scrub, Swamp Oak Floodplain Forest, Themeda Grassland (sea- cliffs and coastal headlands)

Western Eucalyptus diversicolor Tall Forest, Eucalyptus calophylla Asplenium obtusatum subsp. northlandicum, Caladenia Australia Forest-Woodland, Agonis flexuosa Woodland, Eucalyptus busselliana, Caladenia caesarea subsp. maritima, Caladenia marginata Woodland, Dryandra Open Heath, Dense procera, Caladenia viridescens, Caladenia procera, Caladenia Coastal Heath, Eucalyptus gomphocephala Woodland, viridescens, Chorizema varium, Wurmbea calcicola Eucalyptus patens Woodland, Melaleuca thicket Communities: Agonis flexuosa Woodland

Tasmania Dry coastal vegetation, Allocasurina forest, Acacia Brachyloma depressum, Lasiopetalum baueri, Pomaderris longifolia scrub, Melaleuca ericifolia scrub, Stabilised oraria, Veronica plebia, Zieria littoralis dune communities, Coastal Myoporum insulare scrub

Status of Polygala

• currently undeclared any where in Australia • NWR Phase 3b - two Vic CMAs recommended regionally controlled - all other Vic CMAs recommended restricted • 3b Proposal - restricted in Victoria - abandoned • South Australia – listing proposed • control implications Management options

• hand pulling • herbicides • fire • biological • integrated control Hand pulling – difficult for large plants, tedious for small plants – effective – new incursions, isolated plants, sterile plants

Herbicides – susceptible – cut and paint – metsulfuron methyl is effective, high non-target impacts – glyphosate is effective on both seedlings and mature plants – picloram-based gel on freshly cut stumps is effective

Fire – fire-sensitive, high intensity fires are required – suppression of hot summer fires - contributed to expansion – canopy scorch needed

Biological – low herbivory in Australia – potential realised in 1996 – published 2010 (www.australisbiological.com.au/publications) – strong field of candidates Biological control

Duffoemyidia barkeri (longicorn beetle)

Diaphorina petteyi (flower psyllid) Cossidae red crown-moth Lasioptera sp. (shoot gall midge) A C D

Agromyzid (leaf miner)

Aceria myrtifoliae (gall mite)

Uredo polygaliae (rust fungus) Seedling regeneration following removal of adults

Control options:

Hand-pulling? Herbicides? Fire? Natural events?

Point Nepean, Victoria Predicting control scenarios

• modelling - help select best control methods • population responses • need data - construct population models • knowledge gaps – many • e.g. seed longevity, soil-seed bank, seedling behaviour • DPI/CGG/CfOC partnership – 2011 • useful for implementation Buckley Reserve project a study designed to address knowledge gaps

Buckley Reserve, Ocean Grove

Buckley Reserve, Ocean Grove What data was collected

• plant densities • age profiles • phenology • seed bank behaviour • seedling survivorship • seed fall • post dispersal seed losses • cleared vs canopy responses uncleared uncleared

cleared cleared Population structure – Ocean Grove 2010

120

100 infertile

80 fertile

60

Number of plants of Number 40

20

0 < 1mm1 1-5 2mm 6 -103 mm 11- 450 mm >505 mm [ Figure 8. Numberof Polygala myrtifolia seeds in soil from cleared (triangle) and uncleared (circle) plots. Data are mean number of seeds 950 cm-3

20

18

16

14

12

10

8 Number of seeds Number 6

4

2 • persistent seed bank

0 • clearing triggers seed bank depletion 09/2009 12/2009 03/2010 06/2010 09/2010 12/2010 03/2011 06/2011

Date (month/year) P. myrtifolia seedlings on cleared and control plots. Bar colours represent age cohorts over four seasons.

1400 • high emergence on cleared plots 1200 • large losses over summer 1000 • Most germinate cleared 800 winter spring • higher survivorship 600 under canopies

400 Number of seedlings of Number

200

0 1 2 3 4 200 09/09 12/09 03/10 06/10

180

160

140

120 canopy 100 80

60

40 Number of seedlings of Number

20

0 1 2 3 4 09/09 12/09 03/10 06/10 Accumulative germination of fresh P. myrtifolia seed

80

70

60

50

40

30 % germination % 20

10

0 0 10 20 30 40 50 60 70 80 90 100

Days • high levels of viability in fresh seed • germinates rapidly under ideal conditions • very little innate dormancy Seed collection trap

Rat exclusion fence High levels of seed mortality eliasome occur after seed fall

pigeons (turtle doves, rats)

Ants carry seeds short distances and cache seeds

High level of fertility

Most flowers are fertile and produce 2 seeds Flowering/fruiting is very seasonal 40 buds flowers fruit 35

30 • strong 25 seasonality 20 • short peak

15 period % Frequency % 10

5

0 06/09 07/09 09/09 10/09 11/09 12/09 01/10 02/10 03/10 04/10 05/10 06/10 200

180 flowers 160

140 • low seed 120 output 100 seeds (in traps) • high flower

80 failure (verify) fruit

60 Total number Total 40

20

0 06/09 07/09 09/09 10/09 11/09 12/09 01/10 02/10 03/10 04/10 05/10 06/10 Modelling

• Quick method - comparing different control options • 6 months to develop • Can run for any control method - press of a button • Cut down years of trialling each control method • Technology is available for realistic model (Stella software) • Can test how well the model mimics real life • Availability of information of Polygala population dynamics model of a plant life cycle

seeds move in fresh seeds seeds in soil ac cumilation seed ageing

germination1 germination2 seedlings

seedling death growth1

juveniles juvenile death

growth2

adults

adult death month pre predation rate post predation rate

max threshold1

seed predation seed death final s eed

seed fresh soil bank seeds seed ac cumulate seeds move in seeds c arrying capacity

month canopy seed viability

seed predation adult ev aluater Cotyledonous seedlings fire effect2

initial adults canopy germination 2 germination1 summer stress

summer s tres s seedling thinning grow th Gra ph 1 m o nt h growth + p re p re da ti on r at e fire effect1 initial adults fire effect1 s ee d lin g sm o v in g p os t pr ed a tio n ra te T ab le 1 seedling controller m a x th re sh o ld 1

s ee d p re da ti on s ee d d ea th fi na l se e d o ut pu ts s ee d adult ev aluater fr es h s o il ba n k s e ed s Adu lt s ee d a cc u m ul at e + + s ee d s m ov e in s ee d s ca rr yi ng c a pa c ity Total Seedlings T ot al J uv e ni le s

c an o py seed viabi lity m o nt h

+

T ot al p op u la tio n s ee d p re da ti on a du lt e va lu a te r Cot yl ed o no u s s e ed li ng s fi re e ffe c t2

initial adul ts c an o py germination 2 germination1 s um m e r st re ss fi re e ffe c t

fi re e ffe c t3 s um m e r st re ss seedling thinning

g ro wth fi re s ta tu s g ro wth fi re tim e fi re e ffe c t1 initial adul ts fi re e ffe c t1 seedling control ler

a du lt e va lu a te r fi re e ffe c t2 fi re e ffe c t1

See d lin g s s ee d lin g b a nk seedling accumul ation seedling death2 s d yr 1 s d yr 3 s d yr 2 s d yr 6 s d yr 5 s d yr 4 seedlingdeath m2 m1 m3 m6 m4 m5 d4 d1 fi re e ffe c t1 d6 d2 d5 d3 juvenil e death2

fi re e ffe c t1 J uv e ni le b an k juvenil e accumulati on

ju y r9 ju y r1 0 s d yr 7 s d yr 8 ju v en il e ju y r6 ju y r8 s d yr 9 s d yr 10 ju y r1 ju y r3 ju y r5 s d yr 11 ju y r2 ju y r4 g9 g7 jd 1 0 m9 g6 m7 m 1 1 m 1 2 g3 g5 g8 m8 m 1 0 g1 g2 g4 d7 d8 d9 d 10 jd 5 jd 6 jd 7 jd 8 jd 9 d 11 jd 1 jg 2 jd 3 ju v en il e g 10 d e at h jd 1 8 jd 1 7 jd 2 2 jd 2 1 jd 1 4

ju y r1 4 ju y r1 3 ju y r1 2 ju y r1 1 ju y r1 8 ju y r1 7 ju y r2 1 jd 1 1 fi re e ffe c t1 g 13 g 11 init adults g 17 g 12 Adu lt fi re e ffe c t1 ju y r2 2 jd 1 3 g 21 jd 1 2

g 14

g 18 g 16 g 20

initial adul ts s el f g 22 fi re e ffe c t3 th in n in g ju y r2 0 ju y r1 9 ju y r1 6 ju y r1 5

g 19 g 15 Adu lt ju y r2 3 fi re e ffe c t1 a du lt e va lu a te r g 24 jd 1 6 jd 1 5 jd 2 0 jd 1 9

BC ef fe ct s ee d jd 2 3 p ro d uc ti on c an o py s ta tu s 1

fi na l se e d Adu lt a du lt a gi ng fi re e ffe c t1

fi re e ffe c t2 m o nt h m a x se e d c an o py p ro du c tio n

fi re e ffe c t3 m a x se e dl in gs s ee d m o ve o u t

fr es h s e ed s Adu lt

s ee d m o ve o u t

seedling control ler What does the model do?

We can compare the effects of different control methods e.g. biocontrol by seed predator Scenario 1 - 90% of seeds eaten Scenario 2 - 50% of seeds eaten

See model details @ (http://forio.com/simulate/ratnapriya.gajaweera/polyga la-hispeed/run/) Model Results

control Cossidae

Cossidae + 50% seed Cossidae + 90% seed fire

fire + 90% seed removed Key Points

. high post dispersal seed mortality – pigeons and rodents . seed banks are short-lived (5-10 yrs) . soil pathogens and drought – kill seedlings . high seedling mortality (summer) . rapid seed bank depletion when canopy is opened . 2 yrs to sexual maturity . plants are susceptible to herbicides . control activities easier after natural thinning . a combination of control options is better than 1 alone

Acknowledgements

City of Greater Geelong Caring for our Country Program Barwon Coast Committee of Management Bev Wood Troy Gallus Glenelg-Hopkins CMA