WEED RISK ASSESSMENT FOR THE -WAIPORI COMPLEX,

JULY 2009

Report No. 2182

Prepared for:

LAKES WAIHOLA WAIPORI SOCIETY INC. C/- PAULINE BACON 26 NORE STREET WAIHOLA 9073

WILDLAND CONSULTANTS LTD, OFFICE, 764 CUMBERLAND ST, DUNEDIN Ph 03-477-2096, Fax 03-477-2095, email [email protected], website www.wildlands.co.nz 11 Contract Report No. 2182 HEAD OFFICE: 99 SALA STREET,© P.O. 2009 BOX 7137, TE NGAE, ROTORUA. Ph 07-343-9017; Fax 07-343-9018

CONTENTS

1. INTRODUCTION 1 2. METHODS 2 3. ECOLOGICAL CONTEXT 3 3.1 Hydrology 3 3.2 Ecological significance 3 4. LAND TENURE 4 5. VEGETATION TRENDS 5 6. THREATS TO INDIGENOUS VEGETATION AND HABITATS 6 6.1 Indigenous vegetation 6 6.2 Nationally uncommon and regionally important plant species 7 6.3 Fauna 10 7. RECREATIONAL AND CULTURALLY IMPORTANT SITES 10 8. STAKEHOLDER INVOLVEMENT 10 9. WEED CONTROL STRATEGY 11 9.1 The need for weed control 11 9.2 Limitations of current weed control 11 9.3 Weed control in outlying areas 11 9.4 Timing of control 12 9.5 Where should control start? 13 9.6 Control methods 13 9.6.1 Aerially-applied herbicide spray 13 9.6.2 Ground- and water-based control 14 9.6.3 Use of volunteers 14 10. WEED CONTROL PRIORITIES 15 10.1 Pest management strategy for Otago 2008 15 10.2 Setting priorities 15 11. PRIORITY 1 WEEDS 17 11.1 Alder ( Alnus glutinosa ) 17 11.2 Crack willow ( Salix fragilis ) 19 11.3 Grey willow ( Salix cinerea ) 20 11.4 Reed sweet grass ( Glyceria maxima ) 21 12. PRIORITY 2 WEEDS 22 12.1 Elder ( Sambucus nigra ) 22 12.2 Red alder ( Alnus rubra ) 22 12.3 Wilding conifers 22 12.4 Scotch broom ( Cytisus scoparius ) and gorse ( Ulex europaeus ) 23 12.5 Montpellier broom ( Teline monspessulana ) 24 12.6 Silver birch ( Betula pendula ) 24

22 © 2009 Contract Report No. 2182

13. PRIORITY 3 WEEDS 24 14. ESTABLISHMENT OF INDIGENOUS VEGETATION 25 14.1 General 25 14.2 Raupo 25 15. RECORD KEEPING, MONITORING, AND RESEARCH 26 15.1 Records 26 15.2 Monitoring 26 15.3 Research 27 16. OTHER ISSUES 27 16.1 Dead trees blocking waterways 27 16.2 Financial resources 27 16.3 Grazing 28 16.4 Dumping of rubbish 28 17. CONCLUSIONS 28 18. RECOMMENDATIONS 28 ACKNOWLEDGMENTS 29 REFERENCES 30

APPENDIX 1. Weed distribution maps 31

PROJECT TEAM

Steve Rate - Vegetation survey, report compilation. Kelvin Lloyd - Internal peer review.

Reviewed and approved for release by:

______W.B. Shaw Director/Principal Ecologist Wildland Consultants Ltd

 Wildland Consultants Ltd 2009

This report has been produced by Wildland Consultants Ltd for Lakes Waihola Waipori Wetlands Society Inc. All copyright in this report is the property of Wildland Consultants Ltd and any unauthorised publication, reproduction, or adaptation of this report is a breach of that copyright.

33 © 2009 Contract Report No. 2182

1. INTRODUCTION

The Lakes Waihola Waipori Wetlands Society Inc. was formed to promote the sustainable management and enhancement of the Waihola Waipori wetland complex on the near Waihola, Otago. This wetland complex is nationally and internationally significant for its indigenous habitat and wildlife values, fishery and hunting values, and cultural and recreational values. The wetlands are also very significant to local M āori.

One of the key issues identified by the community and stakeholders at a planning meeting in November 2007 was the need to control invasive weeds in the wetlands. In particular, crack willow ( Salix fragilis ), and reed sweetgrass ( Glyceria maxima ) have become major invasive weeds (crack willow for the last 60 years), and there are new threats from species such as alder ( Alnus glutinosa ) and grey willow ( Salix cinerea ).

Control of crack willow has been undertaken by the Otago Regional Council, Department of Conservation, and Fish and Game over the last eight years, but the approach to management of weeds in the wetlands, and protection of biodiversity values has been piecemeal. This is, in part, due to the large size and complex nature of this tidal lowland wetland, and also due to the number of landowners within and adjacent to the wetland and the many different community interests. This has made it difficult to plan for and implement comprehensive weed control within the wetlands.

The Lakes Waihola Waipori Wetlands Society held a recent weed and native vegetation workshop that brought the community together and established a community mandate with the following goal for weed management and indigenous vegetation and wildlife enhancement in the Waihola Waipori wetland complex:

“A wetland system where weeds are being actively controlled according to a plan in order to maintain and enhance the diversity of habitats including a balance of open water and land, for the benefit of native plants, fish, birds and other fauna and for recreational purposes”.

The Society intends to apply for funding to assess, plan, and undertake weed control in the wetland complex. The first task is to undertake a comprehensive weed risk assessment. This report comprises the weed risk assessment, and addresses the following issues:

• Identification of significant weed species in the wetlands that could potentially invade habitats within the project site, either in its present or restored state, and descriptions of the ways in which these species are spread. This report specifically addresses crack willow, grey willow, reed sweetgrass, and alder, but also addresses other weeds that are potentially of concern. • Sources of weed invasion in the vicinity of the project site have been identified. • Zones around the project site that are susceptible to weed invasion are identified. • The distributions of ecological weed species have been described and mapped. • The threats posed by each weed species have been identified, and the potential for eradication or control of these species is assessed.

1 © 2009 1 Contract Report No. 2182 1

• Existing weed sources and areas that are at risk of weed invasion are mapped and/or identified. • Significant ecological features within or near the areas potentially affected by weeds are identified. • Threatened indigenous plants that need to be protected when undertaking weed control, or may be under threat from the spread of weeds, are identified. • The direct and indirect effects of weeds and weed control programmes on indigenous vegetation, habitats, and species are evaluated. • The feasibility of eradicating or controlling the weeds identified is assessed, including follow-up control methods and long-term control. • The potential for weed suppression by establishing indigenous vegetation is assessed. • Control methods are evaluated for potential use in sites of important cultural and cultural harvesting significance (i.e. flax and raupo harvesting by local iwi) and in fishing, hunting, and recreational boating and activity sites.

This report also provides recommendations on: • Whether the proposed project should proceed. • Appropriate timing for the proposed weed control project, taking into consideration the need to eradicate weed sources. • Whether weed control or eradication work should be undertaken within a weed dispersal (buffer) zone and whether this needs to be done before or after other vegetation restoration work.

2. METHODS

Relevant information on ecological values of the Waipori/Waihola wetland complex and adjacent areas was reviewed. This included previously commissioned botanical and hydrological reports. In most cases these covered only parts of the wetland complex.

Stakeholders (Otago Regional Council, Department of Conservation, Fish & Game, local iwi, and other members of the community) were consulted to establish existing knowledge on weeds and potentially affected values.

Field surveys were undertaken during April and June 2009, within the study area marked in Figure 1. The study area conformed to ‘natural’ boundaries along much of its circumference: SH1, Titri Road, Berwick Road, and stop banks. Field work was challenging due the large size of the wetland and the difficulty of traversing dense wetland vegetation and deep drainage ditches on foot. Sites were observed from a boat, traversed on foot, or viewed from a distance using binoculars. Particular attention was paid to the distribution of invasive weed species. The location and identity of threatened and uncommon indigenous plant species was recorded when they were encountered. An attempt was made to visit a representative range of vegetation/habitat types and infestations of each major weed species on the ground.

Weed distributions and important areas of indigenous vegetation and habitat within the Waihola/Waipori wetland complex were mapped using field observations, aerial

2 © 2009 2 Contract Report No. 2182 2

photographs, and photographs taken from the ground and during a helicopter flight. The distribution of potentially invasive weeds in the surrounding landscape was also noted. An assessment of the risks and practicality of weed control was based on field observations, literature review, and stakeholder consultation.

There are several limitations in this current study. Not all sites could be visited and it was difficult to determine some vegetation types from aerial photographs or from a distance. This has resulted in some uncertainty regarding the distributions of weeds and the distributions of uncommon species and vegetation types. These uncertainties are identified in the text and maps contained within this report.

3. ECOLOGICAL CONTEXT

3.1 Hydrology

The major hydrological input into is from the north via the , but smaller flows enter from the Meggat Burn, Taieri Main Drain (Contour Channel), and several small streams. The Waipori River also drains Lake Waipori at its southern end, flowing east before joining the c.10 km from the coast. is drained by the Waihola River, which joins the Waipori River within the wetland complex, between Lake Waipori and the Taieri River. Tidal flows travel up the Taieri River and enter the wetland complex, raising lake levels c.0.5 m at high tide. Lake Waihola has a smaller non-tidal catchment than Lake Waipori, resulting in a much lower mean hydraulic residence time (153 days, compared to 1.9 days for Lake Waipori) (Schallenberg and Burns 2003). Both lakes are shallow: Lake Waipori averages c.1 m deep and Lake Waihola c.2 m deep (Schallenberg and Burns 2003). Predominant land use surrounding the wetland complex is farming (including dairying) and exotic plantation forestry. Water flows bring a significant amount of sediment into the wetland system, resulting in gradual infilling, and there is also input of nutrients from farm run-off.

3.2 Ecological significance

The Waihola/Waipori wetland complex is located within Tokomairiro Ecological District (ED) in the Otago Coast Ecological Region. The complex contains the best examples of lowland wetlands in the Otago Coast Ecological Region (Tangney 1987).

The wetland complex is classified as nationally and internationally significant for its bird habitat (large numbers of waterfowl, threatened species including Australasian bittern Botaurus poiciloptilus , marsh crake Porzana pusilla affinis , spotless crake Porzana tabuensis plumbea , banded rail Gallirallus philippensis assimilis , and fernbird Bowdleria punctata punctata ), fish habitat (unique species assemblage, threatened species including giant kokopu Galaxias argenteus ), botanical values (wetland communities, threatened species), recreational values, and cultural values (Cromarty and Scott 1996, Tangney 1987).

3 © 2009 3 Contract Report No. 2182 3

4. LAND TENURE

Land tenure is a combination of privately-owned land, and land administered by agencies, including Council, Otago Regional Council, Dunedin City Council, Department of Conservation, LINZ, QEII National Trust, and NZ Railways (The Property Group 2007). Conservation units are shown in Figure 1 and are described in Table 1.

1 2 11

9 1

3, 10

4, 8 5 7 6

Figure 1: Conservation units and covenants within the study area (dotted line). Numbers refer to details provided in Table 1.

Table 1: Conservation units and covenants within the Waihola-Waipori study area.

Area Map Number Name Legal Description (ha) 1 H45047 Lake Waipori Wildlife Government Purpose Reserve - 63.73 Management Reserve s.22 Reserves Act 1977 2 H45048 Waipori River Fixed Marginal Strip - s.24(3) 5.00 Conservation Act 1987 3 H45050 Waipori/Waihola Wildlife Local Purpose Reserve - 4.00 Management Reserve s.23 Reserves Act 1977 4 H45054 Lake Waihola Wildlife Government Purpose Reserve - 8.50 Management Reserve s.22 Reserves Act 1977 5 H45057 Waihola Recreation Recreation Reserve - s.17 8.87 Reserve Reserves Act 1977 6 H45060 Waihola Hall Local Purpose Reserve - 0.12 s.23 Reserves Act 1977

4 © 2009 4 Contract Report No. 2182 4

Area Map Number Name Legal Description (ha) 7 H45061 Greenwich Recreation Recreation Reserve - s.17 0.34 Reserve Reserves Act 1977 8 H45084 Lake Waihola-Gillanders Government Purpose Reserve - 26.90 Wildlife Management s.22 Reserves Act 1977 Reserve 9 H45100 Waipori/Waihola Wildlife Government Purpose Reserve - 118.64 Management Reserve s.22 Reserves Act 1977 10 H45106 Public access to Easements that are not part of 0 Waihola/Waipori Wildlife another Conservation Unit Management Reserve 11 5/12/005 Not available QEII open space covenant 325.72

5. VEGETATION TRENDS

The vegetation of the wetland complex has been described previously by Ward and Munro (1985), Tangney (1987), Peach and Johnson (2004), and Johnson (2004). Vegetation/habitat types described by these authors included Carex -dominated sedgelands, raupo reedlands, reed sweet grass reedlands, open water and aquatic vegetation, crack willow forest, kanuka-broadleaf forest, gorse scrub, broom scrub, mudflats, rushlands, Coprosma propinqua scrub, pasture, and lake-edge turf. The same habitats are present today, but from comparison of Tangney’s (1987) maps and vegetation type descriptions with Appendix 2 in Johnson (2004) and the recent field survey, several trends in the distribution and abundance of ecological weeds and some indigenous species are apparent.

Several species have increased in abundance:

• Crack willow (in areas where it has not been controlled) - e.g. c.10 times increase in cover in DOC-administered Lake Waihola-Gillanders Wildlife Management Reserve.

• Alder - rare in 1987, occasional on levees and swamps 2004, large infestation in 2009;

• Silver birch - not mentioned in 1987, rare on levees and in swamps 2004 and 2009;

• Blackberry - not listed in 1987 (listed as rare on levees in 2004). Now common on eastern side of Lake Waipori and occasional at a few other sites (e.g. Duck Consortium property, southern margin of Lake Waihola);

• Grey willow - occasional in sedgeland in 1987, frequent in swamps and on levees 2004 and 2009;

• Pinus spp. - rare in kanuka-broadleaf forest and single tree in back swamp in 1987, rare on islands 2004, scattered on levees and in back swamp in 2009;

• Coprosma propinqua - large parts of areas described as sedgeland in 1987 in the DOC-administered Lake Waihola-Gillanders Wildlife Management Reserve are now Coprosma propinqua -flax shrubland;

5 © 2009 5 Contract Report No. 2182 5

• Harakeke (flax) - up to half the area described as sedgeland in 1987 in the Duck Consortium property is now harakeke flaxland;

• Raupo - many raupo-dominated areas in 1987 have increased in size (e.g. on the south-west margin of Lake Waipori and in many ponds), and raupo has invaded other vegetation types (e.g. large parts of Nevill’s Pond that were sedgeland in 1987 are now raupo reedland). However, raupo cover has remained relatively stable in some areas such as the southern end of Lake Waihola.

At least one species has decreased in abundance:

• Crack willow (at the scale of the whole wetland) - due to herbicide treatment.

6. THREATS TO INDIGENOUS VEGETATION AND HABITATS

6.1 Indigenous vegetation

The greatest threat to indigenous vegetation within the wetland complex is that posed by weed invasion. Aerial spraying of willow regrowth also poses a secondary threat to indigenous vegetation, although the long term consequences of not spraying far outweigh this. It was evident during the field survey that where willow canopy cover was low, non-target species were killed as a result of aerially-applied herbicide. In a field inspection of re-sprayed crack willow on the margins of Lake Waipori, commonly affected species included Carex secta , flax, Coprosma propinqua , and reed sweet grass. Some indigenous species were only partially defoliated (e.g. C. propinqua ) or had repsrouted after total defoliation (e.g. cabbage tree). The At Risk species Urtica linearifolia was common amongst re-sprayed crack willow on the northern and eastern margins of Lake Waipori, indicating that it is either tolerant of spraying, regrows rapidly after initial defoliation, or recolonises the site through water-dispersed seed. Exotic species such as cocksfoot ( Dactylis glomeratus ), Yorkshire fog ( Holcus lanatus ), curled dock ( Rumex crispus ), and spearwort (Ranunculus flammula ), and a native groundsel ( Senecio minimus ), dominated the groundcover amongst dead crack willow and other dead vegetation.

A potential effect of killing indigenous groundcover species is increased susceptibility to weed invasion. However, weedy species similar to those currently present may have also been common prior to weed control. It is likely that indigenous species such as Carex secta , Coprosma propinqua , and flax will recover from herbicide damage at these sites, but that weedy species such as blackberry and reed sweet grass will persist without targeted control. Monitoring of the vegetation following weed control would provide quantitative data on species composition and enable successional patterns to be determined. Replanting to establish a dense cover of indigenous vegetation would reduce the potential for reinvasion of weeds into previously sprayed areas, but is probably only practical on levees and wetland margins.

6 © 2009 6 Contract Report No. 2182 6

6.2 Nationally uncommon and regionally important plant species

The national lists of threatened and uncommon plant species were recently revised by de Lange et al . (2009). One threatened, and several uncommon, species occur within the wetland complex (Table 2). The risk to these species of not undertaking weed control differs between species, although it is generally low.

Pygmy clubrush (Isolepis basilaris ) is listed as ‘Threatened - Nationally Endangered’ in de Lange et al . (2009). This species has been recorded in shallow pools within Titri wetland, about 200 m north of the footbridge (Graeme Loh, Department of Conservation, pers. comm. July 2009; Table 3). Woody weeds tend to be at low abundance in this area and therefore do not currently pose a risk to this species.

Urtica linearifolia (At Risk-Declining) was recorded at several locations during this survey (Table 4, Appendix 1). This species was particularly common on the eastern margins of Lake Waipori, although it is probably more widely distributed within the wetland than these records indicate, because a large portion of the wetland was not searched due to poor access. Urtica linearifolia is capable of persisting under a willow canopy. At several locations it appears to have survived aerial spraying and re-spraying of crack willow. Although "some large populations in the Horowhenua have been destroyed ...through willow control" (NZPCN 2009), the Waihola/Waipori wetland complex most likely supports a well-distributed population of U. linearifolia , which would assist recolonisation in the event of any losses caused by aerial spraying.

Gossamer grass ( Anemanthele lessoniana ) has been planted on Ram Island. This species is listed as ‘At Risk-Declining’, and has few natural populations locally, but is a common garden plant in coastal Otago.

Three other ‘At Risk-Declining’ plant species ( Deschampsia cespitosa , Olearia lineata , Ranunculus limosella ) and two ‘At Risk-Naturally Uncommon’ species (Lepilaena bilocularis , Stuckenia pectinata ) have been previously recorded in the wetland complex (Johnson 2004, Table 2). Three of these are aquatic and unlikely to be threatened by weed encroachment or weed control operations. Deschampsia cespoitosa is described as “occasional in swamps at ” by Johnson (2004). It is unlikely to be threatened by willow control operations or woody weed encroachment because willow cover is generally low in this area. Olearia lineata was recorded on a levee in the wetlands in 1987, but the location was not recorded and it is not known whether this species is still present. If it has survived aerial herbicide spraying, it probably remains very rare at this site.

The Eleocharis sphacelata reedland recorded at E2283805 N5466145 (Appendix 1) is the only known record of this species for coastal Otago (Peter Johnson, pers. comm. 12 May 2009). Crack willow is present on the small ‘island’ it grows around and is not currently threatening the population, although this may change in the long term. Any willow control should be ground-based to avoid potential loss of this habitat/vegetation type. The current landowners have sprayed herbicide on raupo at one end of the island so that it does not encroach on the Eleocharis reedland.

7 © 2009 7 Contract Report No. 2182 7

Table 2: At Risk and uncommon plant species recorded in the Waihola/Waipori wetland complex. *Recorded prior to the current survey (Johnson 2004).

National Threat Species Common Name Abundance/Habitat Classification Anemanthele Gossamer grass Rare/Ram Island At Risk-Declining lessoniana (planted) Deschampsia Tufted hair- Occasional/Swamp At Risk-Declining cespitosa* grass Eleocharis sphacelata Tall spike sedge Rare/Aquatic Not Threatened 1 Isolepis basilaris Pygmy clubrush Threatened - Nationally Endangered Lepilaena bilocularis* Occasional/Aquatic At Risk-Naturally Uncommon Olearia lineata* Rare/Levee At Risk-Declining Ranunculus limosella* Rare/Aquatic At Risk-Declining Stuckenia pectinata* Fennel-leaved Abundant/Aquatic At Risk-Naturally pondweed Uncommon Urtica linearifolia Swamp nettle Rare/Levee and At Risk-Declining swamp 1Regionally uncommon

Table 3: Records for pygmy clubrush ( Isolepis basilaris ) within the Waihola/Waipori wetland complex (from Graeme Loh, Department of Conservation).

GPS coordinates Habitat Date of record E2286361 N5463631 Shallow pool 15 April 2003 E2286331 N5463653 Shallow pool 15 April 2003 E2286319 N5463597 Shallow pool 15 April 2003 E2286338 N5463671 Shallow pool 15 April 2003 E2286330 N5463716 Shallow pool 15 April 2003 E2286327 N5463788 Shallow pool 15 April 2003 E2286503 N5464139 Shallow pool 15 April 2003

Table 4: Location details for swamp nettle ( Urtica linearifolia ) recorded in the Waihola/Waipori wetland complex during the current survey and by Graeme Loh in 2003.

GPS coordinates Habitat Abundance E2284005 N5466330 Levee under dead willow Approximate mid-point of two specimens E2286795 N5466317 Swamp (on Carex secta ) One specimen E2286754 N5466377 Swamp under dead willow 10+ plants E2286259 N5467846 Swamp under dead willow A few plants E2286749 N5467270 Swamp under dead willow 10+ plants E2286707 N5463412 n/a n/a (recorded 15 April 2003) E2286709 N5463416 n/a n/a (recorded 15 April 2003) E2286722 N5463428 n/a n/a (recorded 15 April 2003) E2286700 N5463400 n/a n/a (recorded 15 April 2003)

8 © 2009 8 Contract Report No. 2182 8

Immature specimens of kahikatea (Dacrycarpus dacrydioides ) and pokaka (Elaeocarpus hookerianus ), which are uncommon species in Coastal Otago, have been planted on Ram Island.

Indigenous vegetation types that are uncommon within the wetland complex include a small area of kowhai ( Sophora microphylla )) forest at the northern end of Ram Island, kanuka ( Kunzea ericoides ) forest on Ram Island and Lonely Island, and turf vegetation. None of these vegetation types appear to be threatened by aerial weed control operations. Turfs, which were previously maintained by grazing and ploughing and provided habitat for wading birds such as banded dotterel ( Charadrius bicinctus bicinctus ; Threatened – Nationally Vulnerable), are now being encroached by taller mostly exotic herbs (Graeme Loh, Department of Conservation, pers. comm. July 2009). If these habitats are to be retained, specific management will be required. Ground-based weed control and indigenous plantings on Ram Island and Lonely Island should ensure the persistence of existing indigenous dryland habitats there, although ongoing control of gorse, Scotch broom, and elder would be desirable.

Narrow-leaved lacebark ( Hoheria angustifolia ) is listed as an ‘Important native tree species present on land within Dunedin City’ in Appendix 16B of the Dunedin City District Plan (Dunedin City Council 2006). This species was recorded amongst dead willow c.60 m to the NW of E2287264 N5466931 (three specimens sighted) and c. 20 m to the NW of E2287143 N5466823 ( c.10 specimens), and may be at risk from follow-up aerial spraying of willow now that willow canopy cover is low. Two other species present in the wetland complex are also listed as ‘threatened’ or ‘important’ in the Dunedin City Plan ( Ileostylus micranthus , Sophora microphylla ), but they were recorded just outside the city boundary on Ram Island.

Plate 5: Swamp nettle growing through Carex secta on a levee (left) and Eleocharis sphacelata reedland in Nevill’s Pond (outlined in yellow, right and inset).

9 © 2009 9 Contract Report No. 2182 9

6.3 Fauna

There are not expected to be any threats to fauna as long as all herbicides are registered for use over water and used at the recommended application rates. South Island fernbird (Bowdleria punctata punctata ) were heard frequently during the survey and appear to be unaffected by current weed control operations.

7. RECREATIONAL AND CULTURALLY IMPORTANT SITES

Current recreational activities within the wetland include fishing, boating, and hunting (duck shooting). Weeds threaten these activities by encroaching on access ramps (raupo), ponds, and channels (crack willow, reed sweet grass), accelerating siltation (crack willow), removing fish spawning habitat, and destroying waterbird roosting areas (reed sweet grass). Controlling exotic weed species will improve recreational opportunities within the wetland. However, raupo should not be subject to widespread control as it provides important habitat for indigenous fauna and is able to inhibit reed sweet grass establishment. Recreational activities may enhance the spread of weeds within the wetland, but this is unlikely to be more than the intrinsic dispersal capability if the usual precautions are taken. Planting of crack willow to provide cover for maimais is no longer prevalent (Ian Haland, Fish & Game, pers. comm. 19 June 2009).

Culturally significant sites include those used for harvesting of flax and raupo by local iwi. Both these species appear to be increasing in abundance within the wetland and are unlikely to be threatened by weed control. These species are likely to further increase in abundance with the removal of crack willow. Culturally important indigenous fauna such as tuna (eels) are unlikely to be threatened by weed control.

8. STAKEHOLDER INVOLVEMENT

Involvement of all landowners in the weed control scheme is essential to its long term success. It may be possible to limit spread of weeds from non-participating landowners if channel boundaries are sprayed, although this is a second-best option. Many weed species such as crack willow are present in farmland surrounding the wetland complex, and these could act as sources of invasion in some cases. The Society has been very active in consulting with landowners and this should continue. A ‘bottom-up’ approach will be important for the long-term sustainability of any weed control programme.

Community involvement and backing will be required for project success. The society can be commended for the inclusion of representatives from many interest groups including councils, Department of Conservation, Fish and Game, and local property owners. A challenge for the society is to remain focussed on weed control in the long term and not be distracted by other issues such as siltation, which is an ongoing natural process. In the long term, control of woody weeds will help to solve issues such as dead trees in waterways.

10 © 2009 10 Contract Report No. 218210

9. WEED CONTROL STRATEGY

9.1 The need for weed control

Weeds within the wetland complex pose a threat to indigenous species, communities, and ecological processes. Left unchecked, the vegetation will become dominated by exotic woody weeds such as crack willow, grey willow, and alder. Vegetation dominated by indigenous species will become less common, some indigenous species may become locally extinct, and hydrological patterns will change. Weed control is therefore essential to protect the ecological values of this internationally significant wetland complex.

9.2 Limitations of current weed control

Previous and current control of willow has concentrated on the margins of main waterways and then a progressive spraying of all crack willow infestations starting at the northern end of Lake Waipori. It has successfully reduced the extent of crack willow infestations within the wetland complex. Larger patches of grey willow have also been sprayed from the air. Alder has not yet been targeted except for an unsuccessful test of the efficacy of aerially-applied glyphosate herbicide. Reed sweet grass has had piecemeal control (cutting, spraying) by local property managers/ custodians. Current weed control efforts have the following limitations:

• It has been piecemeal and lacks coordination across landholder boundaries. • Some highly invasive weeds are not targeted (e.g. alder). • Some weeds continue to spread within the wetland (e.g. grey willow). • There is regrowth of previously-sprayed crack willow. • Where canopy cover is low, aerial spraying has a greater impact on non-target species. • There is likely to be continued weed invasion from surrounding areas not subject to control measures.

9.3 Weed control in outlying areas

Many weed species within the wetland can disperse propagules by water. Water flows into the wetland from both upstream and downstream (tidal) sources. As a consequence, re-colonisation by water could occur from the Taieri River, Waipori River, Contour Channel, Meggat Burn, and many local streams (Figure 2). Crack willow and reed sweet grass are two of the weed species that are likely to re-invade the wetland this way. However, there was little evidence seen during the current survey of recent invasion by crack willow. All willow growth on channel margins appeared to be regrowth of sprayed trees. Ideally, major weed species such as crack willow and reed sweet grass would be controlled in the entire catchment, but this is not feasible given its size. It is therefore suggested that a buffer area around the wetland is established in which crack willow is controlled. The Waipori River, Contour Channel, and Meggat Burn have a relatively low abundance of crack willow and control of this species should therefore be feasible within theses areas. As it is not feasible to control crack willow in the entire Taieri River, it is recommended that a crack willow control buffer extends to the Waipori/Taieri confluence. The Taieri River below this confluence is likely to have a relatively high reinvasion risk due to

11 © 2009 11 Contract Report No. 218211

flood flows from the upper Taieri River bringing plant fragments downstream. Wetlands on the western side of Berwick Road in the southwest of the site should also be subject to crack willow control.

Wind-borne seed can also come from outlying areas such as plantation forestry estate and farmland. Invasion of wilding conifers is likely to be an ongoing problem, although currently ecological effects are minor. Some species such as silver birch can be easily controlled within the wetland due to their localised distribution, but reinvasion is likely. The prevailing wind direction is from the south west, and therefore most wind-dispersed seed is likely to establish on south-western margins of the wetland.

9.4 Timing of control

Summer-autumn is the best time to control weeds within the wetland complex, because deciduous species are in leaf and plants are actively growing which increases herbicide uptake. Crack willow should continue to be controlled in early autumn before leaf fall. However, in dry conditions incursion of salt water may occur further into the wetland than usual, causing the willows to shed their leaves early. Willows should not be sprayed at this time because the trees will be dormant (Pete Raal, Department of Conservation, pers. comm. July 2009). Reed sweet grass can also be sprayed in spring. Other species such as gorse and Scotch broom can be controlled at all times of year, depending on the herbicide or method used.

Key

Wind Water Control area

Figure 2: Major weed invasion agents and sources from surrounding areas and suggested control area to slow reinvasion of crack willow.

12 © 2009 12 Contract Report No. 218212

9.5 Where should control start?

The usual recommendation for controlling weeds within wetlands is to start at the top of the catchment and work down so that reinfestation is limited. However, the Waihola/Waipori wetland complex has inflows from both upstream and downstream sources, and weed control in the whole catchment is not feasible due to its size. In addition, some weed species are predominantly wind-dispersed rather than water- dispersed. Table 4 lists the recommended approach for controlling Priority 1 and 2 weed species (Section 10 describes the prioritisation of weeds).

Table 4: Control approach for Priority 1 and 2 weed species. All species will require follow-up control.

Species Approach Alder ( Alnus glutinosa ) Spray entire infestation in a single Sitka spruce ( Picea sitchensis ) operation Remove trees from wetland margins Crack willow ( Salix fragilis ) Ideally, spray entire wetland complex in one operation. Otherwise: 1. Margins of channels 2. Head of Lake Waipori, and the Taieri River (downstream of study area) 3. Move south from Lake Waipori 4. Buffer zone Reed sweet grass ( Glyceria maxima ) 1. Spray large infestations 2. Plant raupo in controlled sites Elder ( Sambucus nigra ) 1. Undertake indigenous tree plantings Scotch broom ( Cytisus scoparius ) when control results in large bare Gorse( Ulex europaeus ) areas 2. Continue control on islands 3. Progressively control on levees moving from northeast to southwest 4. Progressively control on escarpments surrounding the wetland Blackberry ( Rubus fruticosus agg.) Control at known sites Montpellier broom ( Teline monspessulana ) Grey willow ( Salix cinerea ) Start on north-eastern margins and work Wilding conifers systematically towards the southwest Silver birch ( Betula pendula ) Sweep western margins Red alder ( Alnus rubra )

9.6 Control methods

9.6.1 Aerially-applied herbicide spray

Aerial spraying is the most practical and economical way of controlling large infestations and/or widely distributed weeds that occur at low densities. To reduce adverse effects on non-target species, a spray gun rather than a spray boom should preferentially be used for small infestations.

13 © 2009 13 Contract Report No. 218213

Suitable Species for Aerial Control

• Crack willow (herbicide spray). • Grey willow (herbicide spray). • Alder (herbicide spray). • Wilding conifers (herbicide spray).

9.6.2 Ground- and water-based control

Travelling by foot within the wetland complex is very difficult. Ground-based control would therefore be limited to levees and other areas which can be walked or are accessible by boat. Basal bark spraying using Grazon ® in oil is likely to be an efficient and cost-effective technique (Pete Raal, Department of Conservation, pers. comm. July 2009). It may be possible to spray isolated plants on levees along the main channels from a boat. It may not be desirable to control some weeds (e.g. Scotch broom in swamps) by aerial application of herbicide because mortality of non- target species may increase risk of reinvasion by the same or another weed. In this case, cutting plants at ground level and applying herbicide to cut stumps may be the best option.

Species Suitable for Ground-Based Control

• Scotch broom (hand pull, cut and paste stumps with herbicide, backpack spray). • Gorse (hand pull, cut and paste stumps with herbicide, backpack spray). • Montpellier broom (hand pull, cut and paste stumps with herbicide). • Silver birch (fell (remove all green material or swab stump with herbicide) or kill standing by drilling. cutting, or frilling and application of herbicide). • Wilding conifers (hand pull seedlings, fell or kill standing (as for silver birch)). • Priority 3 weeds (fell macrocarpa, remainder hand pull, or cut and paste stumps with herbicide).

Suitable Species for Water-Based Control

• Scotch broom (herbicide spray). • Gorse (herbicide spray). • Crack willow (herbicide spray).

9.6.3 Use of volunteers

The use of volunteers to undertake control work has been previously proposed. Due to access difficulties, the amount of training required, and unavoidable hazards of working in this environment it is recommended that volunteers are only used for weed control on levees, wetland margins, and other easily accessed areas. The most suitable species for volunteers to target would be woody weeds such as gorse, Scotch broom, hawthorn, elder, cotoneaster, sweet brier, blackberry, and small trees (pines, Douglas fir, rowan etc). These species could be controlled by cutting close to ground level and immediately daubing stumps with Vigilant ® gel or other herbicide. It may also be possible to utilise volunteers for follow-up control of alder, crack willow, and

14 © 2009 14 Contract Report No. 218214

grey willow in some areas where access is not difficult, e.g. over half of the large alder infestation can be accessed on foot from the O’Brien property.

10. WEED CONTROL PRIORITIES

10.1 Pest management strategy for Otago 2008

Scotch broom, gorse, and Montpellier broom are the only species recorded in the wetland complex that are listed in the Pest Management Strategy for Otago (currently under review). These species are identified as threats to wetland ecosystems. One proposed gorse and broom rule is applicable to the Waihola/Waipori wetland complex:

(ii) Occupiers outside the urban areas, and outside the Gorse and Broom Free Areas (defined in Maps A1 to A7 in Appendix 3), must destroy gorse and broom within 10 metres of a property boundary where the neighbouring property is clear of gorse and broom within 50 metres of that boundary.

It would be desirable to encourage the Otago Regional Council to enforce this rule, to reduce the potential for invasion of these species into the wetland complex.

10.2 Setting priorities

It is not practicable to eradicate all pest plants from the wetland complex because of its large size, poor access in back swamps, the limitations of available control methods for some species, and reinfestation from outside sources. Management of weeds usually focuses on protecting the most ecologically important sites (site-led control) and/or on the weed species which pose the greatest threat to the wetland (weed-led control). Priority is usually given to halting the spread of existing infestations, controlling new infestations, and investigating future control options (c.f. Airey et al . 2000). This weed risk assessment is weed-led rather than site-led because sites with high ecological values have a low short-term risk from weeds. In addition, it is considered important to build on the control work already undertaken, even if funding is limited in future years.

The following factors have been used to develop a list of priority weeds for control (Table 6):

• Risk of further spread; • Current and future impact on indigenous values; • Distribution (density, extent, and location) of weed infestations; • Practicality and cost of management.

Landholder commitment and availability of funding and resources have not been included in the assessment for each species, as no information was available. Non- woody weeds such as creeping bent (Agrostis stolonifera ) and lotus ( Lotus pedunculatus ) are abundant, but have been excluded from assessments because no practical methods are available for their control in a wetland complex of this scale.

15 © 2009 15 Contract Report No. 218215

Three categories (1-3) have been identified. Priority 1 weeds require immediate action because of their extremely invasive nature and detrimental ecological effects. Priority 2 weeds are currently at low to moderate abundance, are relatively easy to control, and/or have high invasion risk. They require control as finances allow, although undertaking control of blackberry, Montpellier broom, and red alder sooner rather than later would likely reduce control costs in the long term (when they are more abundant). Priority 3 weeds have minor ecological effects. These weeds should be monitored and consideration given to their control once Priority 1 and Priority 2 weeds are being controlled.

Table 6: Priority 1, 2, and 3 weeds in the Waihola-Waipori wetland complex. Each weed is rated on a scale of 1-3 for each criterion, with higher numbers indicating higher risk of spread, higher impacts, more widespread distribution, and greater ease of management.

Weed Species Classification Risk of Risk Spread Impact Distribution Management Total Priority 1 Alder ( Alnus glutinosa ) 3 3 2 2 10 Crack willow ( Salix fragilis ) 3 3 3 1.5 10.5 Grey willow ( Salix cinerea ) 3 2 2 2 9 Reed sweet grass ( Glyceria maxima ) 3 2 2 1.5 8.5 Priority 2 Blackberry ( Rubus fruticosus agg.) 2 1.5 2 1 6.5 Elder ( Sambucus nigra ) 2 1 2 2 7 Gorse( Ulex europaeus ) 2 2 2 2 8 Montpellier broom ( Teline 2 1 1 1 5 monspessulana ) Red alder ( Alnus rubra ) 3 1 1 1 6 Scotch broom ( Cytisus scoparius ) 2.5 2 2 2 8.5 Silver birch ( Betula pendula ) 2 1 1 3 7 Wilding conifers ( Abies alba , 2 1 2 2 7 Chamaecyparis sp., Picea sitchensis , Pinus radiata , Pseudotsuga menziesii ) Priority 3 Bittersweet ( Solanum dulcamara ) 1 1 1 1 4 Cotoneaster sp. 1 1 1 2 5 Elm ( Ulmus × hollandica ) 2 1 1 2 6 Eucalyptus spp . 1 1 1 3 6 Hawthorn ( Crataegus monogyna ) 1 1 1 2 5 Macrocarpa ( Cupressus macrocarpa ) 1 1 1 3 6 Rowan (Sorbus aucuparia ) 1 1 1 3 6 Swamp cypress (Taxodium distichum ) 1 1 1 3 6 Sweet brier ( Rosa rubiginosa ) 1 1 1 2 5

16 © 2009 16 Contract Report No. 218216

11. PRIORITY 1 WEEDS

11.1 Alder ( Alnus glutinosa )

Habitat Damp ground along rivers and in wetlands. Reproduction Wind- and water-dispersed seed, and by suckering. Ecological Effects Can form dense, monotypic stands which overtop and displace indigenous vegetation. Trees trap sediment with their extensive root systems. Field Survey Alder forms one very large infestation to the southwest of Lake Waipori (Plate 1. Appendix 1) and smaller stands on the Lyders property. The large infestation may have started from a line of alder planted along an old drainage ditch, with the rate of expansion accelerating in recent years (Bert O’Brien pers. comm. 18 June 2009). Lone trees are scattered on or near the stop bank in the O’Brien property (to E2284093 N5466649 to the south and E2284725 N5467204 to the north) and in sedgeland and crack willow forest near the main infestation. Under a dense alder canopy, the understorey is dominated by Carex secta (Plate 2), with areas of Yorkshire fog ( Holcus lanatus ) and spearwort ( Ranunuculus flammula ), and shallow pools of water. Other species present include frequent swamp kiokio ( Blechnum minus ), occasional Coprosma robusta × propinqua , and less common peppermint ( Mentha × piperita ), Hypolepis ambigua , Histiopteris incisa , lotus, Coprosma propinqua , and flax. A litter of alder leaves was commonly observed in June 2009. Various age classes of alder are present, with lower stature trees likely to escape initial aerial spraying. Overall, alder forest was very uniform in nature, with very little variation in species composition. In contrast to alder forest at Lake Wairarapa in the North Island, which has a diverse range of native shrubs and some forest trees, such as kahikatea, growing under the alder canopy (DOC 2009), alder forest in the Waipori/Waihola wetland complex does not act as a nursery for indigenous species, probably because indigenous seed sources are no longer available. Current aerial spraying of alder on the Lyders property appears to have been unsuccessful, as trees still have some green leaves. However, from a distance there appears to have been some successful control of alder-crack willow forest alongside the main water channel running through the infestation. Control Aerial herbicide spraying. Initial results of trials in Lake Wairarapa indicate that Garlon ® (a.i. triethylamine salt of triclopyr) has been more effective in controlling alder than Roundup ® (a.i. glyphosate) (Greater Wellington Regional Council 2009). This herbicide is also suited to willow control. Feasibility of Alder is concentrated in one part of the wetland and should therefore Control be a candidate for eradication if a successful aerially-applied herbicide can be found. Follow-up control will be required to target trees missed or not killed in the first herbicide application. Ground- based follow-up control may be feasible, as it is possible to walk to the infestation from the O’Brien property.

17 © 2009 17 Contract Report No. 218217

Recommendations Eradicate alder from the wetland. Contact Greater Wellington Regional Council/DOC staff undertaking control of alder in Lake Wairarapa to help determine the best method of control. Undertake own research regarding alder control. If present, encourage removal of alder from privately-owned farmland surrounding the wetland complex.

W W A

R W

Plate 1: Alder forest (A), crack willow forest (W), and raupo reedland (R) to the south-west of Lake Waipori (Photograph: Ramon Strong).

Plate 2: Alder forest interior, with ground cover dominated by Carex secta .

18 © 2009 18 Contract Report No. 218218

11.2 Crack willow ( Salix fragilis )

Habitat Lowland rivers, streams, ponds, lakesides and other wetland communities, including swamps. Reproduction Vegetative, from shoot fragments dispersed by water. Ecological Effects Shading and displacement of native vegetation; reduction in biodiversity; decreases in water quantity, quality, and flow by changing hydrological patterns. Field Survey Crack willow is widespread in the wetland (Appendix1), in surrounding farmland, and along waterways upstream of the wetland. Current control efforts (aerial spraying with glyphosate) have been very effective in many areas. However, many areas have not been sprayed and regrowth of previously sprayed stands is occurring at several locations. In most cases, regrowth arises from the base of the plants. Death of non-target species as a result of herbicide spraying is mainly restricted to Carex secta in isolated locations. On some levees, Coprosma propinqua and cabbage tree have survived well under sprayed willow, the latter re-sprouting after defoliation and death of some limbs. Death of some C. propinqua and herbicide damage to flax has occurred on the margins of Lake Waipori, but swamp nettle is common in places. Crack willow is likely to reinvade the wetland from surrounding waterways such as the Taieri River. Control Aerial herbicide spraying, including blanket spraying of large areas where there is a dense canopy and spot spraying of outliers and regrowth. Avoid aerial spraying near the Eleocharis sphacelata reedland, and use ground-based methods instead. Feasibility of Current control measures appear to be working, although follow-up Control aerial spraying will be required in some areas. However, aerial herbicide spraying of crack willow regrowth will increase adverse effects on non-target species due to the lower stature and canopy density of sprayed stands. Control of crack willow will need to be ongoing due to reinvasion from surrounding areas. Landowners not currently involved in the willow control programme should be encouraged to become involved. Eradication is unlikely in the medium term, as large areas upstream and downstream of the wetland would also require control of crack willow. Ground-based control of regrowth on a large scale appears to be impractical due to poor access throughout much of the wetland, although ground-based herbicide treatment is an effective technique. Recommendations Continue to control crack willow. Continue to use aerial herbicide application within the wetland complex. Progressively expand willow control to areas outside the complex (upstream and downstream) as resources allow. Ensure that initial control has maximum possible kill rate to decrease costs associated with repeat applications, reduce effects on non-target species, and minimise the likelihood of having to undertake difficult ground-based control. Discourage landowners from planting crack willow.

19 © 2009 19 Contract Report No. 218219

11.3 Grey willow ( Salix cinerea )

Habitat Wide range of habitats, including swamps and riverbanks. Reproduction Seed (abundant but short-lived), fragments, and localised suckering. Wind-dispersed seed can travel tens of kilometres. Seed and fragments can be dispersed by water. Ecological Effects Forms dense thickets, displacing native vegetation, blocking waterways and modifying wetlands. Field Survey Small, live trees are scattered throughout the Sinclair Wetlands, although many trees have been sprayed and killed. Grey willow is also scattered throughout the DOC-administered Lake Waihola- Gillanders Wildlife Management Reserve. Larger areas of grey willow appear to have been the focus of previous (successful) control efforts, but many small (<3 m tall) isolated trees are still present. Grey willow is more common on the western margins of the wetland complex, although scattered c.1.5 m tall saplings were also recorded along the south-eastern margins of Lake Waipori (E2286438 N5465999). Control Aerial spraying of herbicide, using a spray gun for small areas/plants. Ground control: use a method such as drilling or frilling which kills trees while they are standing. Cutting and painting the stump with herbicide should only be used when all plant material can be removed from site (cut material may resprout). Feasibility of Grey willow has the potential to become more abundant than it Control currently is. This extremely invasive species is widespread throughout much of the wetland, often present as isolated small trees, which are difficult to spot and will be difficult to control from the ground. However, infestations are generally not large and starting targeted control operations as soon as possible would be advantageous in terms of resources expended. Grey willow could potentially become as widespread as crack willow, if control efforts do not begin soon. Recommendations Control grey willow in the wetland complex and surrounding areas. The target density for this species should effectively be zero – if plants are seen they should be killed. Control difficult-to-access plants by aerially-applied herbicide, using a systematic sweep of the wetland. Use a spray gun rather than a spray boom on small patches and individual trees. Accessible plants should be targeted from the ground.

20 © 2009 20 Contract Report No. 218220

G

W

R

Plate 3: Reed sweet grass along a drain (R), a lone grey willow (W), and gorse- elder-pohuehue scrub (G) on small cliffs in the Sinclair Wetlands.

11.4 Reed sweet grass ( Glyceria maxima )

Habitat Found in nutrient-rich, slow-flowing rivers and in ponds; prefers full sunlight but tolerant of slight shade. Reproduction Seeds are dispersed by water, mud, or machinery and can remain dormant for years. It can also be dispersed by rhizome fragments. Ecological Effects This species forms monocultures that can reduce plant species diversity and crowd out native wetland vegetation; it is a poor food source and a poor nesting and spawning substrate for wetland wildlife including galaxiid fish; and it spreads aggressively and impedes water flow. Field Survey Reed sweet grass is widespread within the wetland on the margins of waterways and edges of ponds. It is present from the northern margins of Lake Waipori to the southern margins of Lake Waihola. This species appears to be less prevalent where raupo ( Typha orientalis ) reedland is present, as raupo occupies similar habitats but is taller. However reed sweet grass occasionally grows within raupo or forms a band between raupo reedland and a levee. As raupo appears to competitively exclude reed sweet grass, raupo should not be controlled unless there is conflict with other activities at small sites such as encroachment on boat ramps. Extensive areas of reed sweet grass reedland were noted in ponds and back swamps within the Sinclair Wetlands and in a pond within Tatawai Trust land. Reed

21 © 2009 21 Contract Report No. 218221

sweet grass extends into indigenous wetland vegetation in many areas, often along very narrow, sinuous waterways that wind through the wetland complex. In the DOC-administered Lake Waipori Wildlife Management Reserve, cattle have browsed and trampled reed sweet grass, which may aid in its spread. Abundant reed sweet grass was also recorded along the banks of the Taieri River outside the study area. Control Cutting the floating mats and leaving them to rot has been undertaken on Tatawai Trust land. However, this is difficult work, the rotting vegetation smells and persists for a long time, and plants are not killed. Reed sweet grass can be killed by foliar spray of Glyphosate. A surfactant should not be added. Long-term control requires plantings of locally-sourced indigenous species to shade the waterway. Extreme caution should be applied during control efforts so that propagules are not spread. Feasibility of It will probably not be possible to eradicate reed sweet grass within Control the Waipori/Waihola wetland complex due to its abundance, widespread distribution, and ability to tolerate partial shading. However, it should be possible to reduce the size of major infestations and the overall ecological impacts of this species within the wetland. Recommendations Spray large infestations with glyphosate. Encourage raupo establishment in areas formerly dominated by reed sweet grass.

12. PRIORITY 2 WEEDS

12.1 Elder ( Sambucus nigra )

Elder is common on small cliffs and hillslopes on wetland margins, and scattered along levees. In the absence of taller indigenous tree species elder will persist at these sites if not controlled. It is recommended that elder is controlled at the same time as gorse and Scotch broom, and that any cleared areas are restored to indigenous forest using locally-sourced ecologically-appropriate indigenous plant species.

12.2 Red alder ( Alnus rubra )

Red alder was recorded in the southern part of O’Brien property, and appears to be spreading into the wetland. This partially flood-tolerant tree species is commonly associated with wetlands in the Pacific northwest of the United States, establishing in better-drained sites (Ewing 1996). This species, along with many of the wildling conifers present in this area, are likely to have spread from nearby plantation forest. Red alder should be controlled before it becomes more widespread within the wetland complex.

12.3 Wilding conifers

Several silver fir ( Abies alba ) were recorded in the southern part of the O’Brien property. One cypress ( Chamaecyparis sp.) was recorded growing under crack willow forest near the stopbank on the O’Brien property.

22 © 2009 22 Contract Report No. 218222

Sitka spruce (Picea sitchensis ) is present on the Lyders property (Dan Lyders pers. comm. 2 April 2009). This species reproduces by wind-dispersed seed and tolerates moist sites, and may therefore become a greater problem in the future. Current aerial spraying of herbicide does not seem to have affected this species.

Isolated wilding pines (Pinus spp.) are scattered widely throughout the wetland. Some trees present within the wetland are probably old enough to cone. Douglas fir (Pseudotsuga menziesii ) is scattered throughout north-western margins of the wetland complex.

These species are currently having low ecological impacts, but it is recommended that they are controlled within the wetland while they are at low abundance. They are likely to become more widespread within the wetland or on wetland margins if left uncontrolled. Wildling conifers may continue to reinvade the wetland from nearby plantation forest and farmland, and should be controlled on an ongoing basis.

W P

F S D

W

W

Plate 4: Wilding pine (P), crack willow (W), Douglas fir (D), silver fir (F), and silver birch (S) alongside Berwick Road (O’Brien property). Red alder is also present here.

12.4 Scotch broom ( Cytisus scoparius ) and gorse ( Ulex europaeus )

Dense scrub dominated by gorse and/or Scotch broom (Cytisus scoparius ) is present on low cliffs around Ram Island and Lonely Island, along low cliffs on the western

23 © 2009 23 Contract Report No. 218223

edge of the wetland, and at scattered locations to the east of the wetland (Appendix 1). Common associate species are elder and pohuehue. Gorse is also scattered along levees on the main waterways and present as isolated specimens within drier areas of back swamps.

Several areas of Scotch broom scrub also grow in a back swamp to the east of Ram Island (Appendix 1). On the Duck Consortium property, Scotch broom dominates the understorey under crack willow forest, and also forms areas of scrub within back swamp vegetation. Gorse and Scotch broom are scattered along road margins and stop banks bordering the wetland.

It is recommended that large areas of Scotch broom in swamps east of Ram Island and in the Duck Consortium property are controlled. Scotch broom and gorse on Ram Island and Lonely Island should continue to be controlled to remove a seed source, but Scotch broom and gorse have long-persistent seed banks and are likely to re- establish after disturbance. A sustainable solution to gorse and Scotch broom infestations is to establish a shady canopy of indigenous forest above them, as neither tolerate low light levels. Indigenous tree species should be planted where dense infestations of gorse and Scotch broom have been removed, to inhibit reinvasion. It may be possible to use volunteers to control these species on levees and wetland margins.

12.5 Montpellier broom ( Teline monspessulana )

A few individuals of Montpellier broom were recorded on the northeast boundary of the Duck Consortium property along an old roadway. It is recommended that Montpellier broom is eradicated from the wetland while it is at low abundance.

12.6 Silver birch ( Betula pendula )

Silver birch was recorded on the Duck Consortium property and the nearby part of the O’Brien property. It is also scattered along stop banks on the western margins of the wetland complex and present as a specimen tree and as windbreaks on neighbouring properties surrounding the wetland. This species has the ability to form a dense canopy in wetlands. It is recommended that silver birch is controlled within the wetland while its abundance is low. Trees already present will allow seed to spread further into the wetland. Reinvasion is likely from surrounding areas.

13. PRIORITY 3 WEEDS

The Priority 3 weeds listed in Table 6 are currently difficult to control (e.g. bittersweet), have low ecological impacts (e.g. hawthorn), and/or are unlikely to colonise very wet ground (e.g. eucalyptus). However, typically dryland species can colonise dry areas within the wetland (evidenced by the presence of narrow-leaved lacebark), and therefore all woody weed species should be controlled where and when resources allow. Weed-led control and/or eradication objectives would be suitable for many of these species. Indigenous plantings undertaken on the drier margins of the wetland (e.g. Ram Island), on levees, and as part of the proposed Titri restoration project will benefit from control of woody weed species. Herbaceous and grassy

24 © 2009 24 Contract Report No. 218224

weeds such as lotus ( Lotus pedunculatus ; abundant throughout), creeping bent (Agrostis stolonifera ; abundant in some areas), and tall fescue ( Schedonorus phoenix ; abundant in back swamp in Sinclair wetlands, and on levees), as well as the exotic soft rush (Juncus effusus ; frequent on levees) are likely to be having an ecological impact within the wetland, but are not included here as there is no practical control option available for these species at present. Similarly, the aquatic Canadian pond weed ( Elodea canadensis ) is not considered for control at the present time.

Non-local native species have been planted on Ram Island (e.g. Pittosporum sp.). Some of these species (e.g. karamu) can invade indigenous habitats reducing ecological values in the long-term. Consideration should be given to replacing these species with locally-native alternatives.

14. ESTABLISHMENT OF INDIGENOUS VEGETATION

14.1 General

The goal of indigenous forest restoration is to rapidly create a dense indigenous canopy which inhibits weed establishment through shading. Plants used should be indigenous species sourced from Tokomairiro Ecological District and matched to site conditions. The site must be adequately prepared, plantings undertaken at sufficient density (0.5-1 m spacings), and plantings maintained until full canopy closure occurs (usually 2-3 years).

Suitable sites for forest establishment include those covered in gorse and Scotch broom (e.g. Ram Island, Lonely Island, small cliffs on the western margins of the wetland, eastern margins near Waihola township, publicly accessible areas (such as the proposed Titri wetland restoration), highly visible areas (e.g. parts of Waipori/Waihola Wildlife Management Reserve visible from SH1), the area of macrocarpa if it is removed, and levees.

The need for planting should be balanced against the probability that the current vegetation will succeed to vegetation dominated by indigenous species. Natural regeneration and succession of indigenous vegetation is already occurring within the wetland. This should be encouraged through controlling weeds and minimising spraying of non-target species. Species such as flax and raupo can easily be established in appropriate habitat if required at specific sites because of a lack of natural regeneration.

14.2 Raupo

Raupo is increasing in abundance, predominantly in the northern parts of the wetland. This species has high fauna values, providing habitat for birds (e.g. Australasian bittern) and invertebrates, and also acts as a nutrient sink. In addition, establishment of raupo is likely to help control reed sweet grass as both species prefer similar habitats (slow-moving water). It is unlikely that raupo will block main waterways due to swift-flowing currents. The widespread control of raupo is therefore not recommended. However, raupo may become dense in traditional duck shooting ponds or around boat ramps, and local residents have identified some areas in the east of the

25 © 2009 25 Contract Report No. 218225

wetland where raupo is hindering boat access. Control of raupo in some of these areas may be justified, although specific sites should be identified in the management plan for the wetland complex. In sites where raupo is controlled, some areas of raupo should be left to provide habitat for fauna (Murray Neilson, Department of Conservation, pers. comm. July 2009). Planting of raupo for sweet reed grass control should not occur in areas designated for raupo control.

15. RECORD KEEPING, MONITORING, AND RESEARCH

15.1 Records

Detailed records of all control operations should be kept to inform future decision making. These should include records of area covered (GPS coordinates), climatic conditions during control operations, targeted weeds, control method used including herbicide type, costs, and time involved.

15.2 Monitoring

Monitoring should be designed to inform future decision making such as choice of control method and when to undertake future control. Monitoring should never be an end in itself, but designed to answer specific questions required for wetland management. The suggested goals of monitoring are to determine:

• Distribution of weed species within the wetland complex. • Short- and long-term effectiveness of aerial control operations. • Effect of weed aerial control operations on non-target species. • Patterns of vegetation succession following weed control. • Rates of reinvasion of crack willow.

Ideally, a survey of weed distributions within the wetland should be undertaken every five years. For species such as alder and crack willow, the extent of large infestations could be assessed using aerial photography. Particular attention should be paid to the presence of Priority 1 weeds, the distribution of Priority 2 weeds, and the presence of any new invasive weed species.

The success of weed control (particularly that of crack willow, grey willow, and alder) should be monitored from the air to get an overall picture of the extent of control and from the ground to make an accurate assessment of any regrowth and adverse effects on non-target species. Monitoring will need to be undertaken at several intervals after weed control, e.g. 1 and 2 years after initial control.

The effect of weed control on non-target species and patterns of vegetation succession could be monitored by establishing permanent vegetation plots. Replicate 10 m × 10 m plots would need to be established in alder forest and crack willow forest. Some plots should contain swamp nettle so that effects of weed control on this species can be more accurately assessed. This monitoring will determine 1. Whether control of target weeds has been successful, 2. Whether weed control is followed by reinvasion of the same or another weed species, 3. Whether important non-target

26 © 2009 26 Contract Report No. 218226

species have survived weed control operations, and 4. What successional trends follow weed control.

Crack willow reinvasion from outlying areas should be monitored along channel margins within the wetland complex, but only when crack willow is at low abundance within the wetland. Until then it will be difficult to differentiate between reinvasion from outside or within the wetland. A cost:benefit analysis should then be made to assess whether it is more important to control reinvasion in a buffer zone or progressively control crack willow in outlying areas from which invasion is occurring. Due to the large amount of crack willow in some waterways connected to the wetland, and the potential for conflicts with landowners, it is likely that controlling reinvasion within the wetland and suggested buffer areas will be the best option. Monitoring may be as simple as annual counts of live crack willow along randomly selected major waterways.

15.3 Research

The long-term benefits of control of reed sweet grass are uncertain. Research could be undertaken to more accurately determine the advantages and disadvantages of reed sweet grass control. This will require selection of representative sites (e.g. levee on main channel, pond, drainage ditch), replication of treatments (e.g. spraying only, planting of raupo only, spraying with follow-up planting of raupo), control sites (no treatment), and monitoring. Spraying treatments could further be divided into single spray and follow up sprays.

Research relating to new control techniques should be monitored so that better control techniques can be used if they evolve. This should not only include species currently requiring control (e.g. alder), but species where no cost-effective control methods currently exist (e.g. blackberry).

16. OTHER ISSUES

16.1 Dead trees blocking waterways

One of the most contentious issues brought up by local residents is dead willows blocking channels within the wetland. Once dead trees have fallen into the water decomposition is extremely slow (Steve Bryant pers. comm. May 2009). Dead trees can create boating hazards and entrap further material. The cost of removing all fallen trees from waterways would be an unacceptable drain on financial resources. It is therefore recommended that trees are removed only where necessary, e.g. if a major waterway becomes impassable to boats. After long term woody weed control has been undertaken there will be less tree fall into waterways.

16.2 Financial resources

Weed control within the wetland complex is a major undertaking and funding sources should continue to be sought.

27 © 2009 27 Contract Report No. 218227

16.3 Grazing

In the DOC-administered Lake Waipori Wildlife Management Reserve, cattle have browsed, trampled, and defecated in reed sweet grass and raupo reedland. Most of this reserve comprises indigenous vegetation which is adversely affected by grazing. Cattle access is mainly along a constructed levee and its margins, and cattle do not appear to have gone further west than about E2285792 N5465899. If nearby landowner(s) wish to continue stock grazing on the stop bank, it is recommended that fencing is used to exclude stock from the wetland.

Cattle were observed grazing on a privately-owned island in the delta where the Waipori River enters Lake Waipori (part of the Beaton property). Cattle do not appear to have access into the main part of the wetland from this island.

16.4 Dumping of rubbish

Organic and inorganic rubbish has been dumped on the edge of the wetland near the Waipori pump station on the Contour Channel. This rubbish may have come from nearby cribs, although there is public access to this area. Rubbish includes plant fragments of elder, which can resprout. The dumping of rubbish in the wetland should be discouraged.

17. CONCLUSIONS

Weeds are having significant detrimental effects on the ecological, recreational, and cultural values of the internationally significant Waihola/Waipori wetland complex. Currently, the most invasive weeds are alder, crack willow, grey willow, and reed sweet grass. Eradication of alder and control of grey willow and crack willow within the wetland complex can be successfully completed if sufficient resources are made available. No cost-effective control method is available for the eradication of reed sweet grass due to its abundance, widespread distribution, and tolerance of partial shading. However, large infestations can be easily controlled, reducing this species’ adverse ecological effects at selected sites. Lower priority weeds within the wetland - such as gorse, Scotch broom, elder, and wilding conifers - should be monitored and controlled as resources allow. Weed control operations are not expected to adversely affect natural values, as long as ground-based methods (rather than aerial application of herbicide) are used in the vicinity of one uncommon vegetation type.

18. RECOMMENDATIONS

• Continue to seek involvement of remaining landowners.

• Encourage removal of grey willow, alder, and silver birch on all properties on western margins of wetland.

• Landowners and land administrators, including councils, should be encouraged not to plant invasive species such as crack willow, grey willow, alder, and silver birch. Indigenous species, such as kahikatea, kohuhu ( Pittosporum tenuifolium ),

28 © 2009 28 Contract Report No. 218228

cabbage tree, flax, kanuka, manuka ( Leptospermum scoparium ), and kowhai, would be suitable substitutes for screening and erosion control on wetland margins, and would also provide important habitat and food sources for indigenous fauna.

• Define the control area. Include a c. 2-5 km long crack willow control buffer on major waterways entering the wetland complex.

• Give priority to the control of alder, crack willow, and grey willow.

• Eradicate alder from the wetland complex.

• Aim for an effective ‘zero density’ of grey willow in the wetland complex and from surrounding areas.

• Continue crack willow control and expand it to cover the entire wetland complex and within a buffer zone on major waterways.

• Control crack willow beyond the wetland boundaries only when this species is at low abundance within the wetland.

• Crack willow control on the margins of waterways has priority, so that rates of spread are slowed. This has the added benefit of maintaining open channels for recreational use.

• Maximise the kill rate of the target species during the first control operation, so that follow-up control is minimised.

• Priority 2 weed species are controlled where resources allow.

• Priority 3 weed species are monitored and controlled if easy to do so.

• Raupo is only controlled in designated areas that are listed in the management plan.

• The dumping of rubbish within the wetland is discouraged.

• The success of all control operations is monitored.

• Vegetation changes are monitored following weed control.

ACKNOWLEDGMENTS

Thanks to Nicola McGrouther (ORC) for project liaison, and Ramon Strong (ORC) and Pete Ravenscroft (DOC) for photographs, Ian Hadland (Fish and Game NZ) for a boat trip and discussion, landowners (Dan Lyders, Alastair Cameron, Scott Kunac, David Darling, Bert O’Brien, Wendy and John Campbell, Shane Hildred) for access to their properties, Andrew Nicol (ORC) for unlocking the gates along the stopbank, Ngai Tahu and Johnny MacIntyre

29 © 2009 29 Contract Report No. 218229

for access to Sinclair Wetlands, Kai Tahu for access to Tatawai Trust Land, and Steve Bryant for providing access around the wetland by boat. Thanks to Pete Raal, Graeme Loh, Bert Weil, and others for feedback on a draft of this report.

REFERENCES

Airey S., Puentener R. and Rebergen A. 2000: Lake Wairarapa wetlands action plan 2000- 2010. Department of Conservation, Wellington. 68 pp. Cromarty P (comp.) and Scott D.A. (Ed.) 1996: A directory of wetlands in New Zealand. Department of Conservation, Wellington. de Lange P.J., Norton D.A., Courtney S.P., Heenan P.B., Barkla J.W., Cameron E.K., Hitchmough R. and Townsend A.J. 2009: Threatened and uncommon plants of New Zealand (2008 revision). New Zealand Journal of Botany 47 : 61-96. DOC 2009: Lake Wairarapa alder a double-edged sword. Footnotes April 2009. Department of Conservation, Wellington Conservency. http://www.doc.govt.nz/about- doc/news/whats-new/footnotes/footnotes-april-2009/lake-wairarapa-alder-a-double- edged-sword/ . Dunedin City Council 2006: Dunedin City District Plan. Operative 3 July 2006. Dunedin City Council, Dunedin. Ewing K. 1996: Tolerance of four wetland plant species to flooding and sediment deposition. Environmental and Experimental Botany 36 : 131-146 Greater Wellington Regional Council: 2009: Water Supply and Parks January/February 2009 Attachment 1 to Report No. 09.128. http://www.gw.govt.nz/council- reports/pdfs%5Creportdocs%5C2009_128_2_Attachment.pdf Johnson P. 2004: Te Nohoaka o Tukiauau: The Sinclair Wetlands. Landcare Research Contract Report LC0405/052. Prepared for Te Runanga o Ngai Tahu, Christchurch. NZPCN 2009: New Zealand Plant Conservation Network. http://www.nzpcn.org.nz Peach R.M.E. and Johnson P.N. 2004: Lake Waihola, the R.W. Sinclair Wetland: botanical report. Landcare Research Contract Report LC0304/148. Prepared for Department of Conservation, Dunedin. 22 pp. Schallenburg M and Burns C.W. 2003: A temperate, tidal l;ake-wetland complex 2. Water quality and implications for zooplankton community structure. New Zealand Journal of Marine and Freshwater Research 37 : 429-447. Tangney R.S. 1987: The vegetation of Lakes Waihola, Waipori and associated wetlands. Prepared for the Department of Conservation. 21 pp. The Property Group 2007: Land Status Report - Lakes Waihola and Waipori - Otago Land District. Prepared for Otago Regional Council, Dunedin. Ward G. and Munro C.M. 1985: Queen Elizabeth II National Trust Conservation Covenant (H. Sinclair). Unpublished botanical report, in Biological Survey of Reserves format. 4 pp.

30 © 2009 30 Contract Report No. 218230

APPENDIX 1

WEED DISTRIBUTION MAPS

31 © 2009 31 Contract Report No. 218231