The projects below show our restoration work in practice, across settings as different as whitebait spawning habitat at the head of a harbour, purpose-built wetland refuges for a Nationally Critical fish, eroding loess slopes, and a willow-choked braided river. Across all of them the work is the same at heart: bringing habitat back for the species and cultural values that depend on it, through hands-on design and the field surveys, trials and monitoring that show what actually works. Often the evidence overturns the standard instinct – stabilising a bank can worsen erosion, and saving a threatened fish can mean building barriers to keep other fish out – and it’s the willingness to follow it that puts these places back on the right trajectory.
Īnaka ki Whakaraupō – Īnanga Spawning Habitat Design
A ki uta ki tai project with Te Hapū o Ngāti Wheke and the University of Canterbury to restore īnaka spawning habitat and fish passage at the head of Whakaraupō – pairing Western science with mana whenua knowledge to put a mahinga kai taonga back on the right trajectory.
Īnaka ki Whakaraupō is a multi-year restoration project undertaken in partnership with Te Hapū o Ngāti Wheke – mana whenua of Whakaraupō – and the University of Canterbury Marine Ecology Research Group, with EOS Ecology providing the scientific and design lead. Set under the Head of Harbour Project (itself part of the wider Whaka-Ora Healthy Harbour Plan for Whakaraupō/Lyttelton Harbour), the work targets īnaka – whitebait, a mahinga kai taonga whose spawning success at the head of the harbour had declined to the point that egg counts across the focus catchments were running in the tens of hundreds rather than the tens of thousands that healthy populations produce. The project deliberately takes a ki uta ki tai (mountains-to-sea) approach, integrating Western science with the historical knowledge held by Ngāti Wheke to identify where habitat and fish-passage improvements will deliver the most benefit, and to make sure that the priorities for restoration align with both ecological and mana whenua aspirations.
Across the 2020/21 spawning season we undertook saltwater wedge surveys (the upper tidal reach where saltwater meets freshwater and īnaka characteristically spawn), in-stream habitat surveys, and egg counts at five inner-harbour catchments between November 2020 and May 2021. From those surveys we identified three priority sites where targeted habitat improvement would deliver the largest gains. Alongside the field work, EOS Ecology built a multifactor GIS-based prioritisation framework for fish-passage barriers across the wider Whakaraupō catchment – a scoring system that weights mahinga kai value, waterway flow permanence, land protection status, land cover, and landowner engagement feasibility, so that limited remediation effort can be directed where it will most clearly translate into improved access to upstream habitat for adult īnaka returning from the sea.
One of the priority sites – Allandale Stream – became the focus for detailed design drawings, prepared with Lucas Associates as landscape architects. Three remediation focus areas address different parts of the catchment. In the lower tidal reaches, mature ~20-year-old riparian planting has stabilised but also narrowed and deepened the channel beyond what īnaka need; the design re-contours and batters the banks, installs coir logs and replants under resource consent. An existing concrete-and-boulder weir adjacent to farm buildings attenuates sediment usefully but blocks adult īnaka from moving up to the deeper holding pools they use outside the spawning window – so the design remediates the barrier without losing the sediment function. Higher in the valley, a 3.5–4 m gouged channel illustrates a counter-intuitive lesson: that the standard ‘stabilise the banks’ instinct can, in some settings, make stream erosion worse by narrowing and accelerating the flow. Together, the package gives Ngāti Wheke a buildable, science-supported pathway for restoring īnaka to one of the harbours that historically supported them, and the prioritisation framework gives the wider Head of Harbour Project a way to extend the same logic across the other catchments around Whakaraupō.
Papatahora Stream Restoration
Two ecological Assessments of Environmental Effects supporting Te Taumutu Rūnanga’s restoration of Papatahora, a culturally significant small stream flowing through the grounds of Ngāti Moki Marae, as part of the mana whenua-led Whakaora Te Waikēkēwai project.
Papatahora is a small stream on the south-eastern side of Te Waihora/Lake Ellesmere at Taumutu, flowing through the grounds of Ngāti Moki Marae before joining Te Waikēkēwai, which in turn winds through Te Repo Orariki/Taumutu Wetlands and discharges into Te Waihora. The stream is the focus of restoration under Whakaora Te Waikēkēwai – a mana whenua-led project sitting within the broader Whakaora Te Waihora programme that aims to enhance mahinga kai habitat, improve water quality and increase indigenous biodiversity by fencing and restoring waterways with landowners, recreating wetlands, and supporting on-farm actions across the catchment. EOS Ecology was commissioned by Te Taumutu Rūnanga and Environment Canterbury to deliver the ecological inputs for two phases of resource consenting – an initial AEE in 2020 for sediment removal from a ~90 m section of Papatahora downstream of Pohau Road (EOS Ecology Report ENV01-20034-01), and a 2024 update extending the assessment ~950 m upstream of Pohau Road to support a wider set of proposed restoration measures (TET01-23103).
The 2020 work paired in-stream and riparian habitat assessments with kick-net macroinvertebrate sampling and a full fish-survey programme (Gee minnow traps, fyke nets and electrofishing, deployed as site conditions allowed) across Papatahora and its downstream receiving environment in Te Waikēkēwai. The 2024 update added eDNA sampling at two sites in the upstream reach as a time- and cost-effective way to expand the baseline. Both the impact reach and the receiving environment were assessed as ‘moderate’ ecological value, with the at-risk – declining longfin eel among the species recorded. Without mitigation, the proposed sediment removal carries a ‘moderate’ magnitude of adverse effect – primarily potential mortality of stream biota and entrainment of fine sediments depositing downstream – which the Roper-Lindsay et al. (2018) value/magnitude matrix translates to ‘more than minor’ under the RMA. The recommended package of fish relocation, erosion and sediment control procedures, and careful staging and timing of works reduces that to ‘low’ magnitude and ‘less than minor’ effects in an RMA context.
Both AEEs were delivered to tight resource-consenting timeframes (and through Covid-19 lockdown disruption in 2020), with joint pre-survey site visits with the rūnanga and regional council to align expectations from the start. Beyond the formal AEEs, the work has extended into close collaboration with planners, contractors and the Whakaora Te Waihora team on practical sediment-removal methodology, on how ecological design can be embedded into the works themselves to maximise habitat outcomes, and on long-term sediment management – including the proposed construction of a wetland, Te Repo o Papatahora, on the other side of the stream. EOS Ecology has also provided pro bono drone flights to capture video and photos of the restored waterway and wetland following project completion, along with high-resolution georeferenced orthomosaic maps that support project reporting and help guide future catchment-based mahi.
Kōwaro/Canterbury Mudfish Habitat Restoration & Translocation
Designing, constructing and monitoring new habitat for the Nationally Critical Canterbury mudfish, one of New Zealand’s four most threatened freshwater fish, and writing the DOC-approved transfer plan that puts founder populations back into suitable wetland refugia across mid- and inland Canterbury.
The Canterbury mudfish (Neochanna burrowsius), kōwaro, is one of just four New Zealand freshwater fish species classified Threatened – Nationally Critical (with the qualifier ‘conservation dependent’) – the highest threat status the country gives an extant freshwater fish. Its distribution is limited to 16 mid- and South Canterbury catchments, fragmented across an intensifying agricultural landscape; the 2014–2016 Canterbury drought, planned reductions in water-race recharge, and ongoing climate-change-driven weather extremes have all sharpened the trajectory. EOS Ecology was engaged by Environment Canterbury (ECan) to support two parallel ECan-led restoration programmes – the Hekeao Hinds Managed Aquifer Recharge (MAR) project, governed by the Hekeao Hinds Water Enhancement Trust, and the Selwyn-Waihora Targeted Stream Augmentation (TSA) project – both delivered under the Canterbury Water Management Strategy and aligned with Action 4.3 of the NZ Mudfish Recovery Plan to reintroduce mudfish to suitable habitat. The work is undertaken in close collaboration with the Department of Conservation, Selwyn District Council, Fonterra, landowners, and the external Canterbury mudfish specialist consultancy Ichthyo-niche.
On the ground, EOS Ecology has designed and overseen construction of new and restored kōwaro habitat at two mid-Canterbury sites – one delivering more drought-resilient habitat following the 2014–2016 sequence, the other built after the devastating May/June 2021 mid-Canterbury floods – and contributed to a separate restoration and long-term management plan for a kōwaro site within the Selwyn water-race network. UAV drone surveys are used to track plant composition and density at the developing sites so that releases happen only once habitat conditions can actually support a founder population. On the regulatory side, we authored the DOC-approved Canterbury Mudfish/Kōwaro Transfer Plan (ENV01-20056-01, August 2020) for the South Hinds and Broadacres release sites, alongside site-specific Management Plans for the Broadacres release sites on Raywell Farm Stream (ENV01-20056-03) and the South Hinds release pond near Mayfield – each plan setting out a 14-criterion habitat-suitability checklist that has to be met before any fish are moved.
The transfer protocols are deliberately conservative: a 1.5 fish/m² stocking rate (a third of the maximum recorded mudfish density), 50–550 fish per release site released across at least two to four seasons, transfers staged through autumn into early winter so the fish are settled in time for late-winter / early-spring spawning, and an adaptive-management feedback loop that lets the programme pause or accelerate based on monitoring results at the release sites. Founder populations are drawn from a wild population in the Akaunui Wetlands at Eiffelton (for South Hinds) and a captive population at Dunsandel (for Broadacres) that has been held in specialised tanks since being rescued from drying Selwyn waterways during the 2014/15 drought. Because the same eels and trout that would otherwise reach the release sites are themselves kōwaro predators, the plans include the deliberately counter-intuitive step of installing and maintaining fish-passage barriers that keep them out – turning standard freshwater management logic on its head in the service of saving a species that, without this kind of focused intervention, has limited prospects in the wider rural landscape.
Selwyn Water Race Kōwaro/Canterbury Mudfish Habitat
Site-specific habitat protection and restoration for one of the few remaining wild Canterbury mudfish populations, currently sheltering in a terminal section of the Selwyn District Council Ellesmere water race that itself is scheduled to close.
A terminal section of the Selwyn District Council Ellesmere water race scheme – running between Southbridge Dunsandel Road and Leeston Dunsandel Road in mid-Canterbury – supports one of the very few remaining wild populations of kōwaro / Canterbury mudfish (Neochanna burrowsius), one of just four New Zealand freshwater fish classified Threatened – Nationally Critical (with the qualifier ‘conservation dependent’). The site sits at the intersection of competing management drivers: the Department of Conservation needs to protect the remnant population; Selwyn District Council manages the water race for stock-water delivery and plans to eventually close the Ellesmere scheme entirely; Sicon (as SDC’s water-race flow contractor) needs water to reach downstream users; and the landowner is operating a dairy farm whose pivot irrigator crosses the channel in seven places. EOS Ecology was engaged by the Department of Conservation following a multi-party site visit in March 2021, and has been working with DOC, SDC, Environment Canterbury, Fonterra and the landowner ever since to design and implement a site-specific management approach that can hold all of these interests together while securing the future of the population.
Our initial three-tier advice memo (DEP01-21007, March 2021) recommended separating this reach from the wider water-race programme so it could be managed for kōwaro specifically – and that recommendation crystallised into a Management Plan (DEP01-21007-02, June 2022) recognised by all project partners. The plan addresses a core ecological tension: kōwaro need exactly the conditions that standard rural stream and water-race management actively removes. They prefer still or very slow-flowing ‘swampy’ habitats with deep pools, soft bottoms, and extensive aquatic vegetation (which they use both as a food source and as a spawning substrate – eggs are spread widely throughout the vegetation, and the fish delay spawning when it is absent), and they need to be kept apart from predators such as trout. Standard water-race management – mechanical clearance to keep flow moving – is directly at odds with that. The Management Plan therefore proposes a two-stage transition: Stage 1 isolates the site from the wider scheme, improves riparian habitat along the existing channel, and secures an independent flow source via a solar-powered groundwater pump, so the kōwaro habitat no longer depends on the water race that is scheduled to close; Stage 2 then offsets the loss of upstream and downstream habitat by creating ~2.2 ha of new pond and stream habitat on adjacent SDC land, including deeper lined ponds as drought refugia.
EOS Ecology continues to deliver the engineering and ecological details that put the plan into practice. We have designed the soakage-area weir at the downstream end of the protected reach (SEL01-22066, November 2022) – a deliberately permeable structure that holds water on site to facilitate soakage while preventing kōwaro being washed downstream into the dry section during flood events. We have specified the methodology for, and overseen, fish-removal work during pivot-crossing construction (SEL01-22087, methodology February 2022 and implementation memo April 2023), where six undersized polythene-pipe culverts on the landowner’s dairy property are being replaced by single 6.1 m × 0.9 m concrete bridge crossings per pivot – opening up channel flow and reducing the blockage events that had previously inundated adjacent paddocks – with the fish-removal operation conducted under the relevant MPI, DOC and Fish & Game permits. More recent flow-gauging work is informing the design of the long-term groundwater supply that Stage 2 of the plan depends on. Together, these incremental design steps are securing the future of a critically important kōwaro population at exactly the moment when the water race that has been sheltering it is being wound down.
Port Hills Erosion Control Trials
A replicated field trial of five commonly-used erosion-control treatments on Port Hills loess, designed and run with the Cashmere Stream Working Group – generating the quantitative evidence base that fed directly into updates to Canterbury’s Erosion and Sediment Control Guidelines and laid the foundation EOS Ecology later built on in the Cut Slope Erosion trial.
The dispersive loess soils on the lower slopes of Christchurch’s Port Hills are highly erodible the moment they are exposed – and the post-earthquake surge in construction activity on the hills had placed real pressure on the waterways draining them, with clay and fine silt particles entering streams in wet weather and staying in suspension long enough to do serious downstream harm. The Cashmere Stream Working Group of the Christchurch–West Melton Zone Committee commissioned EOS Ecology, on behalf of Environment Canterbury, to design and run a replicated field study of the erosion-control products being used on Port Hills construction sites at the time – so that contractors, councils and consent authorities had real quantitative evidence rather than supplier claims to go on, and so that the findings could feed directly into updates to Canterbury’s Erosion and Sediment Control (ESC) Guidelines and to construction-phase consent conditions.
Across three weekends in February 2016, with volunteer community help organised by the Cashmere Stream Working Group, we tested five erosion-control treatments – topsoiling, straw mulch, coconut fibre blanket, and two hydraulically applied soil stabilisers (WRD-L and Vital Bon-Matt Stonewall) – against bare loess subsoil controls. Each treatment was applied in triplicate over loess subsoil sourced from a nearby construction site, hand-compacted into 2 m × 1 m soil boxes set at an 11° slope; a programmable rainfall simulator delivered 29 mm/hr for one hour while one-litre sediment samples were collected at five-minute intervals and returned to the laboratory for analysis. All five treatments significantly reduced sediment yield against the bare control – straw mulch (95% reduction) and the coconut fibre blanket RECP (94%) sat at the top, with Vital Bon-Matt Stonewall (90%) and topsoiling (86%) close behind; the calcium-lignosulphonate-based WRD-L (48%) clearly trailed the rest.
The trial carried several messages that have shaped Port Hills construction practice since. Even slopes protected with these products still released sediment at concentrations above generally consented limits – meaning a ‘treatment train’ of erosion control plus downstream sediment management (e.g. retention ponds with flocculation) is still essential on any substantial earthworks site. Hydraulically applied stabilisers depend on proper 100% coverage and timely reapplication if they degrade; every product has strengths and weaknesses by setting (straw for low-angle slopes, coconut RECP for steeper ones); every newly exposed loess surface needs cover applied immediately; and the most reliable long-term solution remains rapid establishment of vegetative cover assisted by topsoil. EOS Ecology translated the trial results into a companion memo recommending specific updates to Canterbury’s ESC Guidelines and to construction-phase consent conditions – including loess-specific guidance on soil horizons, dispersive-soil recognition and management, and the appropriate matching of short- and long-term erosion-control measures to construction staging. The 2016 study also laid the empirical foundation that EOS Ecology later built on in the Cut Slope Erosion trial at the Christchurch Adventure Park, which extended this short-term construction-phase focus into testing long-term combinations of products and native plantings on roadside loess cuttings.
Cut-Slope Erosion Control Research
A multi-year replicated field trial pioneering low-cost, soft-engineering solutions to reduce sediment runoff from the bare loess cut faces along Port Hills and Banks Peninsula roads – the first such study to test erosion-control products and native plantings against proper experimental controls.
The loess soils on the lower slopes of Christchurch’s Port Hills and Banks Peninsula are silt-sized, chemically dispersive when wet, and prone to severe erosion the moment vegetation and topsoil are removed – generating sheet-wash and rill erosion at the surface, slips and tunnel gullies at depth, and a fine sediment load that is then very difficult to remove from stormwater once it is in suspension. Exposed roadside cuttings are an obvious source, and with limited room in the narrow road corridors around the harbour for downstream treatment systems, the onus has to fall on reducing sediment runoff at the cut face itself. As part of the Whaka-Ora Healthy Harbour initiative – and with support from the Banks Peninsula Zone Committee – Christchurch City Council and Environment Canterbury commissioned EOS Ecology to design and implement a three-year replicated field study testing ‘soft’ erosion-control measures on loess cut slopes: the first such trial after years of largely anecdotal supplier-led product testing.
After an initial 2015 desktop study and a pilot summarising existing knowledge of erosion around Whakaraupō, EOS Ecology partnered with the soil-science specialists at Manaaki Whenua / Landcare Research and the construction team at Fulton Hogan to develop the experimental design and ready a south-facing cut-loess site within the Christchurch Adventure Park. Site setup in May–June 2019 included rebattering and scraping the face to expose fresh loess subsoil, constructing a stable rock toe at the base (locally sourced volcanic rock, backfilled and planted), installing diversion flumes to redirect upslope water away from the cut face, and putting in a bidim-lined rock channel water table at the toe. Five erosion-control products were then applied to the face in a randomised block design with three replicates each, alongside three untreated control plots – two sprayed hydromulch products with grass-seed mixes, two rolled products (a thin jute mat and a thick wool blanket), and a combined coir-and-hydromulch product. Six native plant species – silver tussock, Banks Peninsula fescue, cutty grass, pig fern, bracken fern and New Zealand ice plant — were plug-planted into each treatment plot, four replicates each. EOS Ecology monitored the plots in detail across 2.5 years, tracking erosion features, product condition and coverage, and plant growth and vigour.
By the end of the monitoring period a clear hierarchy had emerged. The best-performing product treatment was the sprayed Proganics base with Flexterra HP-FGM topcoat and cover-crop seed mix – 91% of the ‘mainly loess subsoil’ grids supported more than 50% vegetation cover by 30 months, and the product itself retained 75–100% coverage in 93% of grids. The thick wool blanket came second on cover and equal-first on erosion attributes; the coir-plus-hydromulch combination came third. Among the plants, silver tussock (Poa cita) was the standout performer across all cover and vigour attributes, with Banks Peninsula fescue (Festuca actae) the next best. The New Zealand ice plant grew vigorously where protected but proved unusable in unprotected settings after wildlife cameras caught possums systematically grazing it – the trial had to substitute Carex comans mid-study. The core conclusion is that managing a loess cut slope effectively needs both a product and the right plant community – the product provides immediate erosion control while the plants establish, and the plants then provide the long-term natural solution. The results have been written up in a technical Year 3 Monitoring Report and a separately published summary document for public dissemination, so that the findings can be applied not just by councils on roadside cuttings but by private landowners around the Peninsula with their own cut-bank challenges.
Hakataramea River Willow Pest Management
A sub-centimetre-resolution UAV survey and image-classification baseline for the Hakataramea Sustainability Collective’s braided-river willow management programme – a repeatable, scientifically defensible record against which future control work can be quantitatively measured.
Braided rivers are one of New Zealand’s distinctive habitat types and a hotspot for indigenous biodiversity – but they’re also one of the country’s most rapidly degrading habitats, with exotic willow species progressively altering river morphology, outcompeting native vegetation, and crowding out the open shingle and braid habitat that braided-river specialists depend on. The Hakataramea River – a braided tributary of the Waitaki in inland South Canterbury – has extensive willow cover in its lower reaches, and the Hakataramea Sustainability Collective (HSC), the catchment community group leading the on-the-ground response, was running an aerial willow-control programme but lacked the quantitative baseline needed to demonstrate how effective the work was. HSC commissioned EOS Ecology to build that baseline using a precise, scientifically defensible and repeatable methodology – one against which the success of the willow control could be measured over time, and which the Collective could carry across the wider catchment as its management programme extends.
Across three days in mid-January 2024 – deliberately timed just four days before HSC’s scheduled aerial willow spray – we flew the 265 ha project management area with a DJI Phantom 4 Pro V2 UAV under Civil Aviation Authority Part 101 rules, capturing 5,876 nadir images at 100 m above ground level (2.7 cm/pixel resolution, significantly finer than the publicly available aerial imagery for the catchment). Twenty-two ground control points positioned with an Emlid Reach RS2 RTK GNSS receiver gave the resulting orthomosaic (generated in Esri’s Drone2Map) survey-grade positional accuracy, and a supervised image classification in ArcGIS divided the area into seven land-cover classes – two braided-channel categories (wet braids, dry braids), four vegetation categories (exotic woodland, dead woody material, grassland, mixed grass/scrub/shrubland), and a manmade category. We treated the area as a quasi-experimental study, splitting it into a 120 ha Downstream Active Management Area (where HSC’s willow control is being concentrated) and a 145 ha Upstream Control Area (left for comparison) – and the contrast was stark, validating the existing management focus: the downstream area carried 38.7% exotic woodland cover (46.6 ha of willows) against just 6.4% in the upstream control area (9.3 ha), and the subsequent 22 January aerial spray covered 27.4% of the downstream area against 1.7% of the upstream area.
The orthomosaic and classification together provide HSC with a precise, georeferenced baseline that any future UAV survey can be quantitatively compared against — change detection rather than qualitative impression. As HSC extends its willow management programme into other parts of the Hakataramea catchment, the same methodology can be applied at low marginal cost; and because the framework is repeatable, the Collective is now positioned to demonstrate efficacy to funders and partners using comparable hectare-and-percentage numbers across years rather than narrative reports of ‘things are getting better’. The methodology itself is broadly transferable – high-resolution UAV mapping paired with supervised image classification of this kind has direct application to wetlands, salt marshes, river riparian zones and any other context where land-cover change needs to be quantified at sub-metre precision over hectare-scale areas – and the EOS Ecology end-to-end capability spans CAA-compliant data acquisition, geospatial processing, and the science interpretation that turns the imagery into management-relevant decisions.