Determining the source of green-lipped mussel spat landing at Ninety Mile Beach

Aotearoa New Zealand’s green-lipped mussel aquaculture industry is worth more than $380M a year, and the government’s aquaculture strategy includes a plan to expand the industry further over coming years. This industry is largely reliant on wild-caught spat (baby mussels): 80% of production uses spat collected from Te Oneroa-a-Tohe / Ninety Mile Beach. Year in and year out, the spat wash up on the shore attached to seaweed and other branching material, from where it is collected and transported to mussel farms nationwide for grow-out. The source populations (adult mussel beds) have not been identified, which poses risks to the long-term sustainability of the mussel aquaculture industry because it means the continued supply of spat cannot be safeguarded.

Backtracking spat landing at Ninety Mile Beach

Mussels spend up to six weeks as larvae and may be transported considerable distances by ocean currents. Our research uses a mixture of mussel genetics (determining the whakapapa or relatedness of mussels from different locations) and ocean models (determining the direction and strength of currents from particle tracking) to identify the likely source beds. Using the ocean models, we backtracked the spat from Te Oneroa-a-Tohe / Ninety Mile Beach to the known beds and found that the source populations are likely to be the beds off Tiriparepa / Scott Point, Ahipara and Herekino.

Mapping the connectivity between mussel beds

Mussel larvae swim up and down in the water column, which affects where they are transported. Using a sophisticated model that incorporates these behaviours, running for 10 years (2008-2017), we forward-tracked the movement of larvae between mussel populations in the wider area.

The results show that there are two blocks of connectivity off the western Northland coast, and that there is limited connectivity between the two. This means that recruitment is likely local – i.e., that self-settlement is very important for local mussel reefs. So, although mussel larvae can theoretically be transported hundreds of kilometres at sea during the month they are in the water column, we now know that for the west coast Northland populations they are not transported very far at all. Kaitaia spat are derived from local mussel beds not from distant beds.

Genetics confirm Kaitaia spat comes from local beds

Preliminary genetic analyses show that wild mussels from Te Oneroa-a-Tohe / Ninety Mile Beach and Kaitaia spat are very similar genetically but are different from mussels from other regions (Cook Strait, Christchurch, Wellington, and Raglan). In conclusion, genetic analyses and physical oceanographic modelling both clearly indicate that Kaitaia spat are derived from local (e.g., Ahipara, Tiriparepa / Scott Point) mussel beds and not from distant (Raglan) or far distant (Cook Strait, Wellington Harbour, Christchurch) mussel beds. This is the first clear demonstration of the source of Kaitaia spat.

Paua connectivity

Blackfoot paua, Haliotis iris, are endemic to Aotearoa New Zealand and support a large commercial and recreational fishery. Wild stocks of many abalone species around the world have collapsed and their cautionary tales show why it is necessary to carefully manage the resource to maintain sustainable exploitation. Major gaps in our knowledge of the biology of Haliotis iris remain, especially regarding their demographic structure, and a better understanding of larval dispersal and wild population connectivity would benefit current management strategies.

Estimated paua population structure
Dashed lines represent previously identified barriers to gene flow. Using models to simulate the dispersal of paua larvae between populations. Combining ocean models and genetic analysis we estimated blackfoot paua population connectivity. We used the numerical ocean model to track paua larvae trajectories between 464 populations over 24 years (1993-2017). To validate the model estimates, we used results from previously published genetic studies.

Maps of connectivity and population structure

The model shows that larvae settle close to their starting location, most often within the first 10 km. However, all the populations remain well-connected through stepping-stones except at the Chatham Islands, which are isolated from other paua populations due to their remoteness. Within mainland New Zealand, we identified three separate areas where local populations are very well inter-connected: the Cook Strait, the Bay of Plenty, and adjacent to Steward Island.

The model estimates match with genetic similarities from sampling sites around the country. Mapping the demographic structure based on estimated connectivity, we grouped the paua into 18 subpopulations (figure on right), matching most of the barriers to gene flow identified in previous studies (dashed lines on figure).

Impact of fronts on population structure

An analysis of the oceanic fronts around Cook Strait and East Cape helps us understand how the demographic structure is formed. We found three recurring fronts over a period of 10 years. The first front wraps around the top of the South Island and prevents larvae spawned in PAU7 from reaching Taranaki populations, creating the barrier observed in the genetic study. The second front helps larvae transition from PAU3 (Kaikoura) to PAU2 on the North Island. The third front corresponds to the weak barrier identified at East Cape.

In conclusion, we mapped the connectivity between paua populations in Aotearoa New Zealand and validated the model estimates with previously published genetic dataset. We found that the population is composed of 18 subpopulations and is strongly influenced by the presence of recurring oceanic fronts and the distance between suitable reefs.