The rāhui question at Te Rākau Bay

What the data do support

• The decline is large and not statistical noise at the longer timescale: 2019 (8.9, CI 7.7–10.1) and 2023 (4.8, CI 3.8–5.8) have non-overlapping confidence intervals, so a real fall of about 46% occurred.
• The last interval is not resolved: 2023 (4.8 ± 1.0) and 2024 (4.1 ± 0.9) overlap, so the further drop cannot be called a real change from these transects alone.
• Fishing effort at the ramp rose from 310 to 830 boat-days, about 2.7×, over the same period that abundance roughly halved. Effort and decline move together, and catch-per-unit-effort would be falling, which usually pushes fishers to fish harder to get the same bag - a reinforcing loop.
• Juvenile numbers collapsed from 5.6 to 1.5 per transect. That is the most serious signal, because recruitment is what refills the legal-size stock.

What weakens the causal claim

• The heatwave is a real rival explanation. It can cause starvation, reduced growth and failed spawning. But the comparison bay separates the signals: it fell only 10% (8.2 → 7.4) while Te Rākau Bay fell 54%. A regional heatwave should have hit both bays; the extra decline at Te Rākau Bay needs a local cause, and harvest is the obvious local pressure.
• The control is imperfect. The bays may differ in habitat, wave exposure, and larval supply; a heatwave can also make pāua less resilient to the same harvest pressure.
• Harvest is under-counted. Ramp counts miss customary take, shore-based gathering and poaching - so the true harvest is larger than 830 boat-days, and the correlation is probably stronger than measured.
• Correlation is not proof. Sedimentation, disease or a recruitment failure driven by oceanographic conditions are not excluded.

Judgement: the evidence supports a precautionary closure, but it does not prove fishing is the sole cause; the honest statement is that fishing is the most likely local driver acting together with the heatwave.

Measure against a pre-agreed baseline

• Keep the same 12 permanent transects and the same February timing, so the comparison is like-for-like. Record legal-size density, juvenile (<60 mm) density, and the full size-frequency distribution, not just a mean.
• Add a no-take reference site elsewhere in the rohe with similar habitat, so regional effects (another heatwave, a poor larval year) can be separated from the effect of the closure.
• Continue ramp surveys and add compliance monitoring, because a closure that is not observed is not a treatment.
• Log environmental covariates: sea surface temperature, sedimentation, swell exposure. Without these, a flat result could be blamed on the rāhui when the real cause was the environment.

Set success criteria in advance, e.g. juvenile density above 4 per transect and legal-size density above 6 per transect within five years.

Timing - the crucial judgement Pāua take 7–9 years to reach 125 mm here. A two-year closure therefore cannot be judged on legal-size abundance; those animals were already in the population when the rāhui began. The detectable early signals are:

• Juveniles: an increase in the <60 mm class within 2–4 years, since settlement and early growth are faster.
• Catch-per-unit-effort and diver observations at the reference site versus the closed bay.
• Size structure: more animals in the 60–110 mm band is the leading indicator that the pipeline is refilling.

Verdict: a two-year rāhui is defensible as a precaution, but evaluating it only on legal-size counts at the two-year mark will produce a misleading 'failure' result. Plan for juvenile and CPUE monitoring at 2–4 years, and a full stock assessment at 7–10 years.

Frame it as what is known, what is uncertain, and what follows either way

Known with confidence: legal-size pāua have roughly halved since 2019 (8.9 → 4.8 by 2023, non-overlapping CIs), juveniles have fallen to a quarter of their 2019 level, and ramp-recorded effort has risen about 2.7×.

Genuinely uncertain: how much of the decline is harvest and how much is the 2022/23 heatwave. The comparison bay (−10%) suggests the heatwave is not the whole story, but it does not give a number for the harvest share. Nor do we know the unrecorded customary and illegal take.

What the fisher is right about: the heatwave is real, documented, and a plausible contributor. Dismissing it would be poor science and would damage trust. But 'all from the heatwave' does not fit the comparison bay.

What the kaumātua is right about: a two-year window is short relative to the 7–9 years pāua take to reach legal size, and short closures can simply move effort elsewhere. A longer horizon with review points may be more honest than a fixed two years.

A defensible recommendation to put to the rūnanga

1. Close the bay now, on a precautionary basis - the direction and size of the decline, and the juvenile collapse, justify acting before certainty arrives.
2. Set review points at 2, 5 and 8 years, with the indicators agreed in advance, and commit to publishing the results either way.
3. Pair the rāhui with monitoring of the unrecorded take, and with a conversation about whether effort displaced to neighbouring bays needs managing.
4. State plainly in the notice that the heatwave is part of the picture; a rāhui that removes one pressure still helps a stressed stock recover, which is a defensible position without overclaiming.