Assessing the performance of pāua (Haliotis iris) fisheries using GPS logger data


Neubauer, P., Abraham, E., Knox, C., & Richard, Y. (2015). Assessing the performance of pāua (Haliotis iris) fisheries using GPS logger data. New Zealand Fisheries Assessment Report, 2015/71. 39 p. Retrieved from


The pāua Haliotis iris dive-logger data collection programme is an industry-led initiative aimed at achieving near real-time monitoring of the fishery. The dive loggers record the position, depth, and duration of individual dives, and allow the individual catch to be recorded. This detailed data recording is expected to allow the management of the fishery to be at finer spatial and temporal scales than is currently possible.

Here, we provide a first description of the dive-logger programme. We assessed the current coverage of the fishery by the logger programme by comparing dive-logger data with data reported on Pāua Catch Effort and Landing Return (PCELR) forms. To determine whether the dive-logger programme is relevant for fisheries management, we used Bayesian linear mixed models to gain an understanding of the relationship between logger-derived effort and catch data, both at large scales (using daily records from all Quota Management Areas (QMAs) and years) and small scales (catches and individual dives aggregated to 1-hectare (ha; 10 000 m2) hexagons).

We found that an increasing number of divers participated in the data-logger programme in the period from the 2010–11 to the 2013–14 fishing year. The catch recorded on data loggers also covered an increasing percentage of the total catch of the pāua fishery, including more than 50% of the total catch in QMAs PAU 7 and PAU 3 in 2012–13 and 2013–14. Some limitations remain, both with reporting (erroneous diver identifiers or catch reported) and with the logger hardware (e.g., missing locations); however, the quality of the data are improving as the coverage increases.

The data provided insights into dive patterns within QMAs, which, although variable, showed some consistent trends towards longer and shallower dives. Models at both large and small scales suggested that dive variables, especially bottom time (the time spent diving relative to the total fishing time), were strongly related to catch. At equivalent total effort (total fishing time), increased bottom times thus predicted increased catches. From the modelling, we recommend bottom time (relative to total fishing time) as a useful indicator of fishery status.

Spatial patterns in the fishery were also investigated using areas and effort concentration measures calculated from kernel utilisation densities (KUDs). On large scales, patchy diving within the total area searched over the day was correlated with lower catches, whereas on the hexagon scale, patchy diving was correlated with higher catches, possibly due to diving on pāua aggregations. We suggest that daily area and concentration indices could serve as indicators of trends in the fishery if monitored over time.

Overall, the logger data provide considerable information about the fishery, and are likely to provide fishery-relevant data that can be used to gain an understanding of patterns and trends in the fishery. As both the quality of the data and the volume of data increase, the proposed metrics will become more useful, and more complex metrics such as spatial catch-per-unit-effort (CPUE) can be constructed to identify local depletion. To illustrate this potential, we estimated a CPUE index at both statistical-area and 1-ha hexagon scales. This estimation showed up to fourfold differences in CPUE among statistical areas within QMAs.