Whitebait are iconic in New Zealand. They comprise a culturally, commercially and recreationally important fishery when netted returning from their marine life phase. The five species comprising the New Zealand whitebait catch are inanga (Galaxias maculatus), koaro (G. brevipinnis), banded kokopu (G. fasciatus), giant kokopu (G. argenteus) and shortjaw kokopu (G. postvectis). Four of the five whitebait species are now ranked as ‘declining or nationally vulnerable’ and there is increasing concern that the fishery has declined considerably over the past few decades due to multiple stressors. Current management of the fishery is based on limited science, regulations are over 20 years old and treat the fishery as a whole apart from the West Coast of the South Island. There is limited information on species composition within the 2.5-3.5 month fishing season. This thesis examines these issues in a geographically-widespread study.

Whitebait were sampled from 92 rivers throughout New Zealand over six months (July to December) in 2015. A subset (8) of these rivers were sampled again during the 2016 whitebait season. Over 95,000 fish from approximately 420 samples were processed (20,000 of which were measured for morphological data). Regional and temporal variations were found in the physical characteristics used to identify species at the whitebait stage. Of the five whitebait species, inanga whitebait were easiest to identify from catches. Differentiating giant kokopu from koaro and banded kokopu whitebait proved difficult. Genetic analysis proved vital for accurate identification of shortjaw kokopu whitebait.

Inanga made up the highest proportion of whitebait in samples from across New Zealand, but koaro and banded kokopu made substantial contributions in some rivers and regions at particular times of the year. Buller was found to have the highest within-region variability in species composition due to the relatively high proportions of non-inanga whitebait. There was a strong positive association between the abundance of koaro and kokopu whitebait in samples with forest cover and unmodified land area. Latitude and coast (east vs. west) were found to affect whitebait length, with fish length increasing with latitude. Non-whitebait species observed in samples included smelt, freshwater shrimp, glass eels, adult eels, juvenile and adult bullies, yellow-eyed mullet, and lamprey.

The timing of giant kokopu and banded kokopu whitebait migrations was earlier in the North Island than in the South Island. Whitebait lengths, body depth, and condition varied throughout the six month period with initially lower lengths and condition increasing and then decreasing again. The peak in length and condition corresponded with the peak migrations.

The species composition of whitebait samples varied significantly in only one of the rivers sampled in successive years. In comparing the current composition to a study from 50 years ago, the whitebait catch included higher proportions of banded kokopu and lower proportions of koaro and inanga than previously. Examining the North Island, and east and west Coast of the South Island separately, there were higher proportions of koaro and banded kokopu, and lower proportions of inanga in 2015 than previously.

My study provides the first New Zealand-wide view of the morphology of whitebait, and new insights into the current species composition of the whitebait fishery. It shows there is substantial spatial and temporal variation in the species composition of the whitebait catch. Other key findings include the discovery of high proportions of non-inanga species in regions other than Buller and Westland and the first genetically confirmed identification of shortjaw kokopu whitebait, and giant kokopu in many rivers. These findings have important implications for freshwater fish conservation and management of the whitebait fishery.