Protected areas (PAs) have become the cornerstone management strategy for terrestrial and increasingly marine biodiversity conservation, yet freshwater protected areas (FPAs or freshwater reserves) remain rare. This is surprising considering that freshwater ecosystems, and the biodiversity they support, are among the most threatened worldwide. While rivers and lakes are more commonly used to define PA boundaries, some freshwater reserves that explicitly protect freshwater species exist. In New Zealand, whitebait reserves (areas closed to whitebait fishing) were established decades ago to protect key species (primarily galaxiids, called whitebait in New Zealand) comprising a culturally and economically important post- larval fishery. However, there has been no follow-up on the effectiveness of these whitebait reserves for conservation and fisheries sustainability since they were established. To address this, I compared whitebait reserve types established to prevent overexploitation of migratory galaxiids (mostly īnanga Galaxias maculatus) from fishing and evaluated their performance as reproductive reservoirs. I studied ten streams that were classified into three a priori types: (a) closed, (b) partially closed and (c) open (fished) streams along >500 km (3° latitude) of coastline. I found that closed streams are effective at protecting G. maculatus and demonstrated that whitebait reserves have many of the ‘reserve effects’ observed in commoner marine reserves, but there are also important differences.

Closed streams had greater abundances, biomasses and egg production of G. maculatus despite having a greater proportion of smaller fish compared to partially closed and open streams. This finding suggested that density-dependent processes may regulate G. maculatus populations in closed streams. Partially closed and open streams had greater richness of fish species than closed streams, and large, predatory fishes were more prevalent in closed streams. There was slight evidence that partially closed streams conferred some fisheries benefits compared to streams open to fishing.

Next, I applied the concepts of density dependence and the theory of compensatory density dependence to gain clarity on what the effects of population density were on individual growth rates and fecundity, and whether density-dependent or density-independent factors enhance or diminish the effects of whitebait reserves. In particular, I examined the additive and interactive effects of individual size, population density and stream temperature on individual growth rates and fecundity. I found that population density and stream temperature interacted to affect growth rates of post-recruit G. maculatus, but juvenile mortality rates in G. maculatus were not strongly affected by density. However, I did not observe negative DD effects on the fecundity of G. maculatus within whitebait reserves. This is a consequence of the amphidromous, annual, semelparous life-history of G. maculatus in New Zealand coastal waterways.

Overall, this doctoral research advances critical knowledge gaps on the effectiveness of whitebait reserves for conservation and fisheries sustainability of G. maculatus. It seems clear that remediation of in-stream and requisite riparian spawning habitats of the current whitebait reserves is necessary for substantially enhancing their reproductive benefits. Rehabilitation of feeding, foraging and spawning habitats will reduce the negative DD effects on individual growth, increase the sizes of spawning fish, increase per-capita egg production, and obviate the prevailing occurrence of ‘sink’ populations with no effective habitat for egg deposition. Such remediation would necessarily involve a more holistic view of whitebait reserves within the context of species’ requirements throughout their entire life-history.