Factors controlling common smelt abundance and rainbow trout growth in the Rotorua Lakes, New Zealand
Blair, J. M. (2012). Factors controlling common smelt abundance and rainbow trout growth in the Rotorua Lakes, New Zealand (Thesis, Doctor of Philosophy (PhD)). University of Waikato, Hamilton, New Zealand. Retrieved from http://hdl.handle.net/10289/6941
Permanent Research Commons link: http://hdl.handle.net/10289/6941
Recreational fisheries are of considerable ecological, economic and cultural importance worldwide, and understanding the factors that influence them is an important goal of fishery managers. The two most important factors influencing the growth of fish are habitat quality and food supply. This study was undertaken to gain a better understanding of how these two factors influence rainbow trout growth in the Rotorua Lakes, central North Island, New Zealand, by surveying prey dynamics, quantifying rainbow trout diet, and assessing the effects of environmental factors and population density on trout growth. Rainbow trout populations in lakes of New Zealand’s central North Island mainly prey upon common smelt (Retropinna retropinna), a small, widespread, pelagic species. Routine monitoring of smelt populations would assist in trout fishery management decisions, especially for optimisation of stocking rates and timing. To recommend an effective capture technique for smelt, we tested purse seining, beach seining, drop netting, and boat electrofishing in Lake Rotoiti. Boat electrofishing in the littoral zone at night allowed us to collect more adult smelt than all other methods, and long boat electrofishing transects across the lake were the most effective method for capturing a wide size range of smelt specimens, including larvae. Most methods also caught other species as bycatch. In terms of the number of smelt caught, the most effective method was (1) boat electrofishing at night, followed by (2) scoop nets at night (i.e. boat electrofishing with the current turned off), then (3) beach seining during the day. The first two sampling methods are suggested as the best methods for collecting data on seasonal dynamics, growth, and reproduction of smelt. For enumeration of the total smelt population, these methods should be carried out in the littoral and pelagic zones and combined with a method such as hydroacoustics that can enumerate smelt in the deeper waters of the pelagic zone. Population dynamics of common smelt in lakes are not well understood. To determine the factors influencing smelt populations in Lake Rotoiti, we examined seasonal changes in habitat and resource use by measuring spatial and seasonal changes in smelt abundance, body condition, and diet. We also characterised seasonal and spatial changes in smelt reproductive state and benthic egg abundance. Smelt abundance in the littoral zone and surface waters of the pelagic zone was highest in autumn, coinciding with peaks in smelt body condition, gonadosomatic index, and benthic egg abundance. Smelt larvae were only found in the pelagic zone, and were more abundant in spring than in summer and autumn. Body condition of smelt varied throughout the year, and was lowest in winter, a period of low abundance of littoral invertebrates and zooplankton. Smelt caught in the littoral zone during the day consumed a range of benthic and pelagic invertebrates and smelt eggs, while at night, smelt caught in the littoral and pelagic zones consumed zooplankton, smelt eggs and larval common bully (Gobiomorphus cotidianus). The amount of food in the stomach relative to smelt mass was higher in the littoral zone than in the pelagic zone, suggesting that food resources in the littoral zone exceed those in the pelagic zone. Predation on zooplankton was highest in winter and spring, and smelt eggs formed a large proportion of smelt diet in autumn and winter. Stomach contents and stable isotope analyses showed that smelt undergo an ontogenetic change in diet, from mainly zooplankton as juveniles to mainly benthic invertebrates as adults. The information obtained in this study is necessary for managing smelt and their predator—rainbow trout—in lakes. To better understand the prey requirements of trout in central North Island lakes, we characterised seasonal and ontogenetic changes in diet and prey energy density of rainbow trout in Lake Rotoiti. Common smelt was the dominant prey item of rainbow trout larger than 200 mm (77.8% of diet by mass), followed by koura (freshwater crayfish Paranephrops planifrons; 6.3%), common bully (5.5%), and koaro (Galaxias brevipinnis; 3.4%). Juvenile rainbow trout (<200 mm) consumed amphipods, aquatic and terrestrial insects, oligochaetes, tanaid shrimps, and smelt. Trout consumed koaro only in autumn and winter; consumption of other species did not vary seasonally. The maximum size of smelt consumed increased with increasing trout size, but trout continued to consume small smelt even as large adults. Consumption of larger prey items (koaro and koura) also increased with increasing trout size. This study indicates the importance of smelt for sustaining rainbow trout populations, as predation on other species was relatively low. These findings provide a basis for bioenergetics modelling of rainbow trout populations in lakes of the central North Island of New Zealand. Though the factors influencing the growth of salmonids in cool-temperate and boreal climates are well understood, we lack an understanding of the influences on salmonid growth in warm-temperate areas, especially in lakes. To determine the combined effects of environmental factors, including habitat, on rainbow trout growth, we investigated the growth patterns of rainbow trout (Oncorhynchus mykiss) in nine warm-temperate New Zealand lakes of contrasting morphometry, mixing regime, and trophic state. Mark-recapture data (some collected by anglers) from hatchery trout releases during eight consecutive years were used to calculate growth parameters and body condition factor. Fish growth rates and condition factors were highest in deep (≥20 m mean depth) lakes of moderate trophic state. Overall, growth rate increased with increasing lake size and volume of favourable habitat (i.e. dissolved oxygen >6.0 mg L-1 and temperature <21°C), but decreased with increasing turbidity, chlorophyll a, and nitrogen concentrations. A classification and regression tree (CART) analysis found that variables describing habitat volume were the most important determinants of trout growth rate, and correlates of trophic state (chl a, conductivity) were important secondary determinants of trout growth rate. These results suggest that lake morphometry and trophic state are important attributes structuring overall habitat quality and thus influencing growth of rainbow trout in lakes in warm-temperate climates. The effects of future ecosystem degradation and climate warming on trout growth are likely to be most severe in shallow, eutrophic lakes. To investigate the carrying capacity and factors affecting growth of rainbow trout in Lake Rotoiti, we employed a bioenergetics model to assess the influence of past stocking rates, timing of releases, and prey abundance on growth and prey consumption. We hypothesised that stocking rates and prey abundance would affect growth and prey consumption by influencing per-capita prey availability, and that the environmental conditions encountered by fish at the time of stocking would affect growth and consumption. Prey consumption of stocked rainbow trout was calculated with the Wisconsin bioenergetics model. We calculated growth trajectories based on data from trout that were stocked into the lake in spring and autumn from 1995 to 2009 and then re-captured by anglers. Diet, prey energy density, body mass lost due to spawning, and lake temperature were measured locally. There was no difference in tag return rate between fish released in spring and autumn. Though trout released in autumn were smaller initially, they grew at a faster rate than trout released in spring. The ratio of observed to predicted change in biomass at the maximum consumption rate by individual trout in the first year of lake residence was negatively correlated with the number of yearlings released in a cohort, suggesting that stocking rates (347–809 fish ha-1 year-1) caused density-dependent effects on growth. Common smelt accounted for 85% of total prey consumption. However, no significant relationship was found between prey consumption by individual trout and adult smelt abundance, suggesting that more detailed investigations of smelt abundance are required to predict trout growth rates. Because there is little risk of non-human predation of stocked fish in Lake Rotoiti, and winter temperatures are mild (11–14°C), stocking smaller trout (~160 mm fork length) in autumn is likely to produce larger fish than stocking larger fish (~200 mm fork length) in spring. Possible reasons for this difference include higher prey abundance in the littoral zone in autumn and more suitable temperature and dissolved oxygen habitat in autumn-winter. These results suggest that optimal stocking strategies in warm-temperate systems may differ to those in cooler temperate regions.
University of Waikato
All items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.
- Higher Degree Theses