Abstract
Life history theory is underpinned by the allocation of finite resources to competing functions. This is most clearly illustrated in comparisons between species which allocate all resources to a single reproductive event (semelparous), and those which spread reproductive effort over time (iteroparous). Galaxias maculatus is a widespread Southern Hemisphere freshwater fish with a highly plastic life history. Populations have been observed to display varying modes of parity from total semelparity to abbreviate iteroparity. The factors influencing this variation remain poorly understood. This thesis investigates the linkages between life history and energetics in G. maculatus.
The life history characteristics of two populations of G. maculatus in southern New Zealand were assessed. The duration of the spawning season, timing of recruitment, and size at maturity was observed to vary across a small spatial scale. Population size-frequencies indicated that both populations were essentially annual, with a far lower frequency of large (presumed 1+) individuals, than observed for other New Zealand populations. Despite this, fish with spent gonads were observed, indicating the capacity for post-spawning survival.
Using laboratory feeding trials, a weight- and temperature-dependent consumption model was developed for G. maculatus. Daily maximum consumption was modelled between 5-25 ℃ for juvenile and adult G. maculatus. The consumption model was combined with published data on respiratory costs for the species, to parameterise a bioenergetics model for G. maculatus. Controlled growth trials at two ecologically relevant ration and temperature levels indicated that the bioenergetics model adequately predicted consumption and growth. Model uncertainty increased with increasing ration size, which reflected increased variance in individual growth at higher rations.
Temporal variation in feeding was assessed in a wild population of G. maculatus. In the period immediately prior to spawning, fish with high gonadosomatic indices had significantly lower gut fullness, indicating that reproductive investment negatively influenced energy acquisition. Monthly estimates of gut fullness were combined with laboratory-derived estimates of gastric evacuation to assess seasonal variation in daily ration. Daily ration was lowest in April, immediately prior to spawning, despite substantially higher water temperatures than in May or June.
Body energy density and tissue proximate composition of wild G. maculatus varied markedly between seasons. Perivisceral fat stores were completely utilised in the lead up to spawning and did not increase post-spawning. On a relative weight basis, males invested more energy into gonad development than females. However, due to differences in gonad energy density, females invested a higher proportion of body energy into their gonads. Females lost an average of 46 % of their total body energy to reproduction. Changes to the population frequency distribution of somatic energy density were observed between May and June, indicating postspawning mortality is influenced by residual somatic energy stores.
The ability of the bioenergetics model to predict the growth and prey consumption of wild G. maculatus was assessed. Model predictions of growth and cumulative consumption over the entire 188 day model test period were accurate, with percent errors of +13.4 % and -4.5 % for growth and consumption respectively. Post-spawning fish were predicted to require very low consumption rates to maintain body weight over the winter period, indicating that over-winter energy limitation is unlikely to increase mortality in surviving post-spawning individuals. The relatively small size post-spawning fish were predicted to reach by the following summer indicates that reliance on size-frequency data to estimate age may underestimate the proportion of 1+ fish in the population.