Fish scales as ecological indicators: empirical approaches to improve their practical application to fish ecology.

Abstract

The collection of scales is common when fish communities are sampled within research and monitoring programmes in freshwater fisheries. Although used primarily to age individual fish, there is increasing evidence of their potential for application to other ecological methods, yet there is also considerable uncertainty in how this can be achieved. Thus, the purpose of this research was to examine how the use of scales within age and growth studies can be enhanced and investigate their application to freshwater trophic ecology, with a particular focus on advancing their utilisation within stable isotope analysis (SIA). The research used fishes of the Cyprinidae family as the focal species. Cyprinids are of substantial global, socio-economic importance as their communities are valued ecologically, commercially and recreationally. The research assessed current methodologies, highlighted extant gaps in knowledge and sought to resolve these issues. It covered work regarding the intrinsic error contained in estimating fish age from scales and improved scale application within SIA through the provision of key data that is currently lacking within the literature. An initial invasion ecology case-study provided new insights into the growth and trophic impacts of a model native and two non-native fishes under three distinct approaches of differing spatial scale and complexity. The results revealed a range of ecological consequences for the native species from the invaders, although the extent of these was also a function of spatial scale. Additionally, a number of procedural concerns relating to the collection of fish age data and current SIA methodologies were highlighted. The use of scales to derive estimates of the ages of fish is well established, with outputs used to address questions on aspects of fish and fisheries ecology, but the process remains prone to inherent errors. The research revealed that precision of growth estimates is significantly influenced by the sub-sampling regime applied. Where individuals are long-lived and slow-growing, sub-sampling strategies that result in few scales being analysed produced imprecise data and potentially erroneous outcomes. Additionally, uncertainty in the accuracy of ageing scales also potentially results from subjective interpretation of scale features. A statistical model was developed to incorporate this uncertainty into analyses, using Bayesian statistics and a bootstrapping methodology, to improve age and growth rate estimates. The model successfully produced error adjusted von Bertalanffy growth parameters. Food web and trophic analyses have traditionally been completed through stomach content analysis, but increasingly SIA is preferred, as it provides greater temporal perspectives and requires smaller sample sizes. In fish studies, dorsal muscle tissue is typically favoured, but this is often collected destructively. The research revealed that non-destructively collected tissues, such as scales, can act as a proxy for muscle and their isotopic values can be converted with minimal error when species-specific factors are used. When stable isotope data are applied to dietary studies, their use in statistical mixing models requires accurate step-wise enrichment values between diet and consumer (i.e. discrimination factors). There is considerable uncertainty in the variability of discrimination factors between species and the influence of their diet. Consequently, specific diet-tissue discrimination factors were produced for a range of cyprinid species and diet was shown to significantly affect diet-tissue discrimination. The application of species-specific values within mixing models can result in significant differences when compared with using standard values and consideration of the influence of diet needs to be made when investigating omnivorous species. The rate of turnover of carbon and nitrogen stable isotopes was also determined and variability between tissues was revealed, indicating that species- and tissue-specific half-lives should be considered when deciding upon experimental time-frames. In summary, the research has provided substantial information targeting extant knowledge gaps relating to the application of scales from cyprinid fishes to ecological studies. Regarding fish age and growth, issues surrounding accuracy and precision of estimates has been tackled, informing researchers of the influence on precision of applying sub-sampling regimes to subsequent growth analyses and providing an original statistical tool that can improve accuracy through producing growth parameters that better reflect inherent errors in fish age data. In contributing to the use of scales in SIA, novel data have been provided that will reduce the requirement for destructive sampling of fishes and enhance present understandings of the significance of species- and tissue-specific discrimination factors and turnover rates.