Library Open Repository

Growth of abalone (Haliotis rubra) with implications for its productivity

Downloads

Downloads per month over past year

Helidoniotis, F (2011) Growth of abalone (Haliotis rubra) with implications for its productivity. PhD thesis, University of Tasmania.

[img] PDF (Whole thesis)
Fay_thesis-.pdf | Download (2MB)
Available under University of Tasmania Standard License.

Abstract

The use of an incorrect growth model in fisheries management may lead to inaccurate predictions about stock productivity. In Australia, three non-nested size-based growth models are generally used to describe the growth of abalone populations: the von Bertalanffy, Gompertz and inverse logistic. The models differ in their description of growth, especially in the juvenile phase. However, while data on juveniles has the greatest discriminating power between models, in reality good data on size distributions and growth of juveniles is uncommon, and this leads to ambiguity in model selection. I use a large dataset (from the Tasmanian Aquaculture and Fisheries Institute) describing sizes and growth of juvenile and adult size classes to systematically resolve model ambiguity for blacklip abalone (Haliotis rubra) populations in Tasmania. Modal progression analysis of bimonthly data collected over two years from the same site identified two cohorts of juveniles between 10 – 75 mm shell lengths. The best statistical model was selected using standard statistical model selection procedures, i.e. Akaike’s Information Criteria and likelihood ratio tests. Despite the large data set of 4,259 specimens, model selection remained statistically ambiguous. The Gompertz was selected as the best statistical model for one cohort and the linear model for the other. Interestingly, the biological implications of the best fitting Gompertz curve were not consistent with observations from aquaculture. The study revealed that slight differences in data quality may contribute to ambiguity in statistical model selection and that biological realism is also needed as a criterion for model selection. The robustness of different growth models to sampling error that is inconsistent between samples was explored using Monte Carlo simulation and cross model simulation. The focus was on simulated length increment data largely from adult size classes (55 – 170 mm shell length) as these data are more commonplace than data from juveniles. Results confirm that the two main shortcomings in length increment data contributing to model misspecification were (i) poor representation of juvenile size classes (< 80 mm) and (ii) low sample size (n<150). Results indicate that when negative growth data are included in the von Bertalanffy model, K increases and L∞ decreases. In reality the true description of growth remains unknown. Given realistic length increment data, there is a reasonable probability that an incorrect growth model may be selected as the best statistical model. This is particularly important, because this study indicates there is a different magnitude of error associated with each growth model. The important overall finding is that while it is possible to make incorrect model selections using customary statistical fitting procedures, departures from biological reality are lower if the incorrect inverse logistic model is selected over the incorrect von Bertalanffy or Gompertz model. The selection of the most appropriate growth model was further tested by fitting each of the three growth models to length increment data from a total 30 wild populations. The inverse logistic was the best statistically fitting model in 23 populations. The combined results from data on the growth of juveniles, cross model simulation, and fitting to data from numerous wild populations systematically revealed that the inverse logistic model was the most robust empirical representation of blacklip abalone growth in Tasmania. With this confidence in the selected model, it was then possible to address two urgent ecological and management issues related to stock productivity; the effect of climate change on growth rates and the success of broad-scale management controls in the presence of fine-scale variability in growth rates.The effect of ocean warming on the growth rates of blacklip abalone populations was explored from the analysis of length increment data from 30 populations across a range of water temperatures. Measurements based on the growth rates of juveniles did not reveal a clear negative relationship between temperature on growth. A decrease in growth rate was observed however it may not be directly attributable to temperature but may be forced by the onset of maturity, which does appear to be directly influenced by temperature. Fine-scale estimates of growth rate are an implicit aspect of evaluating the success of broad-scale management control such as Legal Minimum Length (LML) for harvesting. In reality, it is not possible to obtain fine-scale growth rates given the expense of obtaining empirical length increment data at fine spatial scales. Therefore, an alternative approach was developed that exploited the correlation between the parameters of the inverse logistic model and size at maturity. The approach generated theoretical, fine scale growth parameters and population-specific LMLs for 252 populations around Tasmania. Using population specific size limits, results revealed that 46 populations were unprotected by the current Legal Minimum Length (LML) settings, potentially exposing those populations to overexploitation. The majority of unprotected populations were located in the south west, a region that is economically valuable. An important recommendation from this thesis is that the LML of the economically valuable south-west region should be increased in order to achieve the management goals of the fishery.

Item Type: Thesis (PhD)
Keywords: growth models, abalone, climate change, productivity
Additional Information: Copyright 2011 the Author
Date Deposited: 13 Sep 2011 06:32
Last Modified: 01 Aug 2012 01:05
URI: http://eprints.utas.edu.au/id/eprint/11748
Item Statistics: View statistics for this item

Repository Staff Only (login required)

Item Control Page Item Control Page